Method for determining the measurement data of solar generators for determining the point of maximum power

When using photovoltaic systems, the degree of efficiency is an important criterion. The costs for the solar cells (photovoltaic generator) represent the largest share of the total costs. to take the maximum possible power (with given irradiation conditions) from the cells.

In order to maximize the performance, a current and a voltage measurement is necessary at the same time. Due to the steep switching edges of the converter, the signals to be measured are superimposed by interference. The error can be reduced by averaging over several measurements. When used in solar generators, the determination of the power can easily take some time (if the power is determined every second, this will generally be completely sufficient).

The prior art is now dealt with as follows. US Pat. No. 5,648,731 (DECKER) describes a method for determining the characteristics of solar panels in a working solar-electric system. In particular, the open-circuit voltage and the short-circuit current, the voltage at the point of maximum power and the ratio of the voltage at the point of maximum power to the open-circuit voltage are recorded. These data can be saved or can be processed into a characteristic curve by adapting a curve to the measured values. A special method for determining the required current and voltage values ​​is not specified.

DE 19 502 762 (RUDISCHER) shows a method for operating a photovoltaic system at the point of maximum power. For this purpose, the solar generator is preferably disconnected from the system every minute and the short-circuit current and open-circuit voltage are determined. Characteristic curves of the solar generator are stored in a microcomputer and, after the measurement, the operating point to be taken from the appropriate characteristic curve is approached. Here, too, there is no specific method for recording measurement data.

EP 0 628 901 A2 (CANON KK) shows an interesting method for determining the MPP and for operating the converter itself. The converter is controlled at several operating points and voltage and current are measured. As many points are evaluated as are necessary to lay a given curve through the determined points (eg three points, if a quadratic function is used). Then the maximum for this curve and, from this, the required operating point is determined. One problem with this method is the undisturbed acquisition of the measurement data; however, no special measures are given.

EP 0 326 489 A1 (ROUZIES) or, depending on it, US Pat. No. 4,899,296, a search method for finding the MPP and simultaneous control of the converter is presented. Depending on the switching thresholds, which are adjustable, the operating point is driven in cyclical motion in the direction of the point of maximum power. No special measure is carried out for the required data acquisition.

In EP 0 090 212 A2 (SIEMENS) the operating point of the converter connected to the solar generator is changed by a small value and the reaction of the differential of the power to it is examined. Normal operation is not disturbed by the small change in the setpoint. But it is precisely these small changes that must be measured with as little interference as possible.

In US 4,263,548 (CARLSON) a measuring device for measuring the amplitude of a current or voltage pulse is shown offset in time to a trigger pulse. The measuring device is used to record signal parameters and to check pulse generators. In the present invention, however, the low-noise detection of a current that is separated from this signal is in the foreground. It is not possible to record the current or voltage value at a specific point in the pulse signal, but rather the current in an electronic power converter with as little interference as possible for further processing. The method proposed here in this application does not have to acquire the current signal at very specific times either, since the interference suppression only takes place through the following randomly uniformly distributed sampling.

The method presented in the application works in principle independently of the selected method for determining the MPP, but is best suited for search algorithms that run simultaneously with the operation of the system.

If one looks at the state of the art of the method for finding the point maximum

AT 409 674 B

Power, the importance of a measurement method that is insensitive to interference is of decisive importance for the relevant quantities. The present invention can make a contribution here.

For reasons of measurement technology, if the circuit topology allows, the current detection will be placed in series with the converter inductance, since the current cannot jump there, or in the supply line to the accumulator in a charger. If there is a current detection in the converter, for example in series with an active switch (for the purpose of overcurrent protection, or because the converter is regulated with current mode control), this can also be used. The current flowing immediately before switching off will then be used as the measured value.

The performance recording can be done by a stochastic ergodic multiplication with little disturbance. The measurement signal can be sampled at a specific point in time with reference to the switch-on or switch-off time, or it can be carried out purely by chance and not correlated with the switching signal of the transistor. A noise signal is used as the reference signal for the comparator (s). When measuring power, two statistically independent noise generators are required.

In the case of an analog implementation, a saw tooth will be used as the reference signal. This is realized by a capacitor that is charged via a current source and discharged via a switch. The variation of the reference sawtooth can be achieved by a different slope, which is controlled by a random signal; the basic frequency thus remains constant, but the pulse width jitters. But you can also use a constant current that is proportional to the solar generator voltage and change the clock frequency via noise. The amplitude of the saw tooth is limited. Another possibility for generating the reference signal is to use the interference signal that is caused by the switching process, to which an offset proportional to the current is added.

The measured value of the current channel is compared with the reference signal; if the measurement signal is higher than the comparison signal, a logical one is output and an average value is then calculated (integration, low-pass filter, for example with an RC element). This achieves a low-interference measurement of the current. If you want to record the power, a measuring channel must be set up for both the voltage and the current, with the two comparator output signals via, for example an EXOR gate can be linked and then the averaging takes place. This means that the power is recorded with little interference and can be used for the algorithm for determining the control signal for driving the active switch.

This can be done very easily digitally. The measured values ​​of current and voltage are converted into digital form with an analog / digital converter, digitally generated noise signals or digitally generated, frequency-varying triangular or trapezoidal voltages are used as reference signals, the required comparison (comparator) and the EXOR link are basic commands and the averaging is done by adding (eg 256 values) and subsequent shifting. This can easily be done with a microcontroller with 8 bit ADC. In addition to the simplicity, the advantage is the great insensitivity to interference, since the multiplication is based on a stochastic ergodic multiplication. The principle of the algorithm for determining the performance is summarized in FIG.

In Figure 1, the basic structure of a measuring channel and its integration into the overall system is shown. The solar generator (1) is connected to the input (V_in) the converter structure (2), which can be either a DC / DC (DC voltage converter) or a DC / AC (inverter) converter, connected. At the output of the converter (V_out) the load (3) is connected. The control signal for the active switch (s) is fed via the control signal feed (4) to the converter (2) from the control signal generator (5). Various data from the power section of the converter (2) can be transmitted to the control signal generator (5). The input voltage signal feed (6) is shown as an example. A measuring channel for the current is also shown. The current-proportional, level-matched signal is either fed directly to the comparator (10) via the current signal feed (7), where it is compared with a reference signal from the reference generator (9), or it is converted from analog to digital (8) and with a digital reference signal from the reference generator (9) compared with a comparator (10) which is then digitally executed and for further processing to the block (11) which contains the evaluation, modulator and processing of the control signal, guided.

AT 409 674 B

The method is also useful for recording the current for pure current maximization when charging the battery, in order to record the current without interference.