EP0090212B1 - Apparatus for automatically tracking the optimum working point of a d.c. voltage source - Google Patents

Abstract

Solar generators, fuel cells and similar d-c voltage sources have a current-voltage characteristic, on which at one point ("maximum power point" MPP) the maximum power can be taken from the d-c voltage source. In an arrangement, in which a d-c voltage source feeds a consumer through a controllable power converter, the optimum operating point is automatically set by setting a reference value for the voltage or the current into the converter, and impressing a supplemental reference value temporarily thereon as a disturbance variable at certain time intervals. If due to the impression, the output power of the d-c voltage source increases, the reference value is adjusted in the direction of the supplemental reference value. If, on the other hand, the sign of the power change is negative, the reference value is changed opposite to the sign of the supplemental reference value. After a finite number of reference value changes, the instantaneous operating point is this brought to the optimum operating point. Since the sign of the power change is determined through evaluation of the derivative with respect to time of the actual power value, the amplitude of the disturbance variable can be chosen very small, so that the operation of the consumer is not impaired.

Description

The invention relates to a device for automatically setting the optimum working point of a DC voltage source with an internal resistance, in particular a solar generator. Such a device is already known from EP-A-0 029 743.

Such a DC voltage source can e.g. B. an accumulator, a thermocouple, a fuel cell or in particular a solar generator. Apart from the fact that the power emitted by these DC voltage sources depends on non-electrical parameters, such as the ambient temperature, the internal temperature, the state of charge in the case of accumulators or the irradiated power in the case of solar generators, these DC voltage sources have in common that between their two electrical state variables (output voltage and output current) there is a certain physical relationship, which is usually described in the equivalent circuit diagram by an internal resistance. If, therefore, electrical power is drawn from these direct voltage sources by a direct current controller, a voltage converter or another matching converter, which is fed to a downstream load, the maximum output voltage that can be achieved per se drops the more current is drawn via the voltage converter. Conversely, if the voltage converter is controlled or regulated in such a way that a specific output voltage of the DC voltage source is maintained, the current which can be drawn is thus fixed. The voltage source has only a single electrical degree of freedom, which can be specified as the operating point of the voltage source or the adapter. The power output of such a voltage source is determined by a function of the corresponding degree of freedom, i. H. of the operating point, which, in general, reaches its maximum at a certain value which represents the optimum working point with regard to the utilization of the voltage source ("maximum power point", "MPP"). Especially for voltage sources whose primary energy is free (e.g. solar energy) or practically free compared to the installation effort, it is desirable for the system to be used optimally to keep the system running at full load, i.e. always work at the MPP to get as much electrical energy from the DC voltage source into a load, e.g. B. feed an energy storage.

As a load - in connection with a DC chopper or another DC voltage transformer-DC consumer (e.g. the vehicle electrical system) come into question. The DC voltage transformer can also serve as a charge controller for an accumulator and the accumulator can be followed by a controllable inverter which, for. B. the busbar of an "island network", d. H. a remote group of consumers not supplied by a "public utility grid. If a controllable inverter (generally: a controllable power transformer) is used instead of a DC voltage transformer to convert the primary energy consumed by the DC voltage source into another electrical energy in a controllable manner, so AC consumers, such as feed pumps, can also be considered as consumers, which can serve for further energy conversion, for example the conveying work of a medium.

From DE-A-29 03 559 it is already known to control the power consumption of a load by means of a DC voltage converter connected to a solar generator, the DC voltage converter being supplied with a control voltage by which the output voltage of the solar generator is controlled at the optimum voltage value belonging to the optimum operating point shall be. Accordingly, the control voltage is formed from the control deviation of the generator output voltage from a reference voltage, the reference voltage being supplied by a structurally identical but unloaded solar cell in order to take into account the influences of non-electrical environmental variables. However, the influence of a change in the operating point as a result of the current flowing from the DC voltage source (also referred to as a "panel") with its falling characteristic cannot be adequately taken into account by the artificial reference voltage formed by the unloaded measuring cell. Furthermore, scatter of specimens as a result of the manufacturing tolerances leads to incorrect settings of the working point.

In the case of this known device, it is also no longer possible to find the optimum working point in the event of partial shading or contamination of the solar generator or of the unloaded measuring cell providing the reference voltage.

The preamble of claim 1 takes this prior art into account.

