Distributed control system initialization method

Description: TECHNICAL FIELD [Detailed description of the invention]

【0001】

[Industrial application field]

The present invention relates to a distributed command and control system used in transportation, and more particularly, an address unique to each general-purpose "node" in a distributed control system including a plurality of general-purpose nodes combined with the same distributed communication bus. Regarding how to assign.

【0002】

[Conventional technology]

A common problem in the transportation industry is that distributed communication bus systems have many generic nodes. It can be electrically connected to node). When a distributed communication bus interconnects many general-purpose nodes, it is difficult to assign a unique address to each node in the distributed system. Since it is difficult to specify unique addresses for many general-purpose nodes that are interconnected together, in many electrical systems for transportation applications, each node or member is individually programmed before being placed in the system. Each node can be easily and uniquely identified by modifying each IC or node before it is placed in the system, but these advantages are costly. If the device is programmed with one address and then assembled, a human error could result in the device programmed as "Device A" being misplaced in the wrong place for "Device B" and malfunctioning. It can occur. Such malfunctions can make it impossible to perform easy and cost-effective tests, as well as when testing automotive airbag systems. In addition, the overhead of tracking individually addressed elements prior to assembly can be a significant burden. For example, suppose 16 generic elements are required in the same communication bus system. If these 16 elements are generic parts, then all 16 elements can be cataloged, stored and used as one interchangeable element. If each of the 16 generic elements had to be tagged individually, the process of storing, reusing, repairing, replacing and assembling the system would be significantly more laborious and prone to human error. Become.

【0003】

As an alternative to tagging elements uniquely before assembly, each generic element can be uniquely identified using sockets or physical interconnects uniquely configured for the element. Although it is in a different format than the pre-programmed solution as described above, it has the same disadvantages and risks.

【0004】

In some cases, the distributed communication bus system is circumvented and a point-to-point communication system is selected.In the former, N nodes are each N different and mutually exclusive leads. By one, it is directly coupled to the central controller. While this guarantees unique access to each node, this approach also has many drawbacks. Such N conductive interconnects can quickly become complicated in the design of transportation equipment, and the design cost thereof may become extremely high. Adding wiring and interconnects can significantly increase the weight of the system or vehicle. The density of interconnects within the chassis of the car is limited, and other, more important features take precedence. Also, N connections can require N pins on one IC, and for ICs, the number of pins is the number of pins in many microcontroller and digital signal processor (DSP) designs. Strictly restricted. In general, this powerful approach of connecting unique lines to each node in the system is rapidly becoming impractical.

【0005】

[Problems to be Solved by the Invention]

Generally, in the transportation equipment industry and elsewhere, distributed communication bus architectures and methods that allow general purpose nodes or components to be interconnected via a common communication bus connection while still allowing unique addressing or tagging of general purpose members. On the other hand, there is a need.

【0006】

[Example]

For clarity, the following description will focus on embodiments of the present invention in the vehicle. Examples of vehicles discussed below are shown in FIGS. 1-11. The present invention is applicable to any transportation system having a distributed control and command system and is not intended to be limited to the automotive embodiments described herein. In the distributed command and control system of the present invention, the method of initializing the address of a generic node is not specific to one particular node function and initializes any number of systems as specified below. Can also be used to do.

【0007】

FIG. 1 shows a conventional transport electrical system 10. The system 10 includes, by way of example, a vehicle body system 12, a powertrain system 14, and has a gateway connection 16 that allows selective communication between the powertrain system 14 and the vehicle body system 12.

【0008】

The body system 12 can include any number of subsystems configured as described above. Three specific subsystems are shown in Figure 1 as the airbag system 18, the heater / ventilator / air conditioner (HVAC) system 20 and the power seat system 22.

【0009】

In FIG. 1, the airbag system 18 is configured by an airbag controller 24 to monitor all sensors in the airbag system 18 and control all actuators in the system 18. The airbag controller 24 is coupled with four general purpose airbag actuators 26. Each actuator 26 is general purpose, so if all actuators 26 are coupled together on one common bus, one particular actuator 26 will be uniquely selected to eliminate all other actuators 26. It is not possible. To overcome this, a unique and exclusive line is coupled between the airbag controller 24 and each actuator 26. By using separate leads, the airbag controller 24 can control each actuator 26 independently of all other actuators 26. Also, within the system 18, the airbag controller 24 is directly coupled with two general purpose collision sensors 28.

【0010】

The airbag system 18 requires a large amount of wiring to connect between the airbag controller 24, the actuator 26, and the general-purpose collision sensor 28. Due to such a large amount of wiring, the manufacturing cost and the weight of the automobile increase. In a system such as an automobile chassis, such a large amount of wiring may not be possible due to the limited density of possible interconnections. New designs require significant hardware changes, which can make it difficult to reuse the design. New designs that add functionality, such as the addition of sensors or actuators, may exceed the number of pins available on the controller 24, and the addition of pins also increases the cost of the controller 24.

【0011】

As shown in FIG. 1, system 10 also includes HVAC system 20, which is a distributed command and control system. The HVAC system 20 includes an HVAC controller 30 and five actuators / sensors 32A, 32B, 32C, 32D, 32E connected to the common communication bus 33. When the HVAC system 20 is fully assembled, the actuators / sensors 32A, 32B, 32C, 32D, 32E are identified as not being identical to each other. This can be done using sockets that are physically or electrically different from each other, for example, each socket has a different address pin connection, and each of the actuators / sensors 32A to 32E has a different address. Can be assigned. When using such a unique socket scheme, human error during assembly can cause the HVAC system 20 to fail. Alternatively, the actuators / sensors 32A-32E can be electrically programmed prior to assembly so that the actuators / sensors 32A-32E each have a different address. In this case, the actuators and sensors 32A-32E are no longer universal, and the manufacturer of the transportation equipment must track, store and adjust the five independent components. Therefore, the method illustrated in HVAC system 20 increases the risk of malfunction of the HVAC system by increasing the cost of the manufacturing process and increasing the chances of human error occurring. Repairing, replacing and redesigning systems that include multiple different components in this way is also more cumbersome.

