阅读说明：本技术 一种光伏组件最大功率追踪方法、系统及存储介质 (Photovoltaic module maximum power tracking method, system and storage medium ) 是由 高怀恩 徐诚 易海芒 潘鑫 于 2021-03-29 设计创作，主要内容包括：本发明公开了一种光伏组件最大功率追踪方法、系统及存储介质,其方法包括：基于光伏组件的开路电压,获取若干个电压分量；将若干个电压分量导入原始径向基神经网络模型,输出评估PV特性曲线；从评估PV特性曲线中提取电压极大值集合和电压极小值集合,从电压极大值集合中获取功率全局最大值对应的第一电压；获取电压极大值集合内的所有局部偶对数据,构成第一偶对集合；从第一偶对集合中提取全局最优功率对应的第二电压,判断第二电压是否等于第一电压；若是,将光伏组件的当前最大输出功率对应的电压更新为第二电压,将当前最大输出功率更新为全局最优功率。本发明可适应光伏组件发生环境动态变化的应用场景,实现对光伏电站的精细化管理。(The invention discloses a method, a system and a storage medium for tracking the maximum power of a photovoltaic module, wherein the method comprises the following steps: acquiring a plurality of voltage components based on the open-circuit voltage of the photovoltaic module; introducing a plurality of voltage components into an original radial basis function neural network model, and outputting an evaluation PV characteristic curve; extracting a voltage maximum value set and a voltage minimum value set from the evaluation PV characteristic curve, and acquiring a first voltage corresponding to a power global maximum value from the voltage maximum value set; acquiring all local even pair data in the voltage maximum value set to form a first even pair set; extracting a second voltage corresponding to the global optimal power from the first even pair set, and judging whether the second voltage is equal to the first voltage or not; if so, updating the voltage corresponding to the current maximum output power of the photovoltaic module to be the second voltage, and updating the current maximum output power to be the global optimum power. The photovoltaic power station management system can adapt to the application scene of the photovoltaic module with dynamic environmental changes, and realizes the fine management of the photovoltaic power station.)

1. A method for tracking maximum power of a photovoltaic module, the method comprising:

based on the change of the working environment of the photovoltaic module, taking the open-circuit voltage of the photovoltaic module as a limiting condition, and acquiring a plurality of voltage components;

introducing the voltage components into an original radial basis function neural network model for operation to generate an evaluation PV characteristic curve;

extracting a voltage maximum value set and a voltage minimum value set from the evaluation PV characteristic curve, and acquiring a first voltage corresponding to a power global maximum value from the voltage maximum value set;

local search is carried out on the voltage maximum value set by utilizing a linear search method, all local even-pair data in the voltage maximum value set are obtained, and a first even-pair set is formed;

extracting global optimal power from the first even pair set, simultaneously obtaining a second voltage corresponding to the global optimal power, and judging whether the second voltage is equal to the first voltage or not;

if so, updating the voltage parameter corresponding to the current maximum output power of the photovoltaic module to the second voltage, and updating the current maximum output power to the global optimum power.

2. The maximum power tracking method for the photovoltaic module according to claim 1, wherein before obtaining a plurality of voltage components by taking the open-circuit voltage of the photovoltaic module as a limiting condition, the method comprises the following steps:

acquiring the current maximum output power of the photovoltaic module, and calculating the absolute deviation value between the current maximum output power and the maximum output power at the previous moment;

judging whether the working environment of the photovoltaic module changes or not according to the comparison result of the absolute deviation value and a preset threshold value;

and after the working environment of the photovoltaic assembly is judged to be unchanged, returning to obtain the current maximum output power of the photovoltaic assembly.

3. The method for maximum power tracking of a photovoltaic module according to claim 1, wherein the obtaining of the voltage components with the open-circuit voltage of the photovoltaic module as a limiting condition comprises:

obtaining the open-circuit voltage V of the photovoltaic module_OCAnd the set V is used as the step value of [0, V_OC]N voltage components are detected in the range, where N = (V)_OC/v+1）。

4. The method of claim 3, wherein the step of introducing the voltage components into an original radial basis function neural network model for operation to generate an estimated PV characteristic curve comprises:

setting a radial basis function based on an original radial basis function neural network model, inputting the N voltage components into the radial basis function for operation, and obtaining N power values associated with the N voltage components;

and combining the N voltage components and the N power values to construct an evaluation PV characteristic curve.

