CN107957743A - A kind of photovoltaic maximum power point method for tracing - Google Patents

Abstract

The present invention relates to a kind of photovoltaic maximum power point method for tracing, including：Choose grey wolf initial position；After photovoltaic system stabilization, the output current and output voltage of the corresponding photovoltaic array of each Xi (n) are measured, and calculate corresponding power P i (n)；Maximum corresponding two grey wolves of two Pi (n) are set to wolf king, and record the position of the wolf king；The grey wolf position Xi (n+1) of iteration n+1 times is determined by grey wolf algorithm；When two wolf kings obtain the approximate location of prey, it is believed that be the approximate location for being found that prey, be directly entered at this time by the local search of golden ratio algorithm performs, otherwise enter back into local search until reaching maximum iteration；Whether detection external environment condition undergos mutation.

Description

A photovoltaic power generation maximum power point tracking method

technical field

The invention relates to the technical field of photovoltaic power generation engineering, in particular to a maximum power point tracking method.

Background technique

In order to alleviate the international energy crisis and reduce environmental pollution, countries all over the world are vigorously developing renewable energy. Among various renewable energy sources, solar energy is considered to be an inexhaustible and important environmentally friendly energy source. In order to efficiently utilize solar energy for photovoltaic power generation, maximum power point tracking (MPPT) technology is indispensable. The main problems faced by MPPT technology can be attributed to two points: First, the output characteristic curve is affected by factors such as light intensity, temperature, load, etc., and the output characteristic curve is nonlinear. Second, in order to avoid the hot spot effect, a bypass diode must be connected in antiparallel to each photovoltaic panel, but this leads to the multi-peak state of the output characteristic curve in the case of partial shading.

Traditional MPPT algorithms such as perturbation and observation method and conductance incremental method have a simple structure but may fall into local extremum. Intelligent algorithms such as particle swarm algorithm, firefly algorithm, wolf pack search algorithm, cuckoo algorithm, etc. are faced with complex structure and calculation time. Longer and other problems, the model-based MPPT method needs to go through tedious formula derivation.

Contents of the invention

The purpose of the present invention is to provide a method for tracking the maximum power point of photovoltaic power generation, which can have both tracking speed and tracking efficiency, and can solve the problem that traditional restart discrimination methods face the problem of restart failure under certain circumstances. The specific technical scheme is as follows:

A photovoltaic power generation maximum power point tracking method, comprising the following steps:

1) Select N points within the duty cycle [0,1] as the initial position of the gray wolf;

2) Let the current iteration be the nth iteration, and the position of the gray wolf is Xi ( _n ), i∈[1,2,...N];

3) After the photovoltaic system is stable, measure the output current and output voltage of the photovoltaic array corresponding to each X _i (n), and calculate the corresponding power P _i (n);

4) sorting the P _i (n);

5) Set the two gray wolves corresponding to the largest two P _i (n) as wolf kings, and record the positions of the wolf kings;

6) The gray wolf position X _i (n+1) of iteration n+1 times is determined by the improved gray wolf algorithm, and the decision weight coefficients of the two wolf kings in the improved gray wolf algorithm are dynamically adjusted along with the hunting process, namely and Wherein W ₁ and W ₂ are respectively the decision-making weight coefficients of the two wolf kings, n is the number of iterations, and n _max is the maximum number of iterations;

7) When tracking, the gray wolf will jump to the possible location of the prey; and when the two wolf kings have obtained the approximate location of the prey, it is considered that the approximate location of the prey has been found, and the golden ratio algorithm is used to directly enter local search, otherwise until the maximum number of iterations is reached and then enter the local search;

8) Detect whether there is a sudden change in the external environment: preset the sudden change threshold ε ₀ , let P ₀ and P ₁ be the power before and after the sudden change, respectively, U ₀ and U ₁ are the voltages before and after the sudden change, respectively, and μ be set to 0.001 to prevent zero denominator, judge |(P ₁ -P ₀ )/(U ₁ -U ₀ +μ)|＞ε ₀ , if satisfied, it is considered that there is a sudden change, restart the algorithm to track the maximum power point, otherwise the system is stable at globally optimal duty cycle.

Description of drawings

Figure 1 shows the maximum power point tracking system using Boost circuit

Figure 2 Flowchart of Hybrid Algorithm

Figure 3 Schematic diagram of expanding the initial search interval

Figure 4 P-D curve of photovoltaic array under specific light mutation

Detailed ways

In order to track the global maximum power point accurately and quickly, the present invention proposes a hybrid control method based on modified gray wolf optimization and golden-section optimization (MGWO-GSO). Firstly, the modified gray wolf optimization (MGWO) is used for global search to determine the optimal local area; then, the golden-section optimization (GSO) is used for local search. Different strategies are adopted at different stages to optimize the control effect. In the original gray wolf algorithm, the decision-making weight of the wolf king remains unchanged, which may cause decision-making lag between the wolf kings in the later decision-making. In the present invention, the decision-making weight of the wolf king changes dynamically with the advancement of hunting, which makes the gray wolf have a clearer hunting goal. Moreover, the present invention proposes a novel restart discrimination method based on the slope of the P-U curve to enhance the reliability of the MPPT system when dealing with sudden changes in illumination.

