A method and system for generating weather and ground reflectivity information

[0001] Attempts have been made to obtain a radar signal received from an aircraft radar system, convert it to a reflection value, and store it in a three-dimensional (3D) weather buffer location for the range associated with the radar signal. It has been executed. A three-dimensional weather buffer is an array of computer memory that contains data that describes the distribution of weather reflections in a three-dimensional space. In other attempts, radar return signal power is converted into reflections based on certain assumptions and used to place elements in the 3D weather buffer.

[0002] The challenge with previous attempts is the rationale for the confusion of desired meteorological signals. Prior attempts to prevent basic disruption contamination used antenna beam calculations proximal to the assumed location on the ground, and the ground signal scattering property was then used to determine the degree of signal contamination, which was then it. Used to suppress the signal if it is considered contaminated.

[0003] Therefore, there is a need to identify the weather information more accurately, removing ground disruption pollution from the weather radar signal, or identifying the ground in the weather radar signal.

[0004] US Pat. No. 6,707,415, which is incorporated herein by reference, estimates weather and ground reflections for the purpose of displaying relatively unpolluted weather reflections by ground signal returns. Or describe how to use radar to estimate ground reflections that are minimally contaminated by the weather. The method utilized parameters representing relative uncertainty in the current assessment of weather reflections, and the ground normalized the radar cross section for all modeling locations. These parameters can or cannot be updated in response to new measurements (which reduce the uncertainty of estimates). However, in any case, the starting values for the weather and ground reflection uncertainty parameters must be selected.

[0005] Separate assessments of weather and ground reflections are not complete for any assessment process. From time to time, some meteorological signals have the same result with increasing estimated ground reflections and vice versa.

<p num="0006"> [0006] The present invention is a method, system and computer program product for storing weather radar cross section data in a three-dimensional buffer. The method involves modeling the radar signal scattering properties of the space surrounding the radar / aircraft. Current radar measurements are compared to the predicted measurements using a model of the measurement process. The difference between the current radar reading and the prediction of the reading is used to adjust the stored reflections.</p><p num="0007"> [0007] The distribution of normalized radar cross sections (NRCS) on the surface of the ground is estimated in a manner similar to weather reflection assessment. With respect to the ground, the buffer represents the NRCS distribution in two dimensions to describe the surface of the ground, rather than the three dimensions that describe the three-dimensional weather. Since one of the onboard radar applications is to provide a radar ground map, the present invention provides a ground map as a unique part of the process.</p><p num="0008"> [0008] As immediately acknowledged from the above overview, the present invention is more for storage in a three-dimensional buffer by using radar signal power measurements to estimate weather reflection distribution and ground NRCS distribution separately. Provide accurate information. The process evaluates the distribution using the difference between the expected and radar measurements generated by the model of the measurement process, which behaves according to previously stored estimates of the weather reflection and ground NRCS distribution. Is executed by iteratively updating.</p><p num="0009"> [0009] Other preferred embodiments of the present invention are described in detail below with respect to the drawings below.</p>

<figref num="1">[0010] FIG. 1 is a block diagram of a system that implements the present invention;</figref><figref num="2">[0011] FIG. 2 illustrates a flow chart diagram performed by the system shown in FIG.</figref>

[0012] The present invention is a computer program product for storing systems, methods and 3D radar return data. FIG. 1 illustrates an exemplary system 30 formed by the present invention. System 30 includes a weather radar system 40, a processor 42, a memory 43, a display 44, an inertial navigation system (INS) 46, and a user interface 48 coupled to the processor 42. The display processor 42 is electrically coupled to the radar system 40, the display device 44, the INS 46 and the memory 43. The illustrated radar system 40 includes a radar controller 50, a transmitter 52, a receiver 54 and an antenna 56. To perform radar controller 50 transmitting and receiving signals through antenna 56 based on aircraft data (ie position, direction, roll, yaw, pitch, etc.) received from INS 46 or other aircraft systems. Controls transmitter 52 and receiver 54.

[0013] The radar system 40 receives signals due to pulse scattering transmitted mainly from the external environment of the weather and terrain. The received signal is passed through processor 42, which uses the received signal to update the ground-normalized radar cross section and weather reflection assessment contained in computer memory (three-dimensional 3D buffer). Processor 42 produces an image for display on display device 44 based on any control signal sent from user interface 48 or based on settings within the range of processor 42.