5. The method for tracking the maximum power of a photovoltaic module according to claim 1, wherein the step of performing a local search on the set of voltage maxima by using a linear search method to obtain all local even-pair data in the set of voltage maxima to form a first even-pair set comprises:

based on that the voltage maximum value set contains m voltage maximum values, acquiring an ith local optimal track where the ith (i is more than or equal to 1 and less than or equal to m) voltage maximum value in the m voltage maximum values is located by using a linear search method, and counting the ith group of local even-pair data passing through the ith local optimal track; and sequentially and circularly executing m times to acquire m groups of local even-pair data to form a first even-pair set.

6. The method for maximum power tracking of a photovoltaic module of claim 5, further comprising, after determining whether the second voltage is equal to the first voltage:

if the second voltage is not equal to the first voltage, the original radial basis function neural network model is trained and updated by using the voltage maximum value set and the voltage minimum value set, and then the plurality of voltage components are led into the trained original radial basis function neural network model for operation.

7. The method for maximum power tracking of a photovoltaic module according to claim 6, wherein the training and updating the raw radial basis function neural network model using the set of voltage maxima and the set of voltage minima comprises:

acquiring jth even-pair data to which jth (j is more than or equal to 1 and less than or equal to n) voltage minimum values in the n voltage minimum values belong on the basis that the voltage minimum value set comprises n voltage minimum values; sequentially and circularly executing n times to obtain n even pair data to form a second even pair set;

and importing m groups of local even pair data contained in the first even pair set and n even pair data contained in the second even pair set into the original radial basis function neural network model, and training and updating the original radial basis function neural network model based on a recursive least square method.

8. A photovoltaic module maximum power tracking system, the system comprising:

the photovoltaic component acquisition module is used for acquiring a plurality of voltage components based on the change of the working environment of the photovoltaic module and by taking the open-circuit voltage of the photovoltaic module as a limiting condition;

the characteristic curve generation module is used for introducing the voltage components into an original radial basis function neural network model for operation to generate an evaluation PV characteristic curve;

the extreme value set extraction module is used for extracting a voltage maximum value set and a voltage minimum value set from the evaluation PV characteristic curve and acquiring a first voltage corresponding to a power global maximum value from the voltage maximum value set;

the data searching module is used for carrying out local searching on the voltage maximum value set by utilizing a linear searching method to obtain all local even-pair data in the voltage maximum value set to form a first even-pair set;

the voltage judgment module is used for extracting global optimal power from the first even-pair set, acquiring a second voltage corresponding to the global optimal power, and judging whether the second voltage is equal to the first voltage or not;

and the parameter updating module is used for updating the voltage parameter corresponding to the current maximum output power of the photovoltaic module to the second voltage and updating the current maximum output power to the global optimal power after judging that the second voltage is equal to the first voltage.

9. The photovoltaic module maximum power tracking system of claim 8, further comprising:

and the model training module is used for training and updating the original radial basis function neural network model by utilizing the voltage maximum value set and the voltage minimum value set after judging that the second voltage is not equal to the first voltage, and then returning to the characteristic curve generating module for operation again.

10. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out a method for maximum power tracking of a photovoltaic module according to any one of claims 1 to 7.

Technical Field

The invention relates to the technical field of photovoltaic power generation, in particular to a method and a system for tracking the maximum power of a photovoltaic module and a storage medium.

Background

Photovoltaic module that photovoltaic power plant used comprises a plurality of photovoltaic cell pieces through the series-parallel connection form, receive sheltered from at photovoltaic module's partial battery piece, ageing, damage under the circumstances that physical factor influences such as damage, photovoltaic module's volt-ampere characteristic curve can produce mismatch effect, and also can cause the influence to associated photovoltaic PV characteristic curve according to the different circumstances of mismatch, this photovoltaic PV characteristic curve can not keep original class parabola shape promptly, but has a plurality of convex surfaces and concave surface, can improve the degree of difficulty to photovoltaic module maximum power tracking work undoubtedly greatly. Aiming at the traditional maximum power tracking work, a hill climbing algorithm with a simple and convenient implementation mode is independently adopted to quickly find the pole on the photovoltaic PV characteristic curve, but the algorithm cannot span a local trap and can only stay at a local optimal point, and the method is difficult to adapt to the application scene of the photovoltaic module with dynamic environmental changes.

Disclosure of Invention

The invention aims to overcome the defects of the prior art, and provides a method, a system and a storage medium for tracking the maximum power of a photovoltaic module, which can be suitable for the application scene of the photovoltaic module with dynamic environmental changes so as to realize the fine management of a photovoltaic power station.

