CN106354190B - A kind of photovoltaic maximum power point method for tracing based on multi-objective optimization algorithm - Google Patents

Abstract

The invention discloses a method for tracking the maximum power point of photovoltaic power generation based on a multi-objective optimization algorithm. The method includes the following steps: creating a photovoltaic power generation model, constructing a multi-objective optimization problem; uniformly sampling points for control voltage, and reducing the value of variables Scope; solve the multi-objective optimization problem of the structure, and output the control voltage to obtain the maximum power point. The invention solves the maximum power point of the photovoltaic array based on a multi-objective optimization algorithm, can quickly and accurately track the maximum power point of the photovoltaic array, and improves the efficiency of the photovoltaic power generation system for power generation. At the same time, the invention considers the migration of the maximum power point caused by the change of the illumination intensity and the surface temperature of the solar panel, has extremely high adaptability and accuracy, and can overcome the problem that the global optimum point cannot be found due to local shading.

Description

A method for tracking maximum power point of photovoltaic power generation based on multi-objective optimization algorithm

technical field

The invention relates to the technical field of solar photovoltaic power generation, in particular to a method for tracking the maximum power point of photovoltaic power generation based on a multi-objective optimization algorithm.

Background technique

At present, photovoltaic power generation is the focus of the field of renewable energy, and improving the efficiency of photovoltaic power generation systems is one of the hot issues in current research. One of the effective methods is to track the maximum power point of the power generation system in real time. Since the output characteristics (P-V curve) of photovoltaic cells are nonlinear, and this nonlinearity is affected by factors such as light intensity, battery surface temperature, load, etc., it is very difficult to track the maximum power point in real time. Especially when the photovoltaic power generation system is partially shaded, the P-V curve of the photovoltaic power generation system will present a multi-peak situation. These multi-peaks usually interfere with the maximum power point tracking, greatly reducing the efficiency of the maximum power tracking algorithm, making the system fall into local optimum and unable to find the global optimum.

Traditional maximum power point tracking methods include perturbation and observation (P&O), look-up table method, incremental conductance method (IncCond), etc. They each have advantages and disadvantages. The perturbation-and-observe method is simple to implement and cheap, but it cannot cope with rapid changes in the environment, and it is difficult to determine the tracking step size (too long a step size will cause a sharp change in output power; too short a step size will easily fall into a local optimum and the tracking speed slower). The look-up table method is easy to implement and cheap, but it cannot deal with the partial shading of photovoltaic power generation systems. The incremental conductance method can obtain an effective solution in time when the environment changes rapidly, but its control circuit is complex and it is easy to fall into local optimum. Multi-objective optimization algorithm is an important branch in the field of evolutionary computing, which aims to optimize two or more contradictory objectives at the same time to obtain the best compromise solution. It is widely used in real society and is an effective method to solve optimization problems with multiple objectives in industry.

Utility model content

The purpose of the present invention is to provide a method for tracking the maximum power point of photovoltaic power generation based on a multi-objective optimization algorithm, which can quickly track the maximum power point of photovoltaics for the situation where the intensity of solar radiation changes sharply.

The technical scheme adopted in the present invention is:

A method for tracking the maximum power point of photovoltaic power generation based on a multi-objective optimization algorithm, comprising the following steps A and B;

A: According to the photovoltaic power generation model and the goal of maximum power point tracking, construct two optimization functions

Among them, F ₁ is the error function between the real voltage and current model and the hypothetical control voltage under the unknown voltage and current, that is, the photovoltaic model accuracy function, and F ₂ is the output power under the unknown voltage and current at the current moment, that is, the photovoltaic output power function; V _x and I _x are the model voltage and current, R _s and R _sh are the series and parallel resistances respectively, ns is the number of series resistances, V _t is the thermal saving voltage, I ₀ is the circuit current without light, and I _PV is the photogenerated current;

B: According to the model in step A, the multi-objective optimization algorithm of photovoltaic power generation maximum power point tracking method is transformed into a function optimization problem, that is, to find a set of (V _x , I _x ) that minimizes the value of function F ₁ , and the function F The maximum value of ₂ is enough; any combination of current I _x and voltage V _x is defined to correspond to an individual in the subsequent multi-objective optimization algorithm.