【0012】

FIG. 1 also shows a power seat system 22, which is a control and command system, which has a power seat controller that is uniquely coupled to a plurality of general purpose actuators 36 via separate lines. In addition to the system shown in Figure 1, there are many subsystems and electrical architectures in transportation equipment design that reduce costs, reduce interconnection complexity, and reduce weight. It can be optimized to increase flexibility, reduce the likelihood of human error, and improve the overall performance of the automotive electrical system.

【0013】

Generally, the present invention is a method of interconnecting a plurality of general-purpose nodes into one communication bus structure, in which the nodes are integrated circuits (ICs) and members, which are CPUs, sensors, actuators, Includes electrical, optical or electromechanical devices suitable for switches, or other transportation equipment applications. The main node, the CPU of the system, provides each generic node with a unique address. Such addressing is done after assembly, but when first assembled into the system, the nodes were generic. According to one embodiment, the methods and architectures described here interconnect generic nodes in a tree structure (Figure 2). In another embodiment, the general purpose nodes are interconnected in a ring structure (Fig. 3). The present invention provides a method of identifying each general purpose node, in which case all the nodes are interconnected along one communication bus.

【0014】

In the prior art, the difficulty of addressing general purpose members, as in the HVAC system 20 in FIG. 1, was overcome by making each actuator 32 a different member before assembling. Each actuator 32 was electrically programmed or physically interconnected so that it could be distinguished within the system 20. The airbag system 18 in Figure 1 shows an alternative way to differentiate generic nodes. Here, each general-purpose actuator 26 is uniquely addressed by an exclusive connection to the central controller 24. However, such an interconnect scheme significantly increases the cost of the system and increases the density of interconnects inside the vehicle.

【0015】

The systems and methods described below connect multiple generic nodes together to one identical common bus. The connection is made at the time of physical assembly. During the initialization period, this method determines the distance of each node from the main node. To make this determination, measure the voltage potential, current measurement, optical measurement, temperature measurement, resistance, or parameter change along the length of the communication bus to determine the distance of each node from the main node. It may require a decision-making step.

【0016】

In a distributed control system with a controller and at least one node coupled to one common communication bus, the distributed control system is when each node is at a unique distance measured from the central controller along the communication bus. It has a predetermined topology, and the distance information provides a unique identifier. In one embodiment, the method of initializing the address is to determine the distance from the controller to each node, and this distance represents the relative position of the node in a predetermined topology, and is predetermined. Includes the step of assigning a specific address to the node based on the relative position of the node in the topology. The measurement is performed at each node and the measurement information is stored. The main node assigns an address to each node based on the measurement information. This measurement information can also be adjusted to correct the error. In this way, subsystems with multiple general purpose nodes, often used in transport equipment applications, are designed using a single communication bus that interconnects the general purpose nodes.

【0017】

Using only one communication bus not only reduces manufacturing costs, but also reduces the density of interconnects and reduces the weight of the vehicle. With such a design, it is easier to add new nodes to the system. It can also be designed with general purpose nodes, which allows elements to be easily tracked during manufacturing and replaced for repair and replacement. Also, since all nodes connected to the communication bus are general purpose, human error due to misplacement of elements in the system is eliminated. Summarizing the above, assigning addresses within a distributed system using the unique distance of each node provides a great advantage over the prior art. While automotive applications appear to be the primary application, many other distributed systems, such as home wiring and telecommunications, may benefit from this technology.

【0018】

FIG. 2 shows one type of distributed system with general purpose nodes that can be used according to the present invention. The network of FIG. 2 is called a tree-like network because the communication bus 41 is a single wire that does not take a closed shape. Figure 2 shows the main node 40, which is an electronic control unit (ECU), digital signal processor (DSP), microprocessor, microcontroller device (MCU), custom integrated circuit or similar central processing unit (CPU). ) Or a combination thereof. The main node controls the operation of the node 42 and coordinates the communication between the nodes 42. In FIG. 2, the nodes 42 are all general-purpose members and are coupled to the central communication bus 41.

【0019】

Node 42 is typically purchased as the same functional element. The central communication bus 41 operates using any communication bus protocol and can include one or more leads, but it is desirable to operate according to the communication bus standard of the transport equipment. The node 42 is a general-purpose device that cannot be distinguished from each other except for its position along the physical length of the communication bus 41. Note that you can have multiple types of nodes along a communication bus. The function of each node 42 in the system is determined by its location. For example, if the airbag system is configured as shown in FIG. 2, each node 42 can be an actuator in a specific position. In this case, the node 42 located in the driver's side airbag responds to a different stimulus than the node 42 located in the steering wheel airbag. Node 42 includes actuators such as squibs used to fire airbags or step motors, sensors such as accelerometers or pressure converters, communication devices, and other microcontrollers (MCUs). It can be a DSP, a switch, or any other node along an interconnected network.

【0020】

FIG. 3 shows an alternative distributed system called a ring configuration, which has a main node 31 connected to the communication bus 43 to form a ring. A plurality of general purpose nodes 42 are then coupled to the communication bus 43, as shown in FIG. Similar to FIG. 2, the main node 31 is a central controller or some intelligent device that controls the operation of node 42 and / or monitors node 42. The nodes 42 are all general-purpose devices, except that the physical arrangement on the communication bus 43 is different for each node 42. Note that you can have multiple types of nodes along a communication bus. For example, actuators and sensors are expected to be configured along a single communication bus.

【0021】

In FIGS. 2 and 3, the generic node 42 knows its position, distance or proximity along the communication bus 41 or communication bus 43 and can be used for unique tagging or unique identification. is there. Unless there are two nodes along the respective communication buses 41 and 43 that are the same distance from the main nodes 40 and 50, each node 42 can be individually identified and given a unique address individually, as shown in FIG. And each generic node in the system in Figure 3 can be uniquely and individually accessed to eliminate all other nodes. Also, this address and the location of node 42 in the system are usually related to the function of the device in the application of transportation equipment. For example, if the main node 31 is located on the driver's side dashboard, the driver's front airbag is likely to be the closest device to the main node 31. However, the collision sensor at the rear of the vehicle is likely to be farthest from the main node 31 in the dashboard on the driver's side. Therefore, the main node 31 addresses the node that is determined to be the closest by knowing the design structure as the front airbag of the driver's seat, and the node that is determined to be the farthest, for example, It can be addressed and used as a rear collision detector. To summarize the above, when the main controller knows the layout or topology of the communication bus design of the transportation equipment, the distance of the node can be easily correlated with the function of the node.

【0022】

4 to 10 show (1) how the design of FIGS. 2 to 3 can determine the placement of general purpose nodes along one or more of the communication buses 41,43, and (2) the communication bus. After the distance above is determined, we will show how to assign a unique address that specifies the function to the generic node.

【0023】

FIG. 4 shows an automobile airbag electrical system 50 that is physically located around the chassis of an automobile. FIG. 4 shows the airbag electronic control unit (ECU) 52, which is similar to the main nodes 40 and 50 in FIGS. 2 and 3, respectively. The airbag ECU 52 of FIG. 4 is coupled with one ring-shaped communication bus 53 as shown in FIG. 3 and two tree-shaped communication buses 55 and 57 as shown in FIG. 2, respectively.

【0024】

In ECU 52, starting from the left side of FIG. 4 and proceeding clockwise along the ring-shaped communication bus 53, the ring-shaped communication bus 53 first interconnects the two actuators 54, which are shown in FIG. 2 and Similar to node 42 in Figure 3. The first two actuators 54 in FIG. 4 control the deployment of passenger front airbags. Two actuators 54 are used to operate one airbag, in which case one actuator triggers the airbag in response to a slow collision, according to one embodiment. , Other actuators or both actuators are used during faster collisions. Here, these actuators are separated by a wire loop 58.

【0025】

As shown in FIG. 4, any number of actuators 54 are similarly separated by the wire loop 58. The wire loop 58 increases the length of the communication bus 53 to effectively increase the distance between the two closely spaced actuators 54. In this way, the distance from the ECU 52 to the node can be accurately measured and identified by the ECU 52. Each actuator 54 can then be uniquely identified by a unique distance, which is well above the distance that causes the distance detection error of the electronic system built into the ECU 52. Experiments using state-of-the-art technology have shown that loop 58 should be ensured to separate two adjacent actuators coupled to the communication bus by wiring at least about 15 centimeters. This separation allows each actuator to be positioned to have a unique distance from the controller, the position of each actuator followed by the position of all other actuators along the communication bus 53. Be identified. Other embodiments and other systems can also incorporate wires of lengths that make each node identifiable. That is, it provides a minimum interactuator distance that exceeds the distance at which controller measurement and operational errors can occur.

【0026】

Continuing counterclockwise along the communication bus 53, two actuators 54 separated by a wire loop 58 are located in the upper center of FIG. 4 and one or more sides for the passenger seat of the car. -Control the airbag. Continuing clockwise along the ring-shaped communication bus 53, two actuators 54 separated by another wire loop 58 are used to control the side airbag for the right rear seat. .. Continuing clockwise along the communication bus 53 towards the bottom of FIG. 4, the communication bus 53 couples with two actuators 54 separated by a loop 58, which are for the left rear seat. Control one or more airbags. Still traveling clockwise, two or more actuators 54 and a wire loop 58 between the actuators that are closely spaced allow control of the driver's side airbags. Continuing clockwise towards the airbag ECU 52, the last two actuators 54 and another wire loop 58 are used to operatively connect the last two actuators in the driver's seat. Allows control of the front airbag. As can be seen from FIG. 4, the actuator 54, the wire loop 58 and the ring-shaped communication bus 53 realize the ring-shaped communication bus architecture as shown in FIG.

【0027】

Figure 4 also shows communication bus structures 55,57, each coupled with airbag ECU 52. The communication bus 55 controls the seatbelt pretensioner actuators 54 on both the front and rear seats 56 and the front airbag for the left rear seat. Similarly, the tree-shaped communication bus structure 57 controls the two seatbelt pretensioner actuators 54 for the front seat 56 and the rear seat 56, and the second front airbag for the right rear seat. These tree structures 55,57 are similar to the general tree structure shown in FIG.

【0028】

Returning to the communication bus 53, the airbag ECU 52 uniquely determines the position of each actuator 54 on the communication bus 53. By selectively positioning the loops between the actuators 54, the distance between the actuators is increased so that they cannot be distinguished from each other. The ECU 52 has information about the predetermined topology of the system. The ECU 52 then assigns an address to each actuator 54 based on the distance from the airbag ECU 52. Note that according to one embodiment, this distance is measured in a single direction. Given a unique address for each actuator 54, each actuator 54 can be individually addressed and controlled by the airbag ECU 52 for diagnostic testing and monitoring. The same is true for all actuators 54 along the tree-shaped communication buses 55,57. If each actuator is uniquely placed along the distance of these communication buses 55,57, the actuator 54 on these communication buses can be given a unique address for subsequent operation, monitoring and Can be uniquely identified for diagnostic repair.

【0029】

The application to automotive airbags shown in FIG. 4 is an example of a distributed command and control system incorporating the present invention. Both ring-shaped and tree-shaped systems are considered. In alternative embodiments, these types can be combined or other types of command and control distribution schemes can be implemented.

【0030】

FIG. 5 shows the tree-like communication bus structure 55 or 57 in more detail, including the node 54 and the ECU 52 of the airbag electrical system 50 of FIG. More specifically, the airbag ECU 52 of FIG. 5 includes an ECU communication device 59 and a current / voltage source device 58, each of which is coupled to a communication bus 55. When the system of FIGS. 4 and 5 is assembled for the first time, the generic node 54 in the system is not uniquely tagged. At startup, the current / voltage source device 58 is used to identify the distance of each node 54 along the communication bus interconnect 55 or 57. Then, the ECU communication device 59 gives a unique address to each node 54 on the communication bus 55. After giving these unique addresses to the nodes 54, the normal operating mode is freely started, in which each node is accessed by the address. When first installed in the transportation system, in normal operating mode, control information, even if each node 54 connected to the communication bus 55 has a general purpose structure that is indistinguishable to all other nodes 54. , The monitor information and the diagnostic information are uniquely tagged and transmitted to the individual nodes 54 connected to the communication bus 55.

【0031】

According to one automotive airbag embodiment, each node 54 in FIGS. 4 and 5 includes a communication circuit and an error checking circuit 60 (such as a Cyclic Redundancy Check (CRC) checker). Each node also includes a node address initializer 61 and a squib controller 66. As described above, the communication circuit device 60 enables bidirectional communication of information between the node 54 and the ECU communication device 59. When the system of FIG. 5 is in the initialization operating mode, the communication circuit device 60 can provide the derived digital serial distance data to the ECU communication device 59. Further, the communication circuit device 60 receives a unique address transmitted to the node 54 from the ECU communication device 59 in response to the communication of the distance data, and appropriately sets the route. Also, during normal operating modes, data, addresses and CRC data are sequentially latched as inputs by interface 60 and latched for sequential output. As a minimum, the communication circuit 60 typically has a serial shift register for collecting address bits, a serial shift register for collecting properly addressed data, and serial data output to the communication bus. Has a data register (which may be the same as a data collection register) to provide. Transmission between the airbag ECU 52 and node 54 during normal operating modes, using error checking circuits as shown with communication circuit 60, or CRC checking or other types of error detection and / or correction circuits. Ensure that errors do not cause the airbag safety system to malfunction (eg, activate the airbag when no accident has occurred).

【0032】

The analog distance determination circuit 61 includes a controller 62, a random access memory (RAM) 68, a non-volatile memory portion 70, an A / D converter 64, and a voltage regulator 72. Controller 62 typically controls the operation of node 54 (eg, putting the node into normal or initialized operation mode). When the node 54 is first booted, the controller 62 informs the A / D converter 64 that the initialization operation mode will be started. During the initialization operating mode, the current and voltage source device 58 sends voltage and / current signals along the communication bus 55, which causes the A / D converter 64 to have a constant digital value over time. Can be derived. Note that during the initialization process, the connector 74 is temporarily coupled to the communication bus 55, causing the communication bus to be short-circuited or "closed loop" to determine the current drive distance of the initialization process (Figure). 7). These specific digital values are sent to the RAM 68 via the controller 62. The specific calculation and provision of these digital values and their importance in determining the distance on the communication bus 55 will be described in detail with reference to FIG. 7 below.

【0033】

Once the individual digital distance values are stored in RAM 68, the controller 62 then returns these distance values to the ECU communication device 59 via the communication circuit 60. When the ECU device 52 receives all the distance information from all of the devices 54, the ECU 52 passes the address from the device 56 to all the controllers 62 in each node 54 of the system of FIG. During the operation in which a unique address is passed, each node 54 is uniquely tagged by the ECU communication device 59 using the distance value previously stored by controller 62 as a unique identifier in RAM 68. That is, each node 54 in the system of FIG. 5 has different distance values stored in RAM 68, and these different distance values address one node 54 during the address initialization process. And can be used to eliminate all other nodes 54. If the node 54 is uniquely identified by the distance value retransmitted from the device 56 to the node 54, the ECU communication device 59 gives a unique address value, and only the communication device 60 of the selected node 54 has this address. Receive a value. The controller 62 sends each unique address to the non-volatile memory portion 70 of FIG. In this way, each node 54 is sequentially given a unique address.

【0034】

The non-volatile memory 70 is one or more ferroelectric embedded DRAMs, electrically erasable and writable read-only memories (EEPROMs), or similar non-volatile storage devices. If an address is obtained for use inside the non-volatile memory 70, controller 62 informs the A / D converter 64 to deactivate, and the entire system in Figure 5 is now fully addressed. It has a function and can enter a normal operation mode. In normal operating mode, the ECU communication device can send an information packet containing address, data and error inspection information to the communication bus. Only one node 54, which is tagged by the packet with a unique tag that can be determined by comparing the address of the packet with the contents of the non-volatile memory, processes the data information. Therefore, all generic nodes 54 in the system of FIG. 5 are uniquely tagged by using the architecture shown in FIG.

【0035】

FIG. 5 shows the tree structure of the communication bus 55 and / or 57 of FIG. 4, while FIG. 6 shows the ring-shaped structure of the communication bus of FIG. In general, the architecture in Figure 6 is similar to the architecture in Figure 5, with some minor changes. FIG. 5 relies on an externally connected connector 74 to perform a current-based distance detection operation in order to short-circuit the communication bus 62 of FIG. In FIG. 6, due to the ring-shaped architecture, the switch device 76 is used in the airbag ECU 52 to close loop the communication bus 53. Closed-loop operation is required inside the system during initialization, which will become clear when you consider Figures 7-8.

【0036】

7-8 show interconnected flowcharts that teach how to initialize the structures of 5 and / or 6. In general, FIGS. 7-8 show techniques that use voltage and current processing to make it possible to identify the physically unique distance of each node 54 along the communication bus structure 53 or 55. If this distance is determined by the node itself for each node, the distance value is transmitted from the node 54 to the airbag ECU 52. This distance information is processed by the ECU 52, and each unique distance is assigned to a unique address. These unique addresses are then given to each node 54, and the nodes are addressed to give a unique address by the unique distance information previously transmitted by the node 54. When each node 54 receives a unique address, the unique address is stored in non-volatile memory (NVM) and is permanently used for addressing within the node. At this point, all other generic nodes 54 are given a unique address, which allows normal operation to be initiated by each node 54 and uniquely addressed within the communication bus system.

【0037】

The method of FIGS. 7-8 begins by defining the topology of the airbag system at step 82 of FIG. That is, the automobile designer can determine the number of actuators coupled to each communication bus, the type of communication bus architecture used (ring or tree, or both), and the position of the actuators along the distance of the communication bus. / Actuator function on each communication bus according to distance (eg the closest actuator is the front airbag, the central actuator is the front seat pretensioner, the farthest actuator on the communication bus is the car Is a collision accelerator that detects accidents). Assembling the transportation system, It is important to define the topology before initialization and operation. This is because certain parameters of the software in the ECU 52 need to be programmed according to the topology for proper initialization and operation. For example, for initialization purposes, the ECU 52 may find it useful to know exactly how many nodes 54 are connected to the communication bus. Once the specific number of nodes connected to the communication bus is known, the ECU 52 can perform more efficient initialization operations without skipping the address initialization of some important nodes. Also, if the ECU52 determines that there are N nodes whose distances are ranked from 1 to N, it is useful for the ECU52 to know what function each actuator will perform in its relative position. Will. Which position actuator represents the passenger airbag squib, and which position actuator represents the driver airbag squib, what is the closest and farthest actuator if the system does not know Even if you know, it's almost useless. In any case, the changes to the software in the ECU 52 required to achieve different topologies should be minimized, usually with only one or two to three variable changes in the software of the ECU 52.

【0038】

After the airbag system topology is defined in stage 82, the airbag system is assembled in stage 84. That is, step 84 assembles the actuator 54, the main node ECU 52 and the wiring into the electrical system shown in one or more of FIGS. 2-4. At stage 84, once the system is assembled, ECU 52 is programmed at stage 86 according to the topology of stage 82. At stage 86, the ECU is programmed according to the topology of stage 82, enabling efficient initialization and operation of the system, as described above. For example, the ECU52 has six actuators in the communication bus system, the closest two actuators are passenger airbags, and the next two actuators are side air. The bag is informed that the two farthest actuators are seatbelt restraint devices for the rear seats. Knowing the number of actuators in the system and the desired behavior as a function of the distance along the communication bus allows the ECU to perform efficient and correct initialization behavior on these general purpose actuators. In addition to programming the ECU 52, all actuators, communication buses and ECU devices in the system are powered up at stage 86.

【0039】

In step 88, the current and voltage source 58 of FIG. 5 or 6 provides a reference voltage (Vref) along the respective communication bus 53 or 55 that requires initialization. At this point, the connector 74 of FIG. 5 is not connected to the system and the closed circuit device 76 of FIG. 6 is controlled so that the end of the ring-shaped communication bus 53 of FIG. 6 is disconnected. .. In this state, the reference voltage supplied in the communication bus is supplied to all nodes 54 in the system, assuming that there is no significant current loss due to leakage current (safe and accurate assumption for the scheme in Figure 7). It is desirable that it is the same as the voltage to be applied. With little or no leakage current, the voltages V1 and voltage VN at all N nodes 54 coupled in the system of FIGS. 5 and 6 each register the same input voltage supplied by device 58. Should be, V1 = V2 = ... = VN = Vref. Since node 54 has just been launched, controller 62 in FIGS. 5 and 6 informs the A / D converter 64 that initialization mode is currently in progress. The analog-to-digital converter converts the reference voltage (V1..VN) of each node into a digital voltage reference (Vref) value, and stores this converted result in the RAM 68 via the controller 62. To do. Therefore, each of the N nodes of FIGS. 5 and 6 can store the digital value corresponding to each input reference voltage value in the RAM 68.

【0040】

The digital Vref value stored in RAM 68 in each node 54 plays a vital role. Each A / D converter 64 and each voltage regulator 72 within each node is an incomplete and inconsistent analog device, even if the analog Vref voltage applied to each node 54 is equal or nearly equal. .. The A / D converter 64 and voltage regulator 72 vary from node to node, vary by different temperatures, and have inherent non-linearity, making each node 54 slightly different from all other nodes 54. there is a possibility. To eliminate such non-linearity and inherent errors from the system, the method in Figure 7 uses a reference voltage (Vref). The digital Vref value stored in the RAM device 68 allows the ECU communication device 59 to later eliminate the non-linearity from the system, which can adversely affect the non-linearity (positional determination) at one node. (There is a possibility) can be eliminated as a problem via the ECU52, which makes the distance judgment of each node in the system much more accurate. Therefore, step 90 in FIG. 7 stores the digital Vref value in the RAM position 68 of each node for the purpose of correcting the distance error.

【0041】

After step 90, step 92 signals the end of the Vref recording step by cutting off the supply of current and voltage reference signals from the voltage source 58. This Vref end period is implemented to notify all nodes 54 that the digital Vref recording portion of the initialization algorithm has ended. Typically, stage 90 provides sufficient time for all nodes 54 to be properly processed and to ensure that the digital Vref value is stored before the Vref is blocked at stage 92.

【0042】

In step 94, the airbag communication bus 53 and / or 55 is configured to perform a distance loop measurement using the sourced current supplied as an output from the current source 8. In the case of FIG. 5, in order to enable the required closed loop configuration, the communication bus 62 is provided with a connector 74 and both signals of the communication bus 62 are short-circuited together to form a loop structure. The connector 74 can be simplified as a resistor or wire loop, but the 74 can also include advanced diagnostic electronics. When two or more communication bus signals are generated, the loop closes any two signals (eg power and ground, signal and ground, signal and power, etc.) that are connected to any node on the communication bus. (close You only need to circuit). For the ring configuration of FIG. 6, to achieve step 94 of FIG. 7, in the device 76, both signals of the ring communication bus 53 are shorted together together via an optional resistor device. Inform the circuit. Therefore, in step 94, either the externally connected connector 74 or the internally controlled device 76 forms a closed loop connection on their respective communication bus structures.

【0043】

In step 94, if a closed-loop connection is formed on the communication bus, the current and voltage source device 58 is connected to a fixed reference current (I) via the closed-loop communication bus 53 or 55 in one or more of FIGS. 5 and 6. ) Is transmitted in one direction. Since the communication bus has a resistance value per unit length, the resistance received by each node via the communication bus connection is linearly proportional to the distance of the wiring connected between the device 58 and the node 54. That is, the node 54 closer to the current and voltage source device 58 along the communication bus 53 or 55 receives a different voltage than the node located far away from the current and voltage source device 58. As a specific example, the voltage V1 in FIG. 5 is the differential node voltage that is physically closest to the current and voltage source device 58 and may have a differential voltage value of 100 millivolts (mV). Therefore, the node 54 located farthest from the current and voltage source device 58 may have a differential VN of about 15 millivolts (mV). This different voltage received by each node is called the distance voltage or Vdist of each node. Any node located between the nearest node 54 and the farthest node 54 has a different Vdist voltage value in proportion to the distance of the communication bus, where Vdist is the high voltage value of 100 mV in the above example. , Between low voltage values of 15 mV.

【0044】

In step 98, the controller 62 waits for the settling time in response to the cues from steps 92 to 94 and then for the A / D converter 64 in each node 54 of this analog distance voltage (Vdist). Notify that the converted value be recorded and that the resulting digital distance voltage (Vdist) be stored in RAM 68 along with the previously stored digital Vref value. Therefore, the node 54 closest to the device 58 stores a digital Vdist value approximately equal to 100 millivolts in RAM 68, whereas the farthest node 54 in the system stores a digital Vdist value approximately equal to 15 millivolts. Record in RAM68. Note that two adjacent nodes that are closely coupled along the communication bus may differ in Vdist by only 2-3 millivolts. In such cases, non-linearity between analog circuits within node 54, process differences and similar inherent variations can result in erroneous measurements of 50 mV for the nearest node and 51 mV for the farthest node. (Note that in the concrete examples of FIGS. 5-6, the closest node has more Vdist than the farthest node). However, such errors in Vdist can be quantified by the Vref value, and the Vref value can be used to "correct" the error Vdist value to a "true" or "corrected Vdist value". By combining the Vref value and Vdist value in each RAM68, each node 54 can be uniquely identified by the ECU communication device 59, because no two nodes have the same Vdist value and Vref value. is there. That is, the combination of the Vref value and the Vdist value in the node 54 allows the ECU device 56 to accurately organize the nodes 54 according to their respective distances along the communication bus, and the ECU device 56 can organize each node in the system. Can be uniquely identified.

【0045】

In one form, the Vref and Vdist values are processed by individual nodes to give rise to uniquely processed values, which are the unique addresses of each node. If the node assigns a unique address, the ECU will be informed of what the address is. In another form, the Vdist and Vref values are propagated to the more powerful ECU52, where the ECU mathematically processes the Vdist and Vref values and re-uses the Vdist and Vref values as unique node identifiers. By communicating, each node can be given a unique address. This second form will be described in detail in the remaining stages of Figure 7.

【0046】

Stage 100 begins after a period of time sufficient for each node to perform the appropriate processing in stage 98. At stage 100, connector 74 in FIG. 5 is disconnected from communication bus 55 and / or device 76 in FIG. 6 becomes inoperable, bringing communication buses 53 and / or 55 into a normal open-loop configuration. return. Also, the current and voltage source 58 becomes inoperable in step 100 and is disconnected from the communication bus 53 and / or 55. In essence, the airbag communication bus is transformed back to the normal configuration, where each node 54 transfers the Vdist and Vref values from each RAM 68 via the communication circuit 60 within each node to the ECU communication device. To be able to communicate to 59. Note that the communication circuit 60 can include a serial communication device or other type of device that supports communication over a common bus.

【0047】

The transmission of the Vdist value and Vref value from the node 54 to the ECU device 52 is initiated by step 102 in FIG. In step 102, each node sets a random time to start transmitting the Vdist and Vref values from RAM 68 to the ECU communication device 59. To enable such communication, each node 54 waits for a random time, which is likely to be different for each node, and then the Vdist and Vref values from RAM 68 via controller 62. Provided to the communication circuit device 60. That is, each node sets a random time on the counter and uses this random time to time out the counter (each random time is statistically different for most of the time on all nodes). It should be noted that if the two times are too close, equal, or overlap, the techniques outlined below can be used to compensate for the collision). If the random time times out at the particular node 54, the communication circuit 60 of the particular node begins to continuously provide Vref and Vdist information to the ECU communication device 59. The ECU communication device uses the CRC inspection total to perform processing based on all the information provided from each node 54, and ensures that the Vref and Vdist information of each node 54 is properly received. The node 54 can determine for itself whether or not the information has been properly transmitted by monitoring the presence or absence of the contention of the communication bus for the communication bus 53 and / or 55 via the communication circuit device 60. If the communication circuit 60 of node 54 on the transmission side does not detect the contention of the communication bus and the ECU communication device 59 performs an appropriate CRC inspection total calculation, the information is communicated from node 54 in step 102. Properly transmitted to device 59 and does not require retransmission for initialization.

【0048】

However, in transmissions between two or more nodes, some conflicts may occur statistically from time to time. Therefore, the ECU communication device 59 and the communication circuit device 60 constantly check the communication bus 53 and / or 55 for the presence or absence of the content of the communication bus during the step 106. If one or more nodes 54 detect a collision while transmitting a Vref or Vdist value, each node experiencing the collision will generate a new random delay time (see step 104), and then Attempts to retransmit Vdist and Vref information after an additional random time times out. It is unclear and random how many collisions occur at each initialization, but according to traditional statistical analysis, all nodes 54 cause collisions or errors in the ECU communication device 59. It has been revealed that Vdist and Vref information can be transmitted randomly in a reasonable time well less than 1 second without any problem.

【0049】

The ECU communication device 59 understands the topology of the system defined in stage 82. Therefore, the ECU communication device 59 knows how many nodes 54 exist on the communication bus 53 and / or 55. Therefore, the ECU communication device waits for an appropriate random time until all combinations of Vref values and Vdist values of all N nodes are properly received. If the ECU52 receives less than N correct transmissions to the system, the noise may have caused the node to go unnoticed by the communication bus contention or error, and all N nodes have Retransmission of all nodes must be started again until they are received in random order without CRC error.

【0050】

After such a random interval (a function of communication bus contention and noise error), the ECU communicator 52 will receive and process all digital data pairs of Vref and Vdist. Secured. Therefore, steps 108 and 118 are carried out by the ECU communication device 59 until all N pairs of valid Vdist and Vref information have been completely received by the device 56. When all N pairs of Vref and Vdist are received for all N nodes 54 in the system, step 112 in Figure 7 is performed. In step 112, mathematical calculations are performed by device 56 using the Vdist and Vref values for voltage fluctuations, performance fluctuations, processing fluctuations, A / D non-linearity and other errors and / or inherent. Eliminates the difference between the error / Vdist original data and yields the corrected Vdist data.

【0051】

After the mathematical calculations for such corrections have been performed, step 114 of FIG. 8 is performed. In step 114, the ECU communication device 59 sorts all the corrected Vdist values in descending or descending order. Since the ECU communication device 59 is informed of the topology of the system at stage 82, the airbag system determines which device is the closest device in the system, the function it performs, and which device It identifies whether it is the next closest device in the system, the function it performs, and so on, and finally the farthest device in the system is identified by position and function. Summarizing the above, in step 114, after the alignment is performed, the ECU 52 can determine which node 54 should be given which address (see step 116), and the device 52 is along the communication bus and / or 55. In view of the distance, it is possible to determine the function performed by each of the addressed nodes 54 in the system.

【0052】

Thus, if the corrected Vdist values are sorted by magnitude, step 116 in FIG. 8 is performed to assign a unique address to each node 54 according to function and / or distance. In step 118 of FIG. 8, the ECU communication device continuously transmits a pair of Vref value and Vdist value along the communication bus 53 or 55. These Vref values and the uncorrected or original Vdist values are collected by all communication circuits at node 54. Only one node 54 of all nodes 54 receives a Vref value and an uncorrected Vdist value that match the information previously stored in its RAM 68. This one node 54 is then informed by the controller 62 that it must then receive a unique address to be transmitted and eliminate all other nodes 54. The uncorrected Vdist and Vref information is followed by a unique address, and only one node 54, identified by the Vdist and Vref values, receives the unique address through the communication circuit 60 (see step 122). .. In step 122, a unique address is written only to the non-volatile memory of the selected node 54.

【0053】

Steps 118-124 are repeated N times until all N unique addresses are properly sent to all N nodes connected to the communication bus and a CRC check is performed. A node that experiences a failure during address CRC operation can later retransmit the Vref and Vdist information to the ECU 52, flagging such errors and sending them to the ECU 52. When the one node 54 is flagged for a CRC error, the ECU can transmit the unique address again (necessarily) until all the unique addresses are properly stored in the NVM70 by the node 54. Depending on the Vref and the original Vdist). After all devices 54 have been given unique addresses and have undergone appropriate CRC inspection, each unique address is stored in the non-volatile memory 70 of each node 54. After the unique address is stored in the NVM, the normal mode of operation begins, which allows the transport system to use the ECU device 56 for operation, later reinitialization, diagnostic monitoring, etc. Any of the generic nodes 54 in the can be uniquely identified.

【0054】

FIG. 9 shows an alternative embodiment of the communication bus structure 10 shown in FIG. All elements in FIGS. 5 and 9 that are generic or substantially similar in structure and / or function are indicated by generic reference numbers. One major difference between FIG. 9 and FIG. 5 is that the current and voltage source 58 in FIG. 5 has been replaced by the current lamp supply 134 in FIG. Also, the externally connected connector 74 in FIG. 5 is replaced by an externally connected, more complex reference voltage source (Vref) 136. In addition to these differences, the analog distance determination circuit 61 of FIG. 5 is different from the node address initializer 135 of FIG. In FIG. 5, the device 61 uses the A / D converter 64 during initialization to uniquely determine the physical position or distance of the node 54 along the length of the communication bus 62. In FIG. 9, the peak and hold device 132 replaces the voltage regulator 72 in the A / D converter 64 of FIG. According to the alternative embodiment, a sample hold device can be realized.

【0055】

The structural differences between Figures 5 and 9 appear to be subtle, but the differences have a significant impact on the method. Unlike FIG. 5, in FIG. 9, the reference voltage source 136 provides a reference voltage (Vref) along the communication bus 62, and the current lamp 134 does not provide current. At this point, each node 154 stores a trigger voltage in a capacitor in the peak hold device 132, which voltage is proportional to the reference voltage provided by the device 136. Therefore, each node 54 stores substantially the same trigger voltage inside each device 132.

【0056】

At this point, the current ramp circuit in circuit 134 begins to ramp the current from zero amperes to a predetermined amperage level. Since the resistance length of the communication bus exposed to each node 54 is different, the differential voltage values V1 to VN of each node 54 change at different rates as the current ramps linearly. That is, a node with a larger resistance length of the communication bus has a sharper change in voltage than a node with a lower resistance of the communication bus, and as the current lamp advances, the voltage V1 to VN of the node opens. Starts to grow / spread. In one embodiment, the voltage V1 reaches the trigger voltage and then the other node voltages V2 to VN reach the trigger voltage. At this point, the peak hold device 132 notifies the first V1 node 54 that the trigger voltage has been reached and informs the controller 130 of this event. The controller 130 of the 1st V1 node 54 then sends a signal to the ECU. One way to send this signal is to short-circuit the communication bus 62 via the communication circuit device 60, which causes the ECU communication device 59 to detect an electrical short-circuit circuit change on the communication bus. ..

【0057】

At this point, the current lamp operation via the device 134 is stopped, and the ECU communication device 59 transmits the unique address value to the selected V1 node 54 and stores it in the location 70 of the non-volatile memory. After receiving a valid address verified by CRC operation, controller 130 of the selected node 54 terminates the connection between the peak hold device and the communication bus, and ceases involvement and further current ramp operation. The next current ramp operation then proceeds to the most recently programmed node, and all previously programmed addresses are not involved in the ramp operation. In this new current (I) lamp, all previously addressed nodes are eliminated, so the next ramp operation will detect the next farthest node in the system and the voltage Vn (n) for this node will be. An integer between 1 and N) reaches the trigger voltage in the peak hold circuit 132. In this way, sequential triggering of one node at a time continues until all nodes in the system are sequentially programmed with addresses based on their position / distance along the communication bus.

【0058】

FIG. 10 shows a system similar to the operation of FIG. 9, but FIG. 10 uses the ring configuration of FIG. 6 instead of the tree configuration of FIG. In the ring architecture of FIG. 10, to implement the current ramp process of FIG. 9, the voltage reference source 136 is moved to a position on the chip with the current ramp circuit 134. The communication bus 60 then forms a ring between the current lamp circuit 134 and the voltage reference source device 136. As in the case of FIGS. 5 and 6, the systems of FIGS. 9 and 10 are very similar in structure and method. Therefore, the description with respect to FIG. 9 generally applies to the description with reference to FIG. 10, and the device 136 is internally controlled by the ECU 52 rather than being controlled externally as in FIG. 9.

【0059】

FIG. 11 shows a method of operation that can be used according to the systems of FIGS. 9 and 10 when used in the transport equipment application shown in FIG. In method 200 of FIG. 11, steps 202-206 are similar and substantially similar to steps 82-86 of FIGS. 7-8. In FIG. 11, step 208 enters initialization mode and supplies a reference voltage (Vref) over the communication bus of the airbag system via device 136. As mentioned above, no current flows along the communication bus during step 208, so that all nodes 54 should receive the same reference voltage, assuming very little leakage current. .. In stage 210, all nodes in the system shown in FIGS. 9 and 10 store the trigger voltage, which is a function of the reference voltage (Vref) across the terminals. This trigger voltage is stored in a capacitor or similar element / system in circuit 132.