In order to solve the above problem, the present invention provides a method for tracking maximum power of a photovoltaic module, where the method includes:

based on the change of the working environment of the photovoltaic module, taking the open-circuit voltage of the photovoltaic module as a limiting condition, and acquiring a plurality of voltage components;

introducing the voltage components into an original radial basis function neural network model for operation to generate an evaluation PV characteristic curve;

extracting a voltage maximum value set and a voltage minimum value set from the evaluation PV characteristic curve, and acquiring a first voltage corresponding to a power global maximum value from the voltage maximum value set;

local search is carried out on the voltage maximum value set by utilizing a linear search method, all local even-pair data in the voltage maximum value set are obtained, and a first even-pair set is formed;

extracting global optimal power from the first even pair set, simultaneously obtaining a second voltage corresponding to the global optimal power, and judging whether the second voltage is equal to the first voltage or not;

if so, updating the voltage parameter corresponding to the current maximum output power of the photovoltaic module to the second voltage, and updating the current maximum output power to the global optimum power.

Optionally, before obtaining a plurality of voltage components with the open-circuit voltage of the photovoltaic module as a limiting condition, the method includes:

acquiring the current maximum output power of the photovoltaic module, and calculating the absolute deviation value between the current maximum output power and the maximum output power at the previous moment;

judging whether the working environment of the photovoltaic module changes or not according to the comparison result of the absolute deviation value and a preset threshold value;

and after the working environment of the photovoltaic assembly is judged to be unchanged, returning to obtain the current maximum output power of the photovoltaic assembly.

Optionally, the obtaining a plurality of voltage components with the open-circuit voltage of the photovoltaic module as a limiting condition includes:

obtaining the open-circuit voltage V of the photovoltaic module_OCAnd the set V is used as the step value of [0, V_OC]N voltage components are detected in the range, where N = (V)_OC/v+1）。

Optionally, the introducing the voltage components into the original radial basis function neural network model for operation, and generating the estimated PV characteristic curve includes:

setting a radial basis function based on an original radial basis function neural network model, inputting the N voltage components into the radial basis function for operation, and obtaining N power values associated with the N voltage components;

and combining the N voltage components and the N power values to construct an evaluation PV characteristic curve.

Optionally, the performing local search on the voltage maximum value set by using a linear search method to obtain all local even-pair data in the voltage maximum value set, and forming a first even-pair set includes:

based on that the voltage maximum value set contains m voltage maximum values, acquiring an ith local optimal track where the ith (i is more than or equal to 1 and less than or equal to m) voltage maximum value in the m voltage maximum values is located by using a linear search method, and counting the ith group of local even-pair data passing through the ith local optimal track; and sequentially and circularly executing m times to acquire m groups of local even-pair data to form a first even-pair set.

Optionally, after determining whether the second voltage is equal to the first voltage, the method further includes:

if the second voltage is not equal to the first voltage, the original radial basis function neural network model is trained and updated by using the voltage maximum value set and the voltage minimum value set, and then the plurality of voltage components are led into the trained original radial basis function neural network model for operation.

Optionally, the training and updating the original radial basis function neural network model by using the set of voltage maximum values and the set of voltage minimum values includes:

acquiring jth even-pair data to which jth (j is more than or equal to 1 and less than or equal to n) voltage minimum values in the n voltage minimum values belong on the basis that the voltage minimum value set comprises n voltage minimum values; sequentially and circularly executing n times to obtain n even pair data to form a second even pair set;

and importing m groups of local even pair data contained in the first even pair set and n even pair data contained in the second even pair set into the original radial basis function neural network model, and training and updating the original radial basis function neural network model based on a recursive least square method.

In addition, the embodiment of the invention also provides a photovoltaic module maximum power tracking system, which comprises:

the photovoltaic component acquisition module is used for acquiring a plurality of voltage components based on the change of the working environment of the photovoltaic module and by taking the open-circuit voltage of the photovoltaic module as a limiting condition;

the characteristic curve generation module is used for introducing the voltage components into an original radial basis function neural network model for operation to generate an evaluation PV characteristic curve;

the extreme value set extraction module is used for extracting a voltage maximum value set and a voltage minimum value set from the evaluation PV characteristic curve and acquiring a first voltage corresponding to a power global maximum value from the voltage maximum value set;

the data searching module is used for carrying out local searching on the voltage maximum value set by utilizing a linear searching method to obtain all local even-pair data in the voltage maximum value set to form a first even-pair set;

the voltage judgment module is used for extracting global optimal power from the first even-pair set, acquiring a second voltage corresponding to the global optimal power, and judging whether the second voltage is equal to the first voltage or not;

and the parameter updating module is used for updating the voltage parameter corresponding to the current maximum output power of the photovoltaic module to the second voltage and updating the current maximum output power to the global optimal power after judging that the second voltage is equal to the first voltage.

Optionally, the system further includes:

and the model training module is used for training and updating the original radial basis function neural network model by utilizing the voltage maximum value set and the voltage minimum value set after judging that the second voltage is not equal to the first voltage, and then returning to the characteristic curve generating module for operation again.

In addition, an embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored, where the computer program, when executed by a processor, implements the method for tracking the maximum power of the photovoltaic module described in any one of the above.

In the embodiment of the invention, the description of the evaluation PV characteristic curve by the radial basis function neural network model is updated in real time by adopting an iterative correction mode, meanwhile, any local optimal point of the evaluation PV characteristic curve can be rapidly inquired by combining a traditional linear search method, the maximum output power of the photovoltaic module is accurately tracked by utilizing a global comparison mode, and the method is suitable for an application scene of the photovoltaic module with dynamic environment change, so that the fine management of a photovoltaic power station is realized, the operation and maintenance cost is reduced, and the overall power generation efficiency is improved.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.

Fig. 1 is a schematic flow chart of a method for tracking maximum power of a photovoltaic module according to an embodiment of the present invention;

FIG. 2 is a diagram illustrating extreme value definition on a curve according to an embodiment of the present invention;

fig. 3 is a schematic composition diagram of a photovoltaic module maximum power tracking system according to an embodiment of the invention.

Detailed Description

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

Examples

Referring to fig. 1, fig. 1 is a schematic flow chart illustrating a method for tracking maximum power of a photovoltaic module according to an embodiment of the present invention.

As shown in fig. 1, a method for tracking maximum power of a photovoltaic module includes the following steps:

s101, obtaining the current maximum output power of a photovoltaic module, and calculating an absolute deviation value between the current maximum output power and the maximum output power at the previous moment;

the implementation process of the invention is as follows: with a fixed voltage parameter V_MPP0For the tracking point, acquiring the power value P of the photovoltaic module at the tracking point at the current moment_MPP1Simultaneously, the power value P of the photovoltaic module at the tracking point at the last moment is adjusted_MPP0Then, the absolute deviation value between the two power values is calculated as P_A=|P_MPP0-P_MPP1L. It should be noted that the working environment of the photovoltaic module at the previous time is not changed.

S102, judging whether the working environment of the photovoltaic module changes or not according to the comparison result of the absolute deviation value and a preset threshold value;

the implementation process of the invention is as follows: based on the absolute deviation value P_AIf the value is larger than the preset threshold value, indicating that the working environment of the photovoltaic module at the current moment changes, continuing to execute the step S103; based on the absolute deviation value P_AIf the current time is less than or equal to the preset threshold value, which indicates that the working environment of the photovoltaic module does not change at the current time, the step S101 is executed again, and the power verification at the next time is performed.

S103, based on the change of the working environment of the photovoltaic assembly, taking the open-circuit voltage of the photovoltaic assembly as a limiting condition, and obtaining a plurality of voltage components;

the implementation process of the invention is as follows: obtaining the open-circuit voltage V of the photovoltaic module_OCAnd the set V is used as the step value of [0, V_OC]N voltage components are detected in the range, where N = (V)_OC/v+1）。

S104, introducing the voltage components into an original radial basis function neural network model for operation to generate an evaluation PV characteristic curve;

the implementation process of the invention is as follows: firstly, setting a radial basis function based on an original radial basis function neural network model, inputting the N voltage components into the radial basis function for operation, and obtaining N power values associated with the N voltage components, namely, each voltage component in the N voltage components has a corresponding power value; next, an estimated PV characteristic curve is constructed combining the N voltage components and the N power values, and the P-axis on the curve can be used to delineate the N power values, while the V-axis can be used to delineate the N voltage components.

S105, extracting a voltage maximum value set and a voltage minimum value set from the evaluation PV characteristic curve, and acquiring a first voltage corresponding to a power global maximum value from the voltage maximum value set;

in the embodiment of the present invention, according to the extreme value definition diagram on the curve shown in fig. 2, the extreme value extraction criteria are set as follows: the value of the solid squares is defined as the maximum value on the curve and the value of the solid origin is defined as the minimum value on the curve, without considering the two end points of the curve.

The implementation process of the invention is as follows: firstly, based on the extreme value extraction standard, respectively extