Hybrid control method for maximum power point tracking of satellite photovoltaic energy

Abstract

The invention discloses a hybrid control method for maximum power point tracking of satellite photovoltaic energy. The hybrid control method comprises the following steps: modeling a satellite photovoltaic cell and a DC-DC conversion circuit; carrying out parameter initialization on a grey wolf algorithm, carrying out global search on a grey wolf population by utilizing a Levy flight strategy, and outputting a duty ratio by utilizing the grey wolf algorithm; controlling the output power of a satellite photovoltaic cell by using the output duty ratio, judging whether an output power fluctuation value is smaller than a fluctuation threshold value or not, and if so, switching to a model prediction control algorithm to carry out local search of a photovoltaic maximum power point; and collecting the output power of the satellite photovoltaic cell, calculating the output power and the maximum rated output power to obtain a power change percentage, judging whether the power change percentage is greater than or equal to a set threshold, and if so, restarting the grey wolf algorithm to search the duty ratio. The method has the advantages of being fast in convergence, high in steady-state precision, good in stability, low in requirement for a hardware platform and the like.

Description

Technical Field

[0001] The present invention relates to the field of photovoltaic energy control, and particularly to a hybrid control method for maximum power point tracking of satellite photovoltaic energy. Background Art

[0002] With the continuous development of space technology and applications, the power regulation device of the solar array used in satellites has gradually evolved from the traditional direct energy transfer (DET) method to the more efficient maximum power point tracking (MPPT) method to meet the increasing satellite power requirements. Compared with the traditional DET architecture, MPPT can make more effective use of the power of the solar array. For small, lightweight and high-power-density microsatellites, this can effectively improve the working efficiency of the satellite power system, which is obviously more beneficial.

[0003] Based on its own working principle, the solar array has a working point (Maximum Power Point, MPP) that generates the maximum power. To enable the solar array to continuously operate at the MPP to obtain the maximum power output, a maximum power point tracking algorithm is usually required for control. Classic algorithms include the hill climbing method (HC), the perturb and observe method (P&O), the incremental conductance method (INC), and related improved algorithms. The implementation of these algorithms is relatively simple, but they also have obvious drawbacks, that is, after the algorithm reaches a steady state, continuous steady-state oscillations will occur due to the limitations of the algorithm, resulting in power loss in the photovoltaic system. Although increasing the control frequency can alleviate the oscillations to a certain extent, it will also increase the adjustment time required to reach the steady state.

[0004] In view of the shortcomings of the classic algorithms, some intelligent algorithms based on the principle of optimization have emerged, such as the particle swarm optimization algorithm (PSO), the artificial neural network algorithm (ANN), etc. Optimization algorithms such as PSO are prone to falling into a local optimization state, resulting in the inability to track the maximum power point. The ANN algorithm has better global tracking ability, but it is complex to implement, consumes a large amount of computing resources, and requires a large data set as support. For satellites with limited computing power, it is currently limited in implementation. Summary of the Invention

[0005] The purpose of the present invention is to provide a hybrid control method for maximum power point tracking of satellite photovoltaic energy, aiming to solve the problems of slow convergence speed and easy local optimization of existing MPPT algorithms.

[0006] To achieve the above task, the present invention adopts the following technical solutions:

[0007] A hybrid control method for maximum power point tracking of satellite photovoltaic energy, comprising:

[0008] Model the satellite photovoltaic cells and the DC-DC conversion circuit;

[0009] Initialize the parameters of the grey wolf algorithm, where the parameters include duty cycle, population size, and number of iterations; use the duty cycle as the position parameter of each grey wolf in the grey wolf population; use the output power of the satellite photovoltaic cells as the fitness of the grey wolf individuals;

[0010] Improve the search strategy of the grey wolf algorithm, and use the Lévy flight strategy for global search of the grey wolf population;

[0011] Update the positions and real-time fitness of each grey wolf based on the global search, select the top three grey wolves with the highest fitness and set them as the α wolf, β wolf, and δ wolf in sequence, and output the duty cycle corresponding to the position of the α wolf;

[0012] According to the duty cycle output by the grey wolf algorithm, control the output power of the satellite photovoltaic cells based on the models of the DC-DC conversion circuit and the satellite photovoltaic cells, and determine whether the output power fluctuation value is less than the fluctuation threshold. If it is less, switch to the model predictive control algorithm for local search of the photovoltaic maximum power point;

[0013] Collect the output power of the satellite photovoltaic cells, calculate the percentage change in power with the maximum rated output power, and determine whether the percentage change in power is greater than or equal to the set threshold. If it is greater than or equal to, restart the grey wolf algorithm to search for the duty cycle.