【0060】

If the reference voltage is stored in a capacitor inside the peak hold device 132 of each node 54, step 212 signals the end of the Vref storage phase to node 54, for example by shutting off the Vref. On the other hand, it signals that the next stage of initialization will begin. In step 214, the Vref voltage from device 136 is raised again and the current output from device 134 is ramped from no current to maximum initialization current in step 216 until a trigger is detected. The presence of a fixed voltage 136 and changes in the current 134 ensure that the differential voltage of every node 54 changes at different rates, which means that these nodes differ in their resistive communication bus interface. This is because it is placed in a position.

【0061】

At stage 218, each node constantly compares the voltage across the differential input with the previously stored trigger voltage, the latter being a function of the Vref voltage at stage 218. Since each node 54 receives a different resistance from the communication bus, the current ramp causes the differential input voltage of one node from V1 to VN to reach the trigger point before everything else. Therefore, due to the current ramp, one selected node 54 reaches the trigger voltage stored in the peak hold device 132 before all other nodes 54 reach this value. This selected node can be the node closest to or farthest from ECU 52, depending on the placement of devices 134,136 in the system, and corresponds to the selected node in stage 220 of FIG.

【0062】

Once the selected node has been determined, the controller 130 of the selected node 54 shorts the communication bus via their respective interfaces 60. The ECU 52 detects this short circuit in the communication bus and at step 222 reconfigures the communication bus of the airbag system to its normal configuration. After shorting the communication bus, the controller of the selected node configures the communication circuit to receive a unique address, but all other untriggered nodes 54 ignore the communication of the address. Step 224 of FIG. 11 then provides a unique address from the ECU communication device 59 to one triggered node 54 in the system. This unique address is received by only one trigger / selected node 54, and this unique address is stored in each non-volatile memory 70 of the selected node.

【0063】

At step 226, the selected node is removed from the initialization state, and the peak hold circuit 132 of the selected node 54 is disconnected from the communication bus and is no longer involved in the current lamp operation. The node with the programmed address ignores the operation of the communication bus (ie, the ramp initialization operation) until the ECU 52 sends a "finished initialization" data packet sequence to all nodes 54 on the communication bus. And all other behavior). If the transmission of the address to the selected node 54 fails due to a CRC failure detected in the device 60 of the selected node, the selected node 54 causes the ECU 52 to take such an event by short-circuiting the communication bus. Please note that you can let us know. If the ECU52 notifies that the communication bus has been short-circuited too early during address transmission or during the initial phase of starting the next ramp process, the ECU52 is the last selected node. Know that the address must be propagated to 54 again. Such handshaking is repeated until the address is correctly programmed into the selected node 54, or until the initialization operation times out and the system sends an error message due to the inability to program node 54. It is said. In another embodiment, the acceptance notification can be sent from the device 60 if the acceptance is complete and appropriate, and if the device 60 does not activate the acceptance notification, the information is sent back into the system. it can.

【0064】

At step 228, the ECU communication device 59 knows the structure of the system from step 206 and determines if any other node in the system remains unaddressed. If the node is left unaddressed, steps 208-228 are repeated until every node 54 in the system is sequentially triggered and the ECU 52 sequentially assigns a unique address. If the last node 54 in the system is given a unique address and all CRC values do not indicate an error, then step 230 in Figure 10 puts it into normal operating mode, which causes each node 54 to enter normal operating mode. , The start data packet sent from the ECU 52 along the communication bus changes from the "ignore initialization" state to the "awakened" state. Now all nodes 54 in the system of FIGS. 9-10 will respond to unique addresses in the system.

【0065】

Although the present invention has been described and illustrated with reference to specific examples, it is not intended to limit the invention to these illustrated examples. Those skilled in the art will recognize that modifications and variations are possible without departing from the intent and scope of the invention. Therefore, the present invention is intended to include all variations and variations that fall within the claims.

[Simple explanation of drawings]

[Figure 1]

FIG. 6 is a block diagram of an electrical system of a transport device, including many electrical subsystems constructed by technically known methods.

[Figure 2]

According to the present invention, it is a block diagram of a tree-shaped communication bus system that can be used to interconnect a plurality of general-purpose nodes in a transportation system while individually performing unique addressing for each node.

[Fig. 3]

According to the present invention, it is a block diagram of a ring-shaped communication bus system that can be used to interconnect a plurality of general-purpose nodes in a transportation system while individually performing unique address designation for each node.

[Fig. 4]

FIG. 5 is a top-down schematic of an automotive airbag system 50 that is spatially oriented according to an automotive design according to the present invention.

[Fig. 5]

It is a block diagram of the airbag automobile subsystem of the specific tree-like structure by this invention.

[Fig. 6]

It is a block diagram of the airbag automobile system of the specific ring-shaped structure by this invention.

[Fig. 7]

According to the present invention, a method of configuring the nodes of FIGS. 4 to 6 with unique addresses is shown in an interconnection flowchart.

[Fig. 8]

According to the present invention, a method of configuring the nodes of FIGS. 4 to 6 with unique addresses is shown in an interconnection flowchart.

[Fig. 9]

According to the present invention, alternative embodiments of the systems already shown in FIGS. 5 and 6 are shown in block diagrams.

[Fig. 10]

According to the present invention, alternative embodiments of the systems already shown in FIGS. 5 and 6 are shown in block diagrams.

[Fig. 11]

According to the present invention, an alternative method of the method already shown in FIGS. 7 to 8 is shown in a flowchart.

[Explanation of symbols]

31,40 Main node (not the same as other nodes) 41 Communication bus 1 node 2 Communication bus 50 Automotive airbag system 52 airbag ECU 53 Ring-shaped communication bus 54 nodes 55 Tree structure 56 Front seats, back seats 57 Tree structure 58 Wire Loop (Figure 4), Current and Voltage Sources (Figure 5) 59 ECU communication device 60 Error inspection circuit / communication circuit 61 Node address initializer 62 controller 64 A / D converter 66 squib control 68 RAM 70 Non-volatile memory 72 Voltage regulator 74 connector 76 Closing device