Even according to EP-A1-0 029 743, the power output by the panel to a consumer is set by controlling an inverter in particular so that it becomes maximum. For this purpose, a so-called "maximum power tracker" loads the consumer circuit with a disturbance current Δ I at fixed time intervals and evaluates the panel output voltage change AU that occurs. The "maximum power point" itself is defined as the point on the panel U / I characteristic curve with the slope minus one, which is mapped into a reference voltage derived from the panel output voltage itself. A comparison The differential voltage change dΔU / dΔl resulting from the interference current feed-in with the reference voltage enables the operating point to be shifted in the direction of the "maximum power point".

Due to the interference current, power is withdrawn from the panel that is not available to the consumer. Furthermore, u. U. be readjusted to ensure that this remains a measure of the characteristic point with the slope minus one.

According to DE-A-1 513 195, which corresponds to document US-A-3 384 806, a sinusoidal search signal is superimposed on the panel output current in order to optimize the panel operating point. The phase position of the first derivative of the panel output power relative to the search current is evaluated in a demodulator and it is determined whether the current operating point is to the left or right of the "maximum power point". The pulse duty factor of a transistor used as a controllable power transformer is set by the demodulator output signal such that the panel operating point shifts in the direction of the "maximum power point".

A method described in FR-A-2 175 653 uses even more complicated relationships to define and set the "maximum power point".

The invention has for its object to provide a further device for automatically setting the optimal operating point of a DC voltage source, especially a solar generator. The device according to the invention enables, in addition to practical implementations which are carried out entirely in analog technology, embodiments in which digital and / or program-controlled components can also be used, in particular in the control part.

The object is achieved by a device according to claim 1. Advantageous embodiments are specified in the subclaims.

The starting point is accordingly a DC voltage source, in particular a solar generator, which is followed by a controllable power transformer for feeding a consumer. The transmitter is controlled or regulated so that its power consumption, i. H. the electrical power output by the panel is maximum. For this purpose, for a state variable that determines the operating point of the panel, i.e. the panel voltage or panel current, a corresponding setpoint is specified. An additional value in the sense of a disturbance variable is temporarily applied to this setpoint at certain time intervals and the differential change in the panel power caused thereby is recorded. After the activation of the additional value (removal of the disturbance variable), the setpoint is corrected, i. H. permanently changed, the sign of this setpoint change being chosen to be the same as the sign of the additional value if a positive differential change in the panel power was determined during the connection, i. H. the time derivative of the measured power value caused by the activation is positive. However, if the activation of the additional setpoint has resulted in a negative differential change in the panel power output, the direction of correction (the sign of the change in the setpoint) must be selected opposite the sign of the additional setpoint. So there is a setpoint correction that always leads to an operating point with higher panel performance until the MPP is exceeded. From then on, the further corrections cause the operating point to oscillate around the MPP. The smaller the correction steps and the disturbance variable amplitude (additional setpoint) can be, the smaller the fluctuations caused by these oscillations. According to the invention, it is not the change ΔF of the panel power P itself that is evaluated, but rather its time derivative ei, so that even small disturbance variable amplitudes are sufficient to make an exact qualitative statement about the increase or decrease in the panel power.

The Störgrößenamplituden can be selected so low here that she only 1 ⁰ / _0th preferably less than 1%. Cause a change in the panel performance, i.e. practically no longer disturb the actual panel control.

If, on the other hand, the actual power values were compared with one another before and during the feedforward control, the differences in these actual power values would no longer be detectable with the desired reliability given the accuracy of the usual measuring and evaluation devices.

In a particularly simple device, the amount and sign of the additional setpoint are given the same and fixed for all connections. The amount of the setpoint change itself can be determined as a function of the respective change in the panel power caused by the activation of the additional setpoint, whereby the working point is initially quickly approximated to the MPP in the event of large deviations between the maximum power point and the respective working point. However, the method can be carried out even more simply if the amount of the setpoint changes for all setpoint changes is given in the same and fixed manner, in particular the amount of the setpoint changes can be chosen to be smaller than the amount of the additional setpoint.

The change in the output panel power is preferably determined by differentially evaluating the steady state of the panel power before and after the additional setpoint is applied. To do this, the actual power value (for example, slightly smoothed) that is in the stationary state is detected before a disturbance variable feed-in, immediately before the start of the disturbance variable feed-in into a memory, which places this temporarily stored actual value at the input of a differentiating element until, when the disturbance variable is applied, a steady-state actual power value is set, which then inputs the differentiating element instead of temporarily stored actual power value is switched on. This results in a sudden change in ΔP at the input of the differentiator, which produces a large output signal d Δ P / dt even with a very small ΔP.