At system startup, all elements of barometric pressure (ie, 3D buffer) and all elements of ground (ie, 2D buffer) are initialized. The initialization of each element (ie, the buffer cell) is (1) reflection (with respect to the weather) or normalized radar cross section (NRCS) (with respect to the ground), and (2) crossing the weather and Includes selecting values for the uncertainty parameters associated with each initial value for cells for both ground. Uncertainty represents the processing of how wrong the initial reflection or NRCS value is.

[0015] For ground elements, the first NRCS and / or uncertainty parameters are the surface type (ie, water, forest, city, etc.), terrain shadowing, radar-to-ground element angle of incidence, or reflection. Deductive information about other factors that change can be used. With respect to weather reflections, the deductive information may be uplink weather reflections from other radars and corresponding uncertainty parameters that reflect the downward uncertainty for those elements.

[0016] In certain embodiments, the present invention provides that the NRCS corrects ground element uncertainty parameters based on expectations associated with surface type. For example, water is expected to have a low NRCS value, so if we initialize the water-type ground element to NRCS = 0, the initial NRCS will not accidentally become too much, so the corresponding uncertainty The sex is initialized to a low value. In other embodiments, any deductive information is used to initialize both the element value and its uncertainty parameter, but not for the ground element.

[0017] FIG. 2 illustrates an exemplary method 100 of storing reflection values in a three-dimensional display buffer. First, at block 102, the uncertainty parameters and reflection values associated with the 3D buffer location are initialized to the starting value. For example, a reflection value can be initialized to zero with an uncertainty parameter that is initialized to a relatively large value to indicate the first error that is as large as possible in the reflection value. If the reflection values are initialized with additional information such as uplinking weather reflections from a ground-based radar, the corresponding uncertainty parameter is the diminishing uncertainty of these first reflections. It will be initialized to a lower value to show certainty. Radar cross section uncertainty parameters normalized for ground elements can be initialized based on the type of ground associated with each element. At block 104, the main beam of the antenna is shown in a particular radial direction. At block 106, the radar system then transmits a waveform across the antenna and a plurality of range bins based on the transmitted waveform. Receive a signal sample for bins). At block 108, the antenna main beam is modeled by a two-dimensional array of vectors. Each vector represents an increase in solid angle within the range of the antenna main beam. Each vector has an associated antenna gain value.

[0018] Then, at block 110, the first range bin of the received signal is considered. As part of the consideration of the first range bin, in block 112, the processor 42 calculates the location in the range bin and the corresponding three-dimensional buffer, and the antenna showing the directional increase represented by the array of vectors Used to model the antenna main beam. Then, in block 104, the processor 42 searches for the data stored in the calculated position of the 3D buffer and uses the searched data to predict the radar signal power. At block 116, processor 42 subtracts the signal power prediction from the signal power measured in the range bin, thereby generating an innovation value. At block 118, processor 42 calculates the gain (k) for each calculated buffer location and innovates the calculated gain value to provide the latest version of the scattering parameters (weather reflection or ground NRCS) for each location. .. Next, at block 120, processor 42 updates the uncertainty parameters for each buffer location. In decision block 124, if the current range bin is not the last range bin of the current radiation, process 100 continues to the next range bin of the received signal. Once the next range bin value is retrieved, the process continues to block 112 until the last range bin value is reached. When the final range bin value is reached, the system obtains the next antenna pointing angle (see block 130) and the antenna is directed to the next searched antenna pointing angle, as determined in decision block 124. Return the process to block 104.

[0019] In the following, the method 100 according to the embodiment will be further described. The zero-time equation for determining the reflection value for storage in a three-dimensional storage location is:

<maths num="1"><img id="000002" he="13" wi="137" file="JP6130101B2_D0001.tif" img-format="tif" img-content="drawing" /></maths>

[0020] First, the residual signal is measured to obtain the residual signal or innovation S_measAnd forecast

<maths num="2"><img id="000003" he="10" wi="29" file="JP6130101B2_D0001.tif" img-format="tif" img-content="drawing" /></maths>

Determined by taking the difference between

<maths num="3"><img id="000004" he="7" wi="5" file="JP6130101B2_D0001.tif" img-format="tif" img-content="drawing" /></maths>

Is the reflection evaluation for the i-th grid position before measurement,

<maths num="4"><img id="000005" he="7" wi="5" file="JP6130101B2_D0001.tif" img-format="tif" img-content="drawing" /></maths>

Is the reflection evaluation after measurement, Ki is the filter gain for converting the radar signal into the reflection value, and S_measIs the measured radar signal. h is the element weight that depends on the radar beam shape and the location of the storage location associated with the radar beam shape. The determination of h is described in more detail below. The corresponding equation for updating the ground NRCS value is

<maths num="5"><img id="000006" he="14" wi="75" file="JP6130101B2_D0001.tif" img-format="tif" img-content="drawing" /></maths>

Is. [0021] The filter gain Ki is shown in Eq. (8).

<maths num="6"><img id="000007" he="15" wi="137" file="JP6130101B2_D0001.tif" img-format="tif" img-content="drawing" /></maths>

[0022] The total number of molecules includes both ground NRCS elements and weather reflection elements that contribute to the measurement. [0023] P_iIs the uncertainty parameter of the i-th reflective element (from the weather or the ground),

<maths num="7"><img id="000008" he="6" wi="7" file="JP6130101B2_D0001.tif" img-format="tif" img-content="drawing" /></maths>

Is the estimated additional chaotic noise difference, and the sum is for all reflective elements updated using current measurements. When the i-th reflective element is a ground element, P_iIs the NRCS uncertainty parameter for ground elements that are initialized based on the type of ground associated with each element. In other words, the initialized uncertainty parameters are based on the certainty of how the ground element reflects radar power. For example, if a ground element is associated with water (eg lake, sea, etc.), it is likely that this ground element will provide little or no reflective effect.

The uncertainty value is essentially the variance of the error in the estimated value. At startup, when initialization of the NRCS value occurs, if the surface type is known, this knowledge allows us to place the boundary at the expected error in the initial value for NRCS. For example, if the surface layer is known to be water, the exact NRCS value is likely to be small and the difference between the exact value and the starting value of zero (ie, starting evaluation error) is small, so the NRCS value. Is initialized to zero and the uncertainty parameter is very small. On the other hand, if the element is a city, the exact NRCS value is expected to have a wide range of values and contains very large values. Therefore, the uncertainty (ie, variance) of the zero starting evaluation must be large.

[0025] In addition to updating the reflection element values, the measurements reduce the uncertainty of the evaluation. Using the same model, the change in P is illustrated by the example in Eq. (9):

<maths num="8"><img id="000009" he="15" wi="137" file="JP6130101B2_D0001.tif" img-format="tif" img-content="drawing" /></maths>

Where P_iIs the uncertainty parameter before the measurement,

<maths num="9"><img id="000010" he="6" wi="4" file="JP6130101B2_D0001.tif" img-format="tif" img-content="drawing" /></maths>

Is the uncertainty parameter after the measurement. Another equation for filter gain K and parameter P is:

<maths num="10"><img id="000011" he="15" wi="130" file="JP6130101B2_D0001.tif" img-format="tif" img-content="drawing" /></maths>

and

<maths num="11"><img id="000012" he="15" wi="137" file="JP6130101B2_D0001.tif" img-format="tif" img-content="drawing" /></maths>

[0027] Those skilled in the art will empirically determine certain factors k._kAnd k_pTo decide. Updating the reflection value of a 3D storage location using radar measurements involves the interaction between the corresponding weight (h) and the corresponding uncertainty parameter (P). Larger values of h correspond to locations near the axis of the antenna main beam. As can be seen from equation (10), if the uncertainty parameters are the same with respect to the storage location, the storage location with the larger value of h tends to receive the largest magnitude of the latest version. Similarly, if h is the same with respect to the storage location, the storage location with the highest relative uncertainty will receive the largest up-to-date version.

[0028] In antenna beam scanning through an array of reflective elements (storage locations), elements near the ray axis get a relatively large modern version and a relatively large reduction in uncertainty. As the rays are scanned away from them, they still receive some up-to-date version. However, elements on the other side of the ray with similar h values receive a larger up-to-date version, as their initial evaluation has more uncertainty. The uncertainty parameter provides a record of the classification in which the antenna beam is in the reflection field buffer (three-dimensional buffer).