In order to make the above objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation modes of the present invention will be described in detail below in conjunction with the accompanying drawings. The present invention uses the maximum power point tracking system of the Boost circuit shown in Figure 1, the photovoltaic array adopts a 3×1 photovoltaic array, the switching frequency f=50kHz, C ₁ =100μF, L=0.5mH, C ₂ =100μF, R _load =40Ω. The flow chart of the hybrid algorithm is shown in Figure 2, and the specific scheme is as follows:

The present invention adopts the direct duty cycle control mode, first select N points within the duty cycle [0,1] as the gray wolf initial position X _i (n), i∈[1,2,...N],n Indicates the number of iterations. In the maximum power point tracking algorithm, the global maximum power point is regarded as the prey, and the quality of the prey is replaced by the power;

Select N points within the duty cycle [0,1] as the initial position of the gray wolf Xi ( _n ), i∈[1,2,...N], n represents the number of iterations. In the maximum power point tracking algorithm, the global maximum power point is regarded as the prey, and the quality of the prey is replaced by the power;

After the photovoltaic system is stable, measure the output current I _PV and output voltage V _PV of the photovoltaic array corresponding to each Xi (n), and calculate the corresponding power P _{i (} n);

Sorting the P _i (n), i∈[1,2,...N], n represents the number of iterations;

Set the two gray wolves corresponding to the largest two P _i (n) as wolf kings, and record the positions of the wolf kings and

The gray wolf position X _i (n+1) of the iteration n+1 times is determined by the improved gray wolf algorithm, and the decision weight coefficients of the two wolf kings in the improved gray wolf algorithm are dynamically adjusted along with the hunting process, namely and Wherein W ₁ and W ₂ are respectively the decision-making weight coefficients of the two wolf kings, n is the number of iterations, and n _max is the maximum number of iterations;

When stalking, gray wolves jump to where prey is likely to be. And when the two wolf kings have obtained the approximate location of the prey, it is considered to have found the approximate location of the prey, that is, d _1-2 ≤ 1/80N, where: d _1-2 is the distance between the two head wolves , N is the number of boards connected in series. In this case, it directly enters the local search performed by the golden ratio algorithm, otherwise it does not enter the local search until the maximum number of iterations is reached. However, if the positions of the two head wolves are set as the initial search interval of the golden ratio algorithm, there may be a situation where the global optimal value is not in the initial search interval. Therefore, the length of the initial search interval is expanded _desi to avoid the occurrence of the above situation. The size of d _esi is d _esi =1/40N, and the direction is shown in Figure 3;

Detect whether there is a sudden change in the external environment. When the external conditions change, the maximum global power will change. At this time, the algorithm must be restarted to track the new maximum global power. The restart discrimination mechanism based on detecting the power change rate before and after the mutation is one of the most commonly used restart discrimination methods: |(P ₁ -P ₀ )/P ₀ |＞τ ₀ , where P ₀ and P ₁ are the Power after mutation, _τ0 is the mutation threshold. However, this method has the problem of restart failure in a specific situation similar to that shown in FIG. 4 , for example, the power change before and after the light mutation is so small that the restart discrimination method cannot detect it, that is, τ<τ ₀ . Inspired by the conductance incremental method, the present invention adopts a novel restart discrimination method of the slope of the PU curve to enhance the reliability of MPPT, and the novel restart discrimination method can be expressed as: |(P ₁ -P ₀ )/(U ₁ -U ₀ +μ)|>ε ₀ . In the formula: U ₀ and U ₁ are the voltages before and after the mutation respectively, μ is set to 0.001 to prevent the zero denominator, ε ₀ is the threshold, and ε ₀ is selected as 2 in the present invention. After calculation by the new restart discrimination method, it can be obtained that ε in Fig. 4 is 3.83, which is enough to detect drastic changes in external conditions.

Claims (1)

1. A photovoltaic power generation maximum power point tracking method, comprising the following steps: 1) Select N points within the duty cycle [0,1] as the initial position of the gray wolf; 2) Let the current iteration be the nth iteration, and the position of the gray wolf is Xi ( _n ), i∈[1,2,...N]; 3) After the photovoltaic system is stable, measure the output current and output voltage of the photovoltaic array corresponding to each Xi(n), and calculate the corresponding power P _i (n); 4) Sorting the P _i (n). 5) Set the two gray wolves corresponding to the largest two P _i (n) as wolf kings, and record the positions of the wolf kings; 6) The gray wolf position X _i (n+1) of the iteration n+1 times is determined by the improved gray wolf algorithm. The improvement is: the decision weight coefficients of the two wolf kings are dynamically adjusted with the hunting process, that is, and Wherein W ₁ and W ₂ are respectively the decision-making weight coefficients of the two wolf kings, n is the number of iterations, and n _max is the maximum number of iterations; 7) When tracking, the gray wolf will jump to the possible position of the prey, and when the two wolf kings have obtained the approximate position of the prey, it is considered that the approximate position of the prey has been found, and the golden ratio algorithm is used to directly enter local search, otherwise until the maximum number of iterations is reached and then enter the local search; 8) Detect whether there is a sudden change in the external environment: preset the sudden change threshold ε ₀ , let P ₀ and P ₁ be the power before and after the sudden change, respectively, U ₀ and U ₁ are the voltages before and after the sudden change, and μ be set to 0.001 to prevent zero denominator, judge |(P ₁ -P ₀ )/(U ₁ -U ₀ +μ)|＞ε ₀ , if satisfied, it is considered that there is a sudden change, restart the algorithm to track the maximum power point, otherwise the system is stable at globally optimal duty cycle.