Because e.g. B. with low lighting intensity of a solar generator a setpoint change only leads to correspondingly smaller, difficult to evaluate changes in the panel power, an unchangeable setpoint can be specified as soon as the power output falls below a set minimum value.

The invention is explained in more detail with reference to four figures and an exemplary embodiment.

Figure 1 shows the course of the current-voltage characteristic of a solar generator and the dependence of the panel power on the degree of freedom of the arrangement. FIG. 2 shows a device for carrying out the method and FIG. 3 shows the most important part of an evaluation circuit for detecting the stationary power change. FIG. 4 shows the control of the individual switching elements of the device.

The relationship between the output voltage U (panel voltage) of a solar generator and the current I drawn (panel current) is plotted in FIG. Furthermore, the solar power P, ie the product of the panel voltage and panel current, is shown. The solar power P has a pronounced maximum P _opt , which on the U / I state diagram corresponds to the values U _opt and I _{opt of} the two electrical state variables U and I. The diagrams shown, which differ slightly even for different panels of the same type, are measured at an irradiation of 930 W / m ² , an ambient temperature of 24 ° C and a panel temperature of 36 ° C. If these external, non-electrical parameters are changed, different diagrams result. With the invention, the optimum operating point, which is given by U _opt and l _opt , is set automatically.

In the following, the case is considered that, according to FIG. 2, a solar generator 1 feeds a consumer 3 via an electrical power transformer 2. In the given case, the power transformer is designed as a DC controller and serves as a charge controller of a battery 3. The terminal voltage of the battery changes only very slightly during a disturbance variable connection, so that the electrical power supplied to the battery and which is taken from the solar generator via the DC controller is practically proportional is the charging current of the battery, which can be measured on the measuring cell 4. The input voltage of the battery is also used to supply the operating voltage for the control unit 6 of the DC chopper and the further control devices via a power supply unit 5.

The aim of the control according to the invention is to regulate the state variable U (in this case, the panel voltage) as the reference variable of the arrangement to the optimum working point U _{o t} , which is done by changing the pulse-pause ratio of the switch contained in the direct current controller 2. As a result, the current flowing through the actuator 2 as the manipulated variable of the system is changed in such a way that it corresponds to the desired working point.

The setpoint generator (7) advantageously provides a basic setpoint Ü _o and a correction setpoint U _corr . which are put together to the Sollwe _o + U _corr . It is initially assumed that the device operates at an operating point which deviates from the optimal operating point (maximum power point MPP), which is given by the voltage _o and is permanently set on an adjusting device 7a in the setpoint generator. The arrangement can be operated in a controlled manner, but regulation can also be provided. So z. B. on a comparator 8, the control deviation between the target value U _o and an actual value U tapped off by means of a corresponding measuring element 9, are formed for the panel voltage in order to obtain the control variable of the control device 6 of the actuator 2.

A time stage 10 now generates a disturbance variable (additional setpoint AU '), which is temporarily applied to the setpoint U _o set on the setpoint generator, for example as an interference voltage surge. B. 8 at comparator If the sign of the additional setpoint .DELTA.U 'is negative, it leads in the example shown in Figure 1 case U _o <U _o p _t to a decrease in abgebebenen of solar generator panel power P. The sign of this power change AP', the caused by the difference of the given at a preset operating point U _o panel power P _o and by the disturbance feedforward .DELTA.U 'panel power goes on there, thus indicating must be changed in the direction U _o to _t to an approximation to U _o p to come. An evaluation circuit 11, which evaluates the time derivative of the panel power output before and during the feedforward control, determines the change in the power output of the solar generator caused by the connection.

Depending on the sign of the change in performance determined by the evaluation circuit, the setpoint value U _o supplied by the setpoint generator is then changed. For this purpose, the setpoint generator 7 advantageously an integrator 7b, to which two antiparallel Zener diodes are connected in parallel to limit the voltage. To adjust the setpoint, it can be provided that the evaluation circuit 11 contains at its output a limit indicator 12, which provides the sign of the power change in the form of a digital signal and in a memory, for. B. a flip-flop circuit 13 inputs. The memory output is connected in such a way that a positive or negative voltage ΔU ₀ (corresponding to an increase or decrease in the power) of a constant amount is provided in accordance with the stored signal. After completion of the connection of the additional setpoint ΔU ', the time stage 10 closes a switch 14 between the memory 13 and the integrator 7, so that the integrator is now briefly connected to the voltage provided by the memory as an input voltage with a sign corresponding to the sign of the differential power change. The integrator sums up these short-term voltage surges ΔU _o , so that the integrator output voltage U _corr = ΣΔU ₀ is corrected accordingly as a correction variable for the basic setpoint value Ü _o .