[0014] Further, the modeling of the satellite photovoltaic cells and the DC-DC conversion circuit includes:

[0015] Equivalent the satellite photovoltaic cells to a single diode, and set the equivalent parallel resistance and equivalent series resistance, then there is:

[0016] I = I ph - I d - ((V + R s I) / R sh ) (1)

[0017] I D = I 0 [exp((V + R s I) / V T ) - 1] (2)

[0018] V T = AKT / q (3)

[0019] Where V T、V are the thermal voltage and output voltage of the satellite photovoltaic cell respectively; T is the operating temperature of the satellite photovoltaic cell; q is the nuclear charge amount of the satellite photovoltaic cell area; A is the ideal factor of the diode; K is a constant; I D 、I are the diode current and the output current of the satellite photovoltaic cell respectively; I ph and I 0 represent the photocurrent and reverse saturation current of the solar panel, R s and R sh represent the equivalent parallel resistance and equivalent series resistance respectively;

[0020] From equations (1), (2), and (3), its output characteristic equation can be obtained as:

[0021]

[0022] The output voltage of the satellite photovoltaic cell is converted by a DC-DC conversion circuit and provided to the load; A boost converter in the DC-DC converter is used, and its modeling is as follows:

[0023] The boost converter includes an inductor, a capacitor, and a switch. The capacitor is connected in parallel at both ends of the load, and the on-off connection between the inductor and the capacitor is realized through the switch; When the switch is closed, the satellite photovoltaic cell directly stores energy in the inductor; When the switch is opened, the energy stored in the inductor is transferred to the load through the diode, and the input voltage will also discharge to the load through the diode, and the superposition of the two realizes the boost function; The relationship between its input and output voltages is as follows:

[0024]

[0025] In the formula, V represents the output voltage of the satellite photovoltaic cell, V OUT represents the output voltage of the boost converter, and D is the duty cycle;

[0026] After generating the duty cycle D, a PWM wave is generated through a PWM controller, and then it is output to the switch in the boost converter. By controlling the turn-off of the switch, the impedance of the DC-DC converter is adjusted, so as to track the maximum power point to achieve MPPT.

[0027] Furthermore, the global search-based update of each gray wolf position and the real-time fitness includes:

[0028] S4.1 Calculate the initial fitness of each gray wolf position and sort them. Set the gray wolves with the top three initial fitness values as the α wolf, β wolf, and δ wolf in turn, and determine the positions of the top three gray wolves;

[0029] S4.2, Based on the positions and position ratios of the top three gray wolves, calculate the positions of other gray wolf individuals;

[0030] S4.3. Calculate the fitness of each gray wolf in the population based on the updated gray wolf positions, sort based on the updated fitness, and select the top three gray wolves to become the new α-wolf, β-wolf, and δ-wolf;

[0031] S4.4. Determine whether the duty cycle error is less than a preset threshold or whether the number of iterations has reached the maximum number of iterations according to the duty cycle corresponding to the position of the new α-wolf; if so, terminate the iteration and output the duty cycle corresponding to the α-wolf position; if not, return to S4.3 to perform the iteration again.

[0032] Furthermore, it is set that when the output power fluctuation is less than the fluctuation threshold of 0.5, the model predictive control algorithm is switched to perform more accurate local tracking.

[0033] Furthermore, the specific steps of the model predictive control algorithm are as follows:

[0034] Calculate the reference voltage V of the satellite photovoltaic cell through the fitness and duty cycle output by the gray wolf algorithm 1 and the reference current I 1 , and thus calculate the reference power P 1 ; use these parameters as the input of the model predictive control algorithm for local optimization;

[0035] (1) Measure the output voltage V PV (k) and output current I PV (k) of the satellite photovoltaic cell at the current k moment, and at the same time measure the voltage V C (k) across the capacitor of the boost converter Boost circuit;

[0036] (2) Evaluate the cost function: when the switch in the boost converter is closed, evaluate the predicted reference current I PV (k + 1) and reference voltage V PV (k + 1) of the optical satellite photovoltaic cell output at the next moment through equations (10) and (11); when the switch is off, evaluate the predicted reference current I PV (k + 1) and reference voltage V PV (k + 1) of the satellite photovoltaic cell output at the next moment through equations (12) and (13);

[0037] (3) Calculate the reference current i ref (k):

[0038] First, calculate the output power increment ΔP = V PV (k)·I PV (k) - P 1 ; then judge the value of ΔP. When ΔP ≠ 0, if ΔV PV <0, then i ref (k) = I PV(k) - ΔI, if ΔV PV ≥0, then i ref (k) = I PV (k) + ΔI; when ΔP = 0, if ΔV PV >0, then i ref (k) = I PV (k) - ΔI, if ΔV PV ≤0, then i ref (k) = I PV (k) + ΔI; where, ΔV PV = V PV (k) - V 1 、ΔI = I PV (k) - I 1 ;

[0039] (4) Select the switching state by comparing the cost functions through Equation (14); when F s＝1 is less than F s＝0 , the switching state at the next moment is closed, otherwise the switching state at the next moment is open; then apply the switching state to the boost converter;

[0040] When the boost converter switch is closed, the discrete-time sets of its model voltage and current are given by Equations (10) and (11); when the boost converter switch is open, the discrete-time sets of its model voltage and current are given by Equations (12) and (13):

[0041]

[0042] In the formula: The value of ε is V C (k) / (1 - D)I PV (k); T S is the sampling time, L n and v C (k) are the inductance and output capacitance values of the boost converter respectively; I PV (k) and I PV (k + 1) are the output currents of the satellite photovoltaic cells at time k and k + 1, V PV (k) and V PV (k + 1) are the output voltages of the satellite photovoltaic cells at time k and k + 1 respectively;

[0043] The cost function is as follows:

[0044]

[0045] In the formula, F S＝0 、F S＝1 represent the values of the cost functions when the boost converter switch is closed and open respectively, is I calculated by Equation 12PV (k + 1), I calculated for Equation 10 PV (k + 1);

[0046] By calculating F S＝0 and F S＝1 values, when F S＝1 is less than F S＝0 , the switch state in the boost converter is closed, otherwise it is off.

[0047] A terminal device includes a processor, a memory, and a computer program stored in the memory; when the processor executes the computer program, the hybrid control method for maximum power point tracking of satellite photovoltaic energy is implemented.

[0048] A computer-readable storage medium stores a computer program; when the computer program is executed by a processor, the hybrid control method for maximum power point tracking of satellite photovoltaic energy is implemented.

[0049] Compared with the prior art, the present invention has the following technical characteristics:

[0050] 1. The algorithm has faster convergence and can reach the steady-state value more quickly. This algorithm has been compared with the separate GWO algorithm in experimental verification, which proves that the addition of the MPC controller improves the dynamic response of the algorithm.

[0051] 2. The steady-state accuracy is higher, there is no steady-state oscillation, and the power loss of the system will not be caused by oscillation, thus improving the conversion efficiency. The simulation experimental results of this algorithm show that after reaching the steady state, there is no steady-state oscillation and it has a higher output power.

[0052] 3. Under the working conditions of continuous change of the external environment, the MPPT algorithm can also ensure stable operation and has stronger robustness. This algorithm has been experimentally verified and can still stably and accurately track the maximum power point under the condition of continuous change of the external environment, ensuring the stability of on-orbit mission operation.

[0053] 4. The consumption of computing resources is small and the application cost is low. This algorithm has no complex computational complexity. The mentioned MPC algorithm can minimize its cost function through a relatively simple algorithm internally. And its modeling is relatively simple, with low requirements for the computing hardware processor platform, and little occupation of on-orbit resources, which can save more computing resources for other on-orbit single machines. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 It is a model diagram of a satellite photovoltaic energy supply system based on MPPT;

[0055] Figure 2 It is an equivalent circuit diagram of a classic single-diode photovoltaic cell;

[0056] Figure 3 Output curve graph of the photovoltaic system described in the present invention;

[0057] Figure 4 Schematic circuit diagrams of two operating modes of the Boost circuit;

[0058] Figure 5 Schematic diagram of the position update of the Grey Wolf Algorithm described in the present invention;

[0059] Figure 6 General schematic diagram of the application of the Model Predictive Control Algorithm to an electronic power converter;

[0060] Figure 7 MPPT flow chart based on the Model Predictive Control Algorithm;

[0061] Figure 8 Flow chart of the hybrid algorithm of the present invention;

[0062] Figure 9 Comparison graph of the algorithm described in the present invention and the single Grey Wolf Algorithm;

[0063] Figure 10 Output effect diagram of the algorithm described in the present invention under rapid changes in the external environment. Detailed implementation method

[0064] Based on the analysis of the advantages and disadvantages of existing algorithms, the present invention combines the Grey Wolf Algorithm (GWO) with Model Predictive Control (MPC) and makes innovations and improvements. By utilizing the global search ability of GWO and the advantages of MPC in simple control modeling and strong robustness based on the mathematical model of the converter, the tracking control performance of the MPPT controller is comprehensively improved.

[0065] A hybrid control method for maximum power point tracking of satellite photovoltaic energy, comprising:

[0066] S1, modeling the satellite photovoltaic cells and the DC-DC conversion circuit.

[0067] S1.1, first model the satellite photovoltaic cells using the single diode model, and obtain the relationship between the output voltage and output current of the photovoltaic panel by constructing its output characteristic equation.

[0068] Equivalent the satellite photovoltaic cells to a single diode, and set the equivalent parallel resistance and equivalent series resistance, then there is:

[0069] I = I ph - I d - ((V + R s I) / R sh ) (1)

[0070] ID = I 0 [exp((V + R s I) / V T ) - 1] (2)

[0071] V T = AKT / q (3)

[0072] Where V T , V are the thermal voltage and output voltage of the satellite photovoltaic cell respectively; T is the operating temperature of the satellite photovoltaic cell; q = 1.6022×10 -19 C is the nuclear charge amount of the satellite photovoltaic cell area; A is the ideality factor of the diode; K = 1.3806×10 -22 J / K is a constant value and does not change with the change of the external environment; I D , I are the diode current and the output current of the satellite photovoltaic cell respectively; I ph and I 0 represent the photo-generated current and reverse saturation current of the solar panel, R s and R sh represent the equivalent parallel resistance and equivalent series resistance respectively.

[0073] From equations (1), (2), and (3), its output characteristic equation can be obtained as follows:

[0074]

[0075] S1.2, the output voltage of the satellite photovoltaic cell is converted by the DC-DC conversion circuit and provided to the load; build the model of the DC-DC conversion circuit.

[0076] This solution uses the boost converter in the DC-DC converter (Boost) and models it as follows:

[0077] The boost converter (Boost) includes an inductor, a capacitor, and a switch. The capacitor is connected in parallel across the load, and the connection between the inductor and the capacitor is switched on and off through the switch; when the switch is closed, the satellite photovoltaic cell directly stores energy in the inductor; when the switch is open, the energy stored in the inductor is transferred to the load through the diode, and the input voltage also discharges to the load through the diode, and the two are superimposed to achieve the boost function; its input-output voltage relationship is as follows:

[0078]

[0079] Where V represents the output voltage of the satellite photovoltaic cell, V OUT represents the output voltage of the boost converter, and D is the duty cycle.

[0080] The duty cycle D is generated by the MPPT algorithm constructed by this solution, and then a PWM wave is generated by a PWM controller and output to the switch in the boost converter. The impedance of the DC-DC converter is adjusted by controlling the turn-off of the switch so that its output is V OUT , thereby tracking the maximum power point to achieve MPPT.

[0081] S2, Initialize the parameters of the grey wolf algorithm, where the parameters include the duty cycle, the population size, and the number of iterations; use the duty cycle D as the position parameter of each grey wolf in the grey wolf population; use the output power of the satellite photovoltaic cell as the fitness of the grey wolf individual.

[0082] S3, Improve the search strategy of the grey wolf algorithm, and use the Lévy flight strategy for the global search of the grey wolf population.

[0083] S4, Update the positions and real-time fitness of each grey wolf based on the global search, select the grey wolves with the top three fitness values as the α-wolf, β-wolf, and δ-wolf in turn, and output the duty cycle corresponding to the position of the α-wolf.

[0084] In S4, updating the positions and real-time fitness values of the grey wolves based on the global search is specifically as follows:

[0085] S4.1 Calculate the initial fitness of each grey wolf position and sort them. Set the grey wolves with the top three initial fitness values as the α-wolf, β-wolf, and δ-wolf in turn, and determine the positions of the top three grey wolves;

[0086] S4.2, Calculate the positions of other grey wolf individuals based on the positions and position ratios of the top three grey wolves;

[0087] S4.3, Calculate the fitness of each grey wolf in the population according to the updated grey wolf positions, and sort based on the updated fitness. Select the top three grey wolves to become the new α-wolf, β-wolf, and δ-wolf;

[0088] S4.4, According to the duty cycle corresponding to the position of the new α-wolf, judge whether the duty cycle error is less than the preset threshold, or judge whether the number of iterations has reached the maximum number of iterations;

[0089] If so, terminate the iteration and output the duty cycle corresponding to the position of the α-wolf; if not, return to S4.3 and perform the iteration again.

[0090] Among them, the calculation formula for the positions of the top three grey wolves in S4.1 is:

[0091]

[0092] Among them, D α 、D β and D δ are the distances between the α-wolf, β-wolf, and δ-wolf and the peripheral grey wolf (ω-wolf), A1 , A 2 and A 3 is D α , D β and D δ The coefficient vectors of, X α , X β and X δ are the positions of alpha wolf, beta wolf and delta wolf, that is, the current three optimal solutions; X is the position of the peripheral gray wolves following alpha wolf, beta wolf and delta wolf, A 1 and C 1 are the random vectors of alpha wolf, A 2 and C 2 are the random vectors of beta wolf, A 3 and C 3 are the random vectors of delta wolf, X 1 , X 2 and X 3 respectively represent the next positions of the peripheral gray wolves under the guidance of alpha wolf, beta wolf and delta wolf, X (t+1) represents the position of the updated peripheral gray wolves at the (t + 1)-th iteration.

[0093] Where A 1 , A 2 and A 3 , and C 1 , C 2 and C 3 take the values as shown below:

[0094]

[0095] In the formula: is the convergence factor, which linearly decreases from 2 to 0 during iteration; and are uniform random constants within the range of [0, 1].

[0096] S5. According to the duty cycle output by the gray wolf algorithm, based on the model of the DC-DC conversion circuit and the model of the satellite photovoltaic cell, control the output power of the satellite photovoltaic cell, and determine whether the output power fluctuation value is less than the fluctuation threshold. If it is less, switch to the model predictive control algorithm for local search of the photovoltaic maximum power point.

[0097] In order to make the satellite photovoltaic system work at the maximum power point, it is set to switch to the model predictive control algorithm for more accurate local tracking when the output power fluctuation is less than the fluctuation threshold of 0.5; the specific steps of the model predictive control algorithm proposed in the present invention are as follows:

[0098] Calculate the reference voltage V 1 and reference current I 1, from which the reference power P is calculated 1 ; These parameters are used as the input of the model predictive control algorithm for local optimization; The model predictive control algorithm uses the discrete event model of the system to predict the behavior of the control variable over a future duration; The main steps of the model predictive control algorithm are to predict the circuit behavior so as to respond to the control variable, and the predicted variable obtains the optimal switching state by minimizing the cost function; It mainly includes the following steps:

[0099] (1) Measure the output voltage V PV (k) and output current I PV (k) of the satellite photovoltaic cell at the current k moment, and at the same time measure the voltage V C (k) across the capacitor of the Boost converter circuit;

[0100] (2) Evaluate the cost function: When the switch in the boost converter is closed, the reference current I PV (k + 1) and reference voltage V PV (k + 1) of the output of the satellite photovoltaic cell at the next moment are evaluated through equations (10) and (11); When the switch is off, the reference current I PV (k + 1) and reference voltage V PV (k + 1) of the output of the satellite photovoltaic cell at the next moment are evaluated through equations (12) and (13).

[0101] (3) Calculate the reference current i ref (k):

[0102] First, calculate the output power increment ΔP = V PV (k)·I PV (k) - P 1 ; Secondly, judge the value of ΔP. When ΔP ≠ 0, if ΔV PV < 0, then i ref (k) = I PV (k) - ΔI. If ΔV PV ≥ 0, then i ref (k) = I PV (k) + ΔI; When ΔP = 0, if ΔV PV > 0, then i ref (k) = I PV (k) - ΔI. If ΔV PV ≤ 0, then i ref (k) = I PV (k) + ΔI; where, ΔV PV = V PV (k) - V 1 and ΔI = I PV (k) - I 1 .

[0103] (4) Select the switching state by comparing the cost functions through Equation (14); when F s＝1 is less than F s＝0 , the switching state at the next moment is closed (S = ON), otherwise the switching state at the next moment is open (S = OFF); then apply the switching state to the boost converter.

[0104] To apply MPC to MPPT, a discrete model of different switching states of the Boost converter is required. When the boost converter switch is closed (S = ON), the discrete-time sets of its model voltage and current are given by Equations (10) and (11). When the boost converter switch is open (S = OFF), the discrete-time sets of its model voltage and current are given by Equations (12) and (13):

[0105]

[0106] where: the value of ε is V C (k) / (1 - D)I PV (k); T S is the sampling time, L n and v C (k) are the inductance and output capacitance values of the boost converter respectively; I PV (k) and I PV (k + 1) are the output currents of the satellite photovoltaic cells at time k and k + 1, V PV (k) and V PV (k + 1) are the output voltages of the satellite photovoltaic cells at time k and k + 1 respectively.

[0107] The minimized cost function is as follows:

[0108]

[0109] where F S＝0 , F S＝1 represent the values of the cost functions when the boost converter switch is closed and open respectively, is I PV (k + 1) calculated by Equation 12, is I PV (k + 1) calculated by Equation 10.

[0110] By calculating the values of F S＝0 and F S＝1 , when F S＝1 is less than F S＝0 , the switching state in the boost converter is closed, otherwise it is off.

[0111] During the initial global optimization, the Grey Wolf Algorithm is used for global optimization, which facilitates more rapid and accurate location of the maximum power point. Subsequently, it is determined whether the set number of iterations has been reached. When the set number of iterations is reached, it means that the Grey Wolf Algorithm has locked in the area where the maximum power point is located. At this time, the output voltage and output current of the satellite photovoltaic cell obtained after the previous optimization are used again as the input of the model predictive control algorithm, and then the model predictive control algorithm is switched for more accurate tracking.

[0112] S6. Collect the output power of the satellite photovoltaic cell, calculate the percentage change in power with the maximum rated output power, and determine whether the percentage change in power is greater than or equal to the set threshold. If it is greater than or equal to, restart the Grey Wolf Algorithm to search for the duty cycle.

[0113] Embodiment:

[0114] In the example of the present invention, a satellite photovoltaic energy supply system was modeled in Simulink based on the hybrid GWO-MPC algorithm.

[0115] The satellite energy supply system based on the MPPT algorithm mainly consists of a solar array, an MPPT controller, a DC / DC power converter, and a load. Among them, the function of the solar array is to convert the obtained light energy into electrical energy when the satellite is in orbit with sunlight, providing energy for each on-board unit. The function of the MPPT controller is to collect the output current and voltage of the solar array and use an algorithm based on the maximum power tracking principle to make the solar array always operate at the maximum power point.

[0116] In this embodiment, the relevant parameters are as follows. When the external temperature of the solar cell is 25 °C and the irradiance is 1000 W / m 2 . The maximum power is 179.928 W, the short-circuit current is 5.31 A, and the open-circuit voltage is 44.06 V. The voltage value at the maximum power point is 36.72 V, and the current value at the maximum power point is 4.9 A. The modeling of the solar cell refers to Figure 2 , and the single diode model is adopted.

[0117] Reference Figure 3 , which shows the output characteristic curves of this solar cell under different light conditions (1000 W / m 2 , 800 W / m 2 , 600 W / m 2 ). It can be seen from the figure that as the light intensity decreases, the power output by this solar panel also gradually decreases.

[0118] As an example, the DC-DC circuit is implemented using the Boost boost circuit shown in Figure 4 .

[0119] The inspiration for the GWO algorithm comes from the behavior of grey wolves when hunting prey. Based on a four - level hierarchy, grey wolves hunt prey in groups. The leaders of this group are called alpha (α), and they are responsible for all decisions regarding hunting. The secondary leaders who assist the leaders in making decisions are called beta (β). The grey wolves ranked at the third level of this hierarchy are called delta (δ), and they must obey the alpha and beta. Omega (ω) is the lowest rank in the group and must obey all other dominant wolves.

[0120] Figure 6 General schematic diagram for the application of the MPC algorithm to an electronic power converter. Measured variables X(k), such as voltage and current, are used to perform a set of predictions for each prediction horizon N. Considering the constraints, the cost function associates the predicted values with the reference X * (K + 1). A function is generated that returns the optimal control value when minimized.

[0121] Reference Figure 7 , the figure gives the MPPT flow chart based on the MPC algorithm. It is generally divided into the following stages:

[0122] a) Prediction behavior: After the model is established, by deriving the discrete circuit equations, the behavior of the control variables can be predicted at the next sampling time k.

[0123] b) Calculate the reference current: Through its internal algorithm, the reference current I ref of the current state of the model is obtained. It is used to prepare for the subsequent minimization function.

[0124] c) Minimize the cost function: Then the control input obtained in the optimization is applied to the system to minimize the cost function.

[0125] d) Repeat iteration: This behavior continues until the system obtains the ideal output.

[0126] Reference Figure 8 , the figure is the flow chart of the hybrid algorithm proposed by the present invention.

[0127] This algorithm consists of two parts: global optimization and local optimization. The GWO algorithm undertakes the task of global optimization at the initial stage of optimization. When the output power is less than the fluctuation threshold, it switches to the MPC algorithm for local optimization.

[0128] In this way, the advantages of the two algorithms are combined. The MPPT controller can track the MPP faster and with higher accuracy.

[0129] Combined Figure 9 , it is the comparison chart of the output effects of the hybrid algorithm and the GWO algorithm.

[0130] Comparison of Output Metrics between the Proposed Hybrid Algorithm and the GWO Algorithm in Table 1

[0131]

[0132]

[0133] The output results of this model are compared with those of the MPPT algorithm model based on the Grey Wolf algorithm. For the comparison of the two output results, please refer to Figure 9 . Set the light intensity to 1000 W / m 2 , and the temperature to 25 °C.

[0134] It can be seen from the results that the power change of the MPPT modulation algorithm based on the GWO-MPC hybrid algorithm is smoother and faster. The algorithm proposed in this scheme reaches the steady-state value at about 0.58 s, while the traditional Grey Wolf algorithm in this model reaches the steady-state value at about 1.11 s. And the final output efficiency of this hybrid algorithm reaches 99.98%, and the efficiency of the traditional Grey Wolf algorithm is 99.37%. Generally speaking, this hybrid algorithm has a great improvement compared with the Grey Wolf algorithm.

[0135] Refer to Figure 10 for the output effect diagram of this hybrid algorithm under the rapid change of the external environment. Set the initial irradiance to 1000 W / m 2 , change to 900 W / m at 0.7 s 2 , and suddenly change to 800 W / m at 1 s 2 . This algorithm can achieve the MPPT effect every time the environment changes suddenly, and has good environmental adaptability. During the satellite mission, the solar panels carried by the satellite may reduce the irradiance they receive due to unpredictable factors, and this algorithm can still ensure the stability of the on-board energy supply at this time.

[0136] The method of the present invention is not sensitive to the accuracy of modeling parameters, has strong adaptability, and can effectively improve the efficiency, tracking speed and stability of the on-board MPPT controller.

[0137] The above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A hybrid control method for satellite photovoltaic energy maximum power point tracking, characterized in that: include: Modeling of satellite photovoltaic cells and DC-DC conversion circuits; Initialize the parameters of the gray wolf algorithm, including duty cycle, population size and number of iterations; The duty cycle is used as the position parameter of each gray wolf in the gray wolf population; The output power of satellite photovoltaic cells is used as the fitness of individual gray wolves; Improve the search strategy of the gray wolf algorithm and use the Levy flight strategy to conduct a global search of the gray wolf population; Based on global search, the positions and real-time fitness of each gray wolf are updated, and the top three gray wolves with the highest fitness are selected and set as α wolf, β wolf and δ wolf in turn, and the duty cycle corresponding to the position of α wolf is output; According to the duty cycle output by the Grey Wolf algorithm, the output power of the satellite photovoltaic cell is controlled based on the model of the DC-DC conversion circuit and the model of the satellite photovoltaic cell, and it is determined whether the output power fluctuation value is less than the fluctuation threshold. If so, the model predictive control algorithm is switched to perform a local search for the photovoltaic maximum power point. The output power of the satellite photovoltaic cell is collected and the power change percentage is calculated with the maximum rated output power. It is determined whether the power change percentage is greater than or equal to the set threshold. If so, the gray wolf algorithm is restarted to search for the duty cycle.

2. The hybrid control method for satellite photovoltaic energy maximum power point tracking according to claim 1, characterized in that: Modeling of satellite photovoltaic cells, including: The satellite photovoltaic cell is equivalent to a single diode, and the equivalent parallel resistance and equivalent series resistance are set, then: I=I ph -I d -((V+R s I) / R sh ) (1) I D =I0[exp((V+R s I) / V T )-1] (2) V T =AKT / q (3) Where V T , V are the thermal voltage and output voltage of the satellite photovoltaic cell respectively; T is the operating temperature of the satellite photovoltaic cell; q is the nuclear charge of the satellite photovoltaic cell area; A is the ideal factor of the diode; K is a constant; I D , I are the diode current and the output current of the satellite photovoltaic cell respectively; I ph and I0 represent the photocurrent and reverse saturation current of the solar panel, R s and R sh They represent the equivalent parallel resistance and the equivalent series resistance respectively; From equations (1), (2) and (3), we can get the output characteristic equation:

3. The hybrid control method for satellite photovoltaic energy maximum power point tracking according to claim 1, characterized in that: Modeling of DC-DC conversion circuits, including: The output voltage of the satellite photovoltaic cell is converted by the DC-DC conversion circuit and provided to the load; the boost converter in the DC-DC converter is used and modeled as follows: The boost converter includes an inductor, a capacitor and a switch, where the capacitor is connected in parallel at both ends of the load, and the switch is used to realize the connection between the inductor and the capacitor. When the switch is closed, the satellite photovoltaic cell directly stores energy in the inductor. When the switch is open, the energy stored in the inductor is transferred to the load through the diode, and the input voltage is also discharged to the load through the diode. The two are superimposed to realize the boost function. The relationship between the input and output voltage is as follows: Where V represents the output voltage of the satellite photovoltaic cell, V OUT represents the output voltage of the boost converter, and D is the duty cycle; After the duty cycle D is generated, a PWM wave is generated through the PWM controller, which is then output to the switch in the boost converter. The impedance of the DC-DC converter is adjusted by controlling the switch to be turned off, thereby tracking the maximum power point to achieve MPPT.

4. The hybrid control method for satellite photovoltaic energy maximum power point tracking according to claim 1, characterized in that: The updating of the position and real-time fitness of each gray wolf based on global search includes: S4.1 calculates the initial fitness of each gray wolf position and sorts them, sets the gray wolves with the top three initial fitness as α wolf, β wolf and δ wolf in turn, and determines the positions of the top three gray wolves; S4.2, based on the positions and position proportions of the first three gray wolves, calculate the positions of other gray wolves; S4.3, calculating the fitness of each gray wolf in the population according to the updated gray wolf position, and selecting the top three gray wolves to become the new α wolf, β wolf and δ wolf based on the updated fitness ranking; S4.4, based on the duty cycle corresponding to the new α wolf position, determine whether the duty cycle error is less than the preset threshold, or determine whether the number of iterations has reached the maximum number of iterations; if so, terminate the iteration and output the duty cycle corresponding to the α wolf position; if not, return to S4.3 and iterate again.

5. The hybrid control method for satellite photovoltaic energy maximum power point tracking according to claim 1, characterized in that: It is set that when the output power fluctuation is less than the fluctuation threshold of 0.5, it switches to the model predictive control algorithm for more accurate local tracking.

6. The hybrid control method for satellite photovoltaic energy maximum power point tracking according to claim 1, characterized in that: The specific steps of the model predictive control algorithm are as follows: The reference voltage V1 and reference current I1 of the satellite photovoltaic cell are calculated by the fitness and duty cycle output by the grey wolf algorithm, and the reference power P1 is calculated accordingly; these parameters are used as the input of the model predictive control algorithm for local optimization; (1) Measure the output voltage V of the satellite photovoltaic cell at the current time k PV (k) and output current I PV (k) and measure the voltage V across the capacitor of the boost converter Boost circuit at the same time. C (k); (2) Evaluation cost function: When the switch in the boost converter is closed, the reference current I of the photovoltaic cell output of the optical satellite at the next moment is evaluated and predicted by equations (10) and (11): PV (k+1) and reference voltage V PV (k+1); when the switch is turned off, the reference current I output by the satellite photovoltaic cell at the next moment is estimated and predicted by equations (12) and (13) PV (k+1) and reference voltage V PV (k+1); (3) Calculate the reference current i ref (k): First calculate the output power increment ΔP = V PV (k) I PV (k)-P1; secondly, determine the value of ΔP. When ΔP≠0, if ΔV PV <0, then i ref (k) = I PV (k)-ΔI, if ΔV PV ≥0, then i ref (k) = I PV (k)+ΔI; when ΔP=0, if ΔV PV >0, then i ref (k) = I PV (k)-ΔI, if ΔV PV ≤0, then i ref (k) = I PV (k)+ΔI; where ΔV PV =V PV (k) -V1, ΔI = I PV (k)-I1; (4) Select the switching state by comparing the cost function through equation (14); when F s＝1 Less than F s＝0 When , the switch state at the next moment is closed, otherwise the switch state at the next moment is open; then the switch state is applied to the boost converter; When the boost converter switch is closed, the discrete time sets of its model voltage and current are given by equations (10) and (11); when the boost converter switch is open, the discrete time sets of its model voltage and current are given by equations (12) and (13): Where: The value of ε is V C (k) / (1-D)I PV (k); T S is the sampling time, L n and v C (k) are the inductance and output capacitance of the boost converter, respectively; I PV (k) and I PV (k+1) is the output current of the satellite photovoltaic cell at time k and time k+1, V PV (k) and V PV (k+1) are the output voltages of the satellite photovoltaic cells at time k and time k+1 respectively; The cost function is as follows: Where F S＝0 、F S＝1 They represent the cost function values ​​when the boost converter switch is closed and open, respectively. I calculated by formula 12 PV (k+1), I calculated by formula 10 PV (k+1); By calculating F S＝0 and F S＝1 When F S＝1 Less than F S＝0 When , the switch state in the boost converter is closed, otherwise it is off.

7. A terminal device, comprising a processor, a memory and a computer program stored in the memory; characterized in that: When the processor executes the computer program, it implements the hybrid control method for satellite photovoltaic energy maximum power point tracking according to any one of claims 1-6.

8. A computer-readable storage medium, wherein a computer program is stored in the medium; characterized in that: When the computer program is executed by a processor, the hybrid control method for satellite photovoltaic energy maximum power point tracking according to any one of claims 1 to 6 is implemented.

Images