[0029] If the radar antenna is indicated by a particular command to obtain a measurement of the backscattered signal as a function of range, the average of these measurements is described by the radar equation. Considering only weather scatter, in one embodiment the radar equation is:

<maths num="12"><img id="000013" he="11" wi="139" file="JP6130101B2_D0001.tif" img-format="tif" img-content="drawing" /></maths>

here,

<maths num="13"><img id="000014" he="5" wi="6" file="JP6130101B2_D0001.tif" img-format="tif" img-content="drawing" /></maths>

Is the signal power, K is a constant that includes the effects of transmission power, loss, etc., Wr (r) is a weighting function in the appropriate range with respect to the time of the received signal power, and G (Ω). ) Is the antenna gain as a function of the direction Ω, and Z (r, Ω) is the reflection distribution.

[0030] In certain embodiments, the received power is

<maths num="14"><img id="000015" he="12" wi="138" file="JP6130101B2_D0001.tif" img-format="tif" img-content="drawing" /></maths>

Approximated by: Here, δr is an increase in the range, and δΩ is an increase in the solid angle. The range weighting function and the antenna gain function are evaluated at discrete points in spatial coordinates (r, Ω). For simplicity, the range weighting function Wr is approximated as roughly constant over the range interval of the range δr. Thus, the above approximation arises:

<maths num="15"><img id="000016" he="12" wi="137" file="JP6130101B2_D0001.tif" img-format="tif" img-content="drawing" /></maths>

The sum over i is close to the integration over the antenna main beam in the angular space (each period of the sum representing the increment instruction in space). Using the bias function expansion of the reflection field, the above sums are divided into groups of sums and made over a particular realization of the bias function. If the bias function is just a rectangular prism of the unrestricted example (called a voxel and the reflection is considered a constant within the range of the voxels), then a single weight for the sum of each group to the voxels. Is determined. Other bias functions can be used. In this case, hk's is:

<maths num="16"><img id="000017" he="11" wi="137" file="JP6130101B2_D0001.tif" img-format="tif" img-content="drawing" /></maths>

[0032] Summing is performed only through the incrementing main beam that enters a particular voxel. Separation of ground scatter from weather scatter can be achieved by simultaneously estimating the distribution of ground NRCS in a weather reflection field assessment. Therefore, the ground radar map is generated as part of estimating the weather reflection field.

[0034] The process for assessing ground NRCS is similar to the process for assessing weather reflections. The received power from backscattering on the ground is

<maths num="17"><img id="000018" he="11" wi="139" file="JP6130101B2_D0001.tif" img-format="tif" img-content="drawing" /></maths>

Approximated by:

<maths num="18"><img id="000019" he="6" wi="16" file="JP6130101B2_D0001.tif" img-format="tif" img-content="drawing" /></maths>

Range r and azimuth angle

<maths num="19"><img id="000020" he="6" wi="4" file="JP6130101B2_D0001.tif" img-format="tif" img-content="drawing" /></maths>

NRCS on the ground at. Antenna gain is evaluated in the direction of the ground at this range and azimuth angle. Topographic elevations are already known. The constant K contains the increment and radar design parameters used in the numerical integration. By extending the NRCS distribution for a set of bias functions, Equation 10 is assigned to:

<maths num="20"><img id="000021" he="9" wi="126" file="JP6130101B2_D0001.tif" img-format="tif" img-content="drawing" /></maths>

[0035] The two components are separable by estimating the weather and ground scattering power. Separation is made because changes in the ground scattering signal with changes in the antenna pointing angle (particularly tilt) change in a predictable way. By making many comparisons of the measurements predicted with different antenna pointing angles, the process iteratively reaches a solution that separates the weather and ground scatter characteristics. The illustrated scanning strategy is to start with a low tilt angle and proceed towards a higher tilt angle. In this example, a good assessment of ground scatter is achieved and is not contaminated by weather scatter (lower tilt angle). These ground scatter assessments can then be used to predict the relative contribution of ground scatter to the next radar measurement, thereby indicating which part of these measurements is of weather scatter. Clarify what you can think of.

[0036] Although preferred embodiments of the invention have been illustrated and described above, many modifications can be made without departing from the spirit and scope of the invention. For example, deductive information is used to initialize both estimates and the associated uncertainty parameters of weather reflections and ground NRCS.

[0037] Embodiments of the present invention are specified by the following claims.