The setpoint U _o is thus changed by a constant, fixed, predetermined correction amount ΔU _o after each connection. After a finite number of such correction steps, each consisting of a temporary activation of the additional setpoint value ΔU 'and a subsequent permanent setpoint value change by ΔU _o , the maximum power point MPP is thus achieved and the operating point can only be slightly optimized around this for all further connections Commute working point.

The sign and amount of the additional setpoint value ΔU 'is fixed in the given case by the time step 10. Because of the highly sensitive differential detection of the change in performance can .DELTA.U selected 'so that caused by the feed-forward change in the output voltage U o 1% to a maximum of 1%, the voltage U _o p _t in the MPP. The setpoint change ΔU _o is determined by the closing time of the switch 14 and is advantageously chosen such that ΔU _{o is} somewhat smaller than ΔU '.

The time stage 10 also controls a switching device consisting of two switches 16a and 16b within the evaluation circuit 11. In the given case, a current measuring element is sufficient for the evaluation circuit 11 to detect the power output of the DC voltage source, since the terminal voltage of the consumer is therefore the battery input voltage when open - And switching off the disturbance variable remains practically constant and a slow change in the terminal voltage dependent on the state of charge of the battery is of no importance for the differential power change. In other cases, current and voltage must be recorded and multiplied with one another in order to record the power or its differential change.

The switch 16a, which is opened immediately before or at least when the additional setpoint starts to be connected, connects the measuring element 4 (or a downstream actual value smoothing element 17 with a small time constant) to a spoke 18a, in which the one measured before the activation is then connected value of the power output belonging to a steady state of the panel is stored. Even before the end of the connection, as soon as the arrangement has settled to a new stationary value belonging to U _o + ΔU ', the switch 16a is closed again and the memory 18a contains the new stationary measured value. A differentiating element 18c is connected downstream of the memory, wherein the memory and differentiating element can be combined to form a common differentiating device 18, which is shown in FIG. 3.

Memory 18a and switch 16a cooperate in such a way that at the input of the differentiator 18c the respective power measurement value before opening the switch, with the switch open the measured value measured and stored immediately before the disturbance variable connection and after the switch is closed again the measured value, now at U _o + ΔU 'belonging measured value are supplied. Since these measured values are obtained in each case in steady-state conditions, the differentiator 18c thus only records the change in the steady-state power P _stat caused by the disturbance variable, ie its change ΔP _stat . This can e.g. B. are formed by means 18b and is present after the switch 16a is closed again as a surge and is differentiated. The differential change in the steady-state power output of the DC voltage is therefore present at the output of the differentiating element 18c.

According to FIG. 3, the capacitor 31 connected upstream of an operational amplifier 30 acts as a memory which charges when the highly insulating switch 16a is closed in accordance with the input signal applied and retains this charge virtually unchanged until the switch 16a is closed again. The operational amplifier 30 is designed via the capacitance 31 and the resistor 34 as a differentiator and via the RC circuit 33, 34 as additional smoothing. The switch 16b, which is actuated and actuated together with the switch 16a via a control signal S1, prevents currents from flowing out of the differentiating device 18 into a downstream smoothing element 18d during the opening time of 16a. This smoothing member 18d can e.g. B. consist of a passive low-pass filter and an active smoothing element and serve a superimposed AC component of the differentiator output voltage, the harmonics of the actual power value prevails to smooth.

The limit value detector 12 already mentioned detects the sign of this (smoothed) change in power and leads, via the connection already described, by means of the elements 13 and 14 to readjusting the correction setpoint U _corr or the setpoint U _o by the voltage ΔUo.

Furthermore, a further limit value indicator 19 is provided, which checks the actual value of the output power for falling below a minimum value, closes a bypass switch 20 on the integrator 7 and thus disengages the means for adjusting the setpoint value U _o as soon as the output power of the solar generator is so low that a perfect detection in the evaluation circuit 11 is no longer possible.

The interplay between connection of the additional setpoint ΔU 'and readjustment of the setpoint takes place in work cycles which are specified by the time control circuit 10. The duration of such a cycle can e.g. B. 2 seconds and divided into 256 time steps by a corresponding oscillator with a downstream counter.

If there is a risk of disturbances in the power acquisition due to the operating cycle of the actuator 2, the oscillator 21 can be matched to the actuator cycle. The addresses of a memory 22 are controlled in succession with the oscillator pulses, in which the corresponding control pulses for the tracking control are stored for each time step. FIG. 4 shows an example of the course of the corresponding control signals as a function of the time steps n.

At the beginning of a cycle, the initially closed switching device 16a, 16b is opened (control signal S1) and the additional setpoint ΔU 'of the addition point 8 is applied immediately thereafter (voltage S2). If the panel has settled to a steady state actual power value in accordance with the new voltage setpoint U _o + ΔU ', the switching device 16a is closed - with ΔU' still applied. The input voltage of the differentiating element 18c thereby jumps to the new actual power value and a pulse is generated at the differentiator output and the smoothing element 18d, the sign of which is evaluated by the threshold value element 12. When the voltage of the smoothing element has increased approximately to its maximum value, the memory 13 is opened briefly with the control signal S3 and the pending output signal of the threshold value element 12 is stored for the duration of one cycle. Subsequently, the disturbance variable connection ΔU 'is ended and the correction of the setpoint U _corr begins. For this purpose, the output of the memory is given to the integrator 7b for a fixed correction time, the output voltage U _{corr of} which changes by the voltage time area ΔU _o belonging to the signal S4.

The control of the DC chopper shown here primarily acts on the transmitted current via the pulse-pause control, the voltage being set according to the load resistance. Of course, other power converters can also be used.

The device thus makes it possible to track the working point to the optimum working point, all displacements of the optimal working point being taken into account automatically.

Claims (6)

1. A device for automatically adjusting the optimum operating point of a d.c. voltage source (1) having an internal resistance, in particular a solar generator, comprising:

a) an energy transmitter connected to the output of the d.c. voltage source (1), comprising a control device (6) and provided with a theoretical value (U_o) for the output voltage (U₁) of the d.c. voltage source (1),

b) a measuring component (4) which measures a quantity which corresponds to the power (P) emitted from the d.c. voltage source (1) to the load (5, 3), in particular a battery,

c) a theoretical value forming device (7) for the theoretical value (U_o),

d) a superimposition device which temporarily changes the operating point of the d.c. voltage source (1) by superimposing a disturbance variable (ΔU') onto the energy transmitter (2),

e) an analysis circuit (11) which comprises a differentiator device (18) and which determines the sign of the change in the power emitted from the d.c. voltage source (1) produced as a result of the superimposition and which comprises adjusting means (13,14) which adjust (S4) the theoretical value forming device (7) in dependence upon the sign, characterised by

f) a store (18a) provided in the differentiator device (18) which, at the start of a superimposition, samples and stores (S1) the value (P) occurring across the measuring component (4),

g) a switch (16a) provided at the input of the analysis circuit (11) which decouples the analysis circuit from the measuring component (4) at the start of a superimposition, and recouples (S1) the analysis circuit before the end of the superimposition,

h) means (18b) in the differentiator device (18) which determine the change in power (ΔP) thus produced before the end of a superimposition and supply this to a differentiator element (18c)

i) a sign store (13) arranged in the adjusting means (13, 14) into which the sign of the time coefficient of the change in power (sign [5/]) detected via a comparator (12) is input (S₃) at the end of a superimposition and

k) a time stage (10) as superimposing device which emits control signals (S1, S3) for the store (18a), the switch (16a) and the sign store (13) in time-coordinated fashion.

2. A device as claimed in Claim 1, characterised in that by way of disturbance variable the time stage (10) superimposes (8) an additional theoretical value (AU') of a determinate magnitude upon the theoretical value (U_o) for a predetermined period of time.

3. A device as claimed in Claims 1 or 2, characterised in that the time stage (10) contains a store (22) driven by a clock generator (21) and in which the control signals (S1 to S4) required to superimpose the disturbance variable and subsequently adjust the theoretical value forming device are stored in a store-programmed form.

4. A device as claimed in one of the preceding Claims, characterised in that the measuring component (4) is connected into the circuit between the energy transmitter (2) and the load (5,3).

5. A device as claimed in one of the preceding Claims, characterised in that a limit value alarm (19) de-actuates (switch 20) the adjusting means (13,14) when a minimum power output value is undershot.

6. A device as claimed in one of the preceding Claims, characterised in that the theoretical value forming device (7) forms the theoretical value (U_o) as the sum of a fundamental theoretical value (U_o) and a correcting theoretical value (U_korr), for the formation of the fundamental theoretical value, contains a setting-up device (7a) and for the formation of the correcting theoretical value contains an integrator (7b), upon which a predetermined input voltage (ΔU_o) with the sign stored in the sign store (13) is temporarily superimposed following each superimposition of the additional value (Fig. 4, Signal S4).

Images