Described B step comprises the following steps:

Step 1: Determine the value range of voltage and current at the maximum power point of the photovoltaic power generation system by sampling the control voltage, current and output power in the photovoltaic power generation system multiple times;

Step 2: Substitute the data collected in step 1, the value range of the voltage and current at the maximum power point, the parameters of the algorithm termination condition, and N individuals into the multi-objective optimization algorithm, and finally output N corresponding control systems with large output power and small model error voltage current pair;

Step 3: Substitute the voltage and current values of N individuals obtained from step 2 into the photovoltaic power generation model

Among them, _IPV is the photogenerated current, _Io is the output current without light, _Np is the number of parallel resistors R _sh , N _s is the number of series resistors R _s , V _t is the junction temperature voltage, and the assumption on the left side of the model is calculated The error of the circuit from the true current.

Step 4: Select the individual whose calculation error is less than 1% as the candidate solution set,

Step 5: Select the voltage value corresponding to the maximum output power from the candidate solution set as the photovoltaic maximum power point tracking control voltage at the current moment.

The multi-objective optimization problem constructed in the claims

Include the following steps:

A1: According to the value range of the determined control voltage _Vx and current _Ix , randomly initialize N individuals, each individual is a combination of a set of voltage and current, and N individuals form the initial population;

B1: according to the initial population obtained in step A1, first adopt simulated binary crossover to generate N new individuals, then carry out polynomial variation to these N new individuals, and the N individuals obtained after the polynomial variation form the offspring population;

The simulated binary cross is as follows:

Randomly select two individuals x ₁ and x ₂ from the population initialized in A1, randomly select a number k on [0,1], and obtain new solutions x ₁ ' and x ₂ ' according to the following form, where X ₁ '= 0.5[(1+k)x ₁ +(1-k)x ₂ ]; X ₂ '=0.5[(1-k)x ₁ +(1+k)x ₂ ]; The polynomial variation is specifically as follows:

Randomly select an individual z, and mutate it into a new individual z', where z'=z+p(z _max -z _min ), where

p ₁ =(zz _min )/z _max -z _min , p ₂ =(z _max -z)/(z _max -z _min )

z _min and z _max are the minimum and maximum values that z can obtain respectively, u is a random number in [0, 1] p ₁ and p ₂ are transitional parameters,

C1. Merge the initial population generated in A1 and the offspring population generated in step B1 to obtain a new population containing 2N individuals, and select N individuals with better quality from it: the selection criteria are non-domination relationship and sparsity relationship , the non-dominated individuals are selected first, and the individuals with a relatively sparse distribution are secondly selected;

D1. For the N non-dominated and relatively sparse individuals obtained in step C1, judge whether the number of algorithm iterations R1, C1, and D1 reaches the critical value. If the number of iterations is not reached, continue to iterate B1, C1, and D1. Step; otherwise, the N non-dominant and relatively sparse individuals selected and retained in step C1 are the individuals required in step 2 of claim 2.

The output voltage and current relationship model of the photovoltaic power generation system is as follows:

Among them, I _PV is the photogenerated current, I _O is the output current without light, N _p is the number of parallel resistance R _sh , N _s is the number of series resistance R _S , V _t is the junction temperature voltage, V is the control voltage, F is Output current; when F=0, the open circuit voltage V=V _max can be obtained according to the dichotomy method, and the control voltage can be output through uniform sampling points Obtain the corresponding output flow I _x =[I ₁ I ₂ ... I _k ], further calculate the output power as P _x = [P ₁ P ₂ ... P _k ] (P _i =V _i ·I _i ), according to the output voltage P The maximum value P _j ＝max[P ₁ P ₂ … P _k ], the range of the maximum photovoltaic power point is V _opt ∈[V _j-1 V _j+1 ], which is 2/k of the original search range, k is the number of sampling points.

The multi-objective optimization algorithm uses non-dominated sorting and environment selection. The population size used in this algorithm is 30, and the number of iterations is 100. The new individual generation method adopted is to simulate binary crossover and polynomial mutation; meanwhile, in order to speed up the optimization efficiency, Set the population size of the multi-objective optimization algorithm to 30, and the termination condition is that the number of iterations is less than 100 or the optimal solution change is less than 0.001;

From the solution set obtained by solving in step 5, the voltage of the individual with the largest output power is selected as the control voltage of the photovoltaic array.

In the step 1, the steps of determining the value range of the control voltage and the current are as follows:

First, evenly collect points for the control voltage and other gaps:

According to the output power corresponding to the sampling point, the optimal control voltage range is obtained: according to the mathematical model of the photovoltaic power generation system, the open circuit voltage is calculated when the model current is 0, and several model voltages are uniformly selected within the range from the open circuit voltage to 0;

Then, calculate the model currents corresponding to the selected model voltages, and further calculate the output powers;

Finally, according to the current maximum output power, determine the range of voltage and current corresponding to the maximum output power point, and use this range as the range of variables to be optimized, that is, the value range of the control voltage and current.

Based on a multi-objective optimization algorithm, the present invention does not need to sample the voltage and current data of the photovoltaic power generation system during the calculation process, and at the same time avoids the problem that the photovoltaic voltage and current model cannot be directly solved, and can track the global maximum power point in real time and quickly to avoid photovoltaic power generation. The system falls into a local optimum in the case of partial shading, which greatly improves the power generation efficiency of the photovoltaic power generation system. At the same time, the invention takes into account the changes in the intensity of light and the surface temperature of the solar panel, has extremely high adaptability and accuracy, and can overcome the problem that the global optimum point cannot be found due to local shading.

Description of drawings

Fig. 1 is the flow chart of the steps of the inventive method;

Fig. 2 is a flow chart of the maximum power point tracking algorithm of the present invention;

Fig. 3 is the flow chart of the multi-objective optimization algorithm involved in the present invention;

Fig. 4 is a flow chart of the effective solution selection method after the multi-objective algorithm ends in the present invention.

detailed description

As shown in Figures 1, 2, 3 and 4, the present invention includes a method for tracking the maximum power point of photovoltaic power generation based on a multi-objective optimization algorithm, comprising the following steps:

A: According to the photovoltaic power generation model and the goal of maximum power point tracking, construct two optimization functions (photovoltaic output power and photovoltaic model accuracy)

Among them, F ₁ is the error function between the real voltage and current model and the assumed control voltage under the unknown voltage and current, F ₂ is the output power at the current moment under the unknown voltage and current, V _x and I _x are the model voltage and current, R _s and R _sh is the series and parallel resistances, ns is the number of series resistances, V _t is the heat-saving voltage, I ₀ is the circuit current without light, and I _PV is the photogenerated current;

B: According to the model in step A, the multi-objective optimization algorithm of photovoltaic power generation maximum power point tracking method is transformed into a function optimization problem, that is, to find a set of (V _x , I _x ) that minimizes the value of function F ₁ , and the function F The maximum value of ₂ is enough; any combination of current I _x and voltage V _x is defined to correspond to an individual in the subsequent multi-objective optimization algorithm.

Described B step comprises the following steps:

Step 1: Determine the value range of voltage and current at the maximum power point of the photovoltaic power generation system by sampling the control voltage, current and output power in the photovoltaic power generation system multiple times;

Step 2: Substitute the data collected in step 1, the determined control voltage, the value range of current, the parameters of the algorithm termination condition and N individuals into the multi-objective optimization algorithm, and finally output N corresponding to large output power and small model error individual;

Step 3: Substitute the voltage and current values of N individuals obtained from step 2 into the photovoltaic power generation model

(IP _PV is the photogenerated current, I _o is the output current without light, N _p is the number of parallel resistors R _sh , N _s is the number of series resistors R _s , V _t is the junction temperature voltage), and calculate the assumption on the left side of the model The error of the circuit from the true current.

Step 4: Select the individual whose calculation error is less than 1% as the candidate solution set,

Step 5: Select the voltage value corresponding to the maximum output power from the candidate solution set as the photovoltaic maximum power point tracking control voltage at the current moment.

The multi-objective optimization algorithm described in step 2 optimizes the multi-objective optimization problem constructed in claim 1

Include the following steps:

A: According to the value range of the determined control voltage _Vx and current _Ix , randomly initialize N individuals, each individual is a combination of a set of voltage and current, and N individuals form the initial population;

B: According to obtaining initial population in step A, first adopt simulated binary crossover to generate N new individuals, then carry out polynomial variation to these N new individuals, the N individuals obtained after polynomial variation constitute subpopulation;

The simulated binary cross is as follows:

Randomly select two individuals x ₁ and x ₂ from the population initialized in A, randomly select a number k on [0,1], and obtain new solutions x ₁ ' and x ₂ ' according to the following form, where x ₁ '= 0.5[(1+k)x ₁ +(1-k)x ₂ ]; x ₂ '=0.5[(1-k)x ₁ +(1+k)x ₂ ]; the polynomial variation is as follows:

Randomly select an individual z, and mutate it into a new individual z', where z'=z+p(z _max -z _min ), where

p ₁ =(zz _min )/z _max -z _min ,p ₂ =(z _max -z)/(z _max -z _min )

z _min and z _max are the minimum and maximum value that z can take respectively, and u is a random number in [0,1]. (p ₁ and p ₂ are transitional parameters)

C. Merge the initial population generated in A and the offspring population generated in B to obtain a new population containing 2N individuals, and select N individuals with better quality from it: the selection criteria are non-domination relationship and sparsity relationship, Prioritize the selection of non-dominated individuals (non-dominated means that there is no individual whose output power and model error are better than the current individual), and secondly select individuals with a relatively sparse distribution (farther away from other points);

D. For the N non-dominated and relatively sparse individuals obtained in step C, judge whether the number of algorithm iterations B, C, and D reaches the critical value. If the number of iterations has not reached, continue to iterate steps B, C, and D Otherwise, the N non-dominant and relatively sparse individuals selected and retained in step C are the individuals required in step 2 of claim 2.

The control voltage-current relationship model of the photovoltaic power generation system is

(I _PV is the photogenerated current, I _o the output current without light, N _p is the number of parallel resistance R _sh , N _s is the number of series resistance R _s , V _t is the junction temperature voltage, V is the control voltage, F is the output current). When F=0, the open circuit voltage V=V _max can be obtained according to the dichotomy method, and the control voltage can be output through uniform sampling points Get the corresponding output flow I _x = [I ₁ I ₂ ... I _k ], further calculate the output power as P _x = [P ₁ P ₂ ... P _k ] (P _i = V _i ·I _i ), according to the output voltage P The maximum value P _j ＝max[P ₁ P ₂ … P _k ], the range of the maximum photovoltaic power point is V _opt ∈[V _j-1 V _j+1 ], which is 2/k of the original search range, k is the number of stepping points.

The multi-objective optimization algorithm uses non-dominated sorting and environment selection (environment selection is based on non-dominated sorting and congestion distance selection), the population size used in this algorithm is 30, the number of iterations is 100, and the new individual generation method adopted is simulated binary Crossover and multinomial mutation. At the same time, in order to speed up the optimization efficiency, the population size of the multi-objective optimization algorithm is set to 30 (the population contains 30 randomly generated individuals defined in step 1), and the termination condition is that the number of iterations is less than 100 or the optimal solution change is less than 0.001 ( This value corresponds to the fastest solution setting within the acceptable error range);

From the solution set obtained by solving, the voltage of the individual with the largest output power is selected as the control voltage of the photovoltaic array.

The steps for determining the value range of the control voltage and current are as follows:

First, evenly collect points for the control voltage and other gaps:

According to the output power corresponding to the sampling point, the optimal control voltage range is obtained: according to the mathematical model of the photovoltaic power generation system, the open circuit voltage is calculated when the model current is 0, and several model voltages are uniformly selected within the range from the open circuit voltage to 0;

Then, calculate the model currents corresponding to the selected model voltages, and further calculate the output powers;

Finally, according to the current maximum output power, determine the range of voltage and current corresponding to the maximum output power point, and use this range as the range of variables (voltage and current) to be optimized, that is, the value range of the control voltage and current.

When the present invention was specifically implemented, an example was given to illustrate:

Step 1: According to the current photovoltaic power generation system, construct a mathematical model of the photovoltaic power generation system, and construct a two-objective multi-objective optimization problem. The mathematical model of the system is and the power generation system model is:

where R _s ＝0.018Ω, R _sh ＝135Ω, k=1.38×10 ^-23 J/K, q＝1.6022×10 ^-19 C, N _p ＝N _s ＝5, I _or ＝6×10 ^-9 A, I _stc =19.4A, K _I =1.7×10 ^-3 , A=1.92,

The multi-objective optimization problem constructed according to this model is

Where V _x and I _x are output voltage and output current.

Step 2: Calculate and obtain the open circuit output voltage according to the mathematical model of the photovoltaic power generation system. When I=0 in the model, construct When F=0, the open circuit voltage V=V _max can be obtained according to the dichotomy method, and it is uniformly selected within the maximum output voltage range Get the corresponding output circuit I _x = [I ₁ I ₂ ... I _k ], further calculate the output power as P _x = [P ₁ P ₂ ... P _k ] (P _i = V _i × I _i ), according to the output voltage P The maximum value P _j ＝max[P ₁ P ₂ … P _k ], the range of the photovoltaic maximum power point is V _opt ∈[V _j-1 V _j+1 ], I _opt ∈[I _j+1 I _{j- 1} ];

Step 3: See question (1) for the optimization goal, the variables are V _opt ∈ [V _j-1 V _j+1 ], I _opt ∈ [I _j+1 I _j-1 ], and the termination condition is set as the number of iterations is less than 100 Or the optimal solution change amount is less than 0.001, the population size is 30, and the multi-objective optimization algorithm is used to solve the output voltage of the maximum power point; among them, the multi-objective algorithm multiplication operator parameters used are: the cross distribution factor is 20, and the variation distribution factor is 20, the crossover probability is 1.0, and the mutation probability is 0.5.

Step 4: According to the solution set calculated by the multi-objective optimization algorithm, select the individual with F ₁ <0.01, select the individual with the largest output power among these individuals, and output the voltage of the individual as the control voltage.

Step 5: When the environment changes, collect the light intensity and the surface temperature of the photovoltaic panel, and repeat steps 1 to 4 to adjust the parameters in real time and track the maximum power point.

As shown in Figure 2, photovoltaic maximum power point tracking consists of two parts. Module 1 completes the construction of multi-objective optimization problems, and module 2 completes the solution of multi-objective optimization problems.

The specific multi-objective optimization algorithm involved is shown in Figure 3. According to the known output voltage and output current range, 30 individuals are randomly generated. Each individual contains two attributes of voltage and current, and the corresponding target value is true. Voltage error and predicted power, by generating a new solution, selecting a better solution and iterating in turn, finally a series of individuals with small error and high power will be obtained, and then the voltage at the point with the largest predicted power will be selected as the output, as shown in Figure 4 Show.

The above is only a preferred implementation example of the present invention, and is not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included in the protection of the present invention. within range.

Claims (7)

1. a kind of photovoltaic maximum power point method for tracing based on multi-objective optimization algorithm, it is characterised in that including following Step A-B：

A：According to the target of photovoltaic generation model and maximum power point tracking, two majorized functions are constructed

Wherein F₁It is under unknown voltage and electric current, real voltage current model is with assuming control voltage error function, i.e. photovoltaic mould Type precision function, F₂It is power output of the current time under unknown voltage electric current, i.e. photovoltaic output power function；Vx and Ix are Model voltage and electric current, R_sAnd R_shIt is series connection and parallel resistance respectively, ns is series resistance quantity, V_tIt is to save thermal voltage, I₀It is nothing Circuital current under illumination, I_PVIt is photogenerated current；

B：According to step A model, the photovoltaic maximum power point method for tracing of multi-objective optimization algorithm is set to be changed into function Optimization problem, as find one group of (V_x,I_x) cause function F₁Numerical value it is minimum, function F₂Numerical value maximum；Definition is appointed Anticipate an electric current I_xWith voltage V_xThe corresponding follow-up multi-objective optimization algorithm of combination in an individual.

2. the photovoltaic maximum power point method for tracing according to claim 1 based on multi-objective optimization algorithm, it is special Sign is：Described step B comprises the following steps 1- steps 5：

Step 1：By repeatedly in photovoltaic generating system acquisition control voltage, electric current and power output, determining photovoltaic generating system The span of maximum power point voltage electric current；

Step 2：Data, the span of maximum power point voltage electric current, the parameter of algorithm end condition that step 1 is collected Multi-objective optimization algorithm is substituted into individual, the N number of corresponding output power of final output and the small control voltage electricity of model error Stream pair；

Step 3：The individual voltage and current value obtained from step 2 is substituted into photovoltaic generation model

Wherein, I_PVIt is photogenerated current, I_oOutput current under no light, N_pIt is parallel resistance R_shQuantity, N_sIt is series resistance R_s Quantity, V_tIt is the warm voltage of section, calculates the error that circuit and real current are assumed on the left of the model；

Step 4：Choose above-mentioned calculation error individual corresponding to less than 1% and be used as candidate's disaggregation；

Step 5：The maximum magnitude of voltage of the corresponding power output of selection is concentrated as current time maximum photovoltaic power point from candidate solution Tracing Control voltage.

3. the photovoltaic maximum power point method for tracing based on multi-objective optimization algorithm described in claim 2, its feature exist In：The multi-objective optimization question constructed in the step A Comprise the following steps：

A1：According to the control voltage V of determination_x, electric current I_xSpan, random initializtion individual, each individual is one group The combination of voltage and current, individual constitute initial population；

B1：According to initial population is obtained in step A1, first intersected using simulation binary system and produce N number of new individual, then to this N Individual new individual carries out multinomial variation, and the individual obtained after multinomial variation forms progeny population；

Described simulation binary system intersects specific as follows：

Two individual x are randomly choosed in the population initialized from A1₁And x₂, a number k is randomly choosed on [0,1], according to such as Lower form obtains new explanation x₁' and x₂', wherein X₁[(1+k) x of '=0.5₁+(1-k)x₂]；X₂[(1-k) x of '=0.5₁+(1+k)x₂]； Described multinomial variation is specific as follows：

To randomly choosing individual z, new individual z ', wherein z '=z+p (z are sported_max- z_min), wherein

p₁=(z-z_min)/z_max-z_min, p₂=(z_max-z)/(z_max-z_min)

z_minAnd z_maxIt is the minimum and maximum value that z can take respectively, u is a random number p in [0,1]₁And p₂It is transitional ginseng Number,

C1. merge in A1 caused progeny population in caused initial population and step B1, obtain one it is individual comprising 2N New population, the individual of N number of better quality is selected from the inside：The standard of selection is non-dominance relation and degree of rarefication relation, is preferentially selected Non-dominant individual is selected, secondly selection is distributed more sparse individual；

D1. to the N number of non-dominant and more sparse individual obtained in step C1, evaluation algorithm iteration B1, C1, D1 step Whether the number of process reaches critical value, if being not reaching to iterations, these steps of continuation iteration B1, C1, D1；Otherwise, The N number of non-dominant and more sparse individual for selecting to retain in step C1 is required in step 2 described in claim 2 Individual.

4. the photovoltaic maximum power point method for tracing based on multi-objective optimization algorithm according to claim 2, its feature It is：The photovoltaic generating system output voltage that uses and current relationship model for

Wherein I_PVIt is photogenerated current, I_oOutput current under no light, N_pIt is parallel resistance R_shQuantity, N_sIt is series resistance R_s's Quantity, V_tIt is the warm voltage of section, V is control voltage, and F is output current；Can be in the hope of open-circuit voltage V according to dichotomy when making F=0 =V_max, pass through uniform sampling site output control voltageOutput stream I corresponding to obtaining_x =[I₁ I₂ ... I_k], it is P further to calculate power output_x=[P₁ P₂ ... P_k](P_i=V_i·I_i), according to output voltage Maximum P in P_j=max [P₁ P₂ ... P_k] to obtain maximum photovoltaic power point location be V_opt∈[V_j-1 V_j+1], should Scope is the 2/k of initial search scope, and k is the number of sampling site.

5. the photovoltaic maximum power point method for tracing according to claim 1 based on multi-objective optimization algorithm, it is special Sign is：The multi-objective optimization algorithm employs non-dominated ranking and environmental selection, the population scale that this algorithm uses for 30, Iterations is 100, and the new individual producing method used is intersected for simulation binary system and multinomial variation；Meanwhile in order to accelerate Optimization efficiency, the population scale for setting multi-objective optimization algorithm are 30, and end condition is that iterations is less than 100 times or optimal Solve knots modification and be less than 0.001.

6. the photovoltaic maximum power point method for tracing according to claim 3 based on multi-objective optimization algorithm, it is special Sign is：The solution that acquisition is solved in step 5 is concentrated, and is chosen the maximum individual voltage of power output as photovoltaic array and is controlled electricity Pressure.

7. the photovoltaic maximum power point method for tracing according to claim 6 based on multi-objective optimization algorithm, it is special Sign is：Determine that control voltage, the span step of electric current are as follows in the step 1：

First, uniform sampling site is carried out to gaps such as control voltages：

The power output correspondingly obtained according to sampling site is optimally controlled section where voltage：According to photovoltaic generating system mathematical modulo Type, the open-circuit voltage when model current is 0 is calculated, some model voltages are uniformly chosen in the range of open-circuit voltage to 0；

Then, model current corresponding with some model voltages chosen is calculated, and each output is further calculated Power；

Finally, according to current peak power output, voltage corresponding to maximum output power point, section where electric current are determined, will The section is control voltage, the span of electric current as the range of variables of problem to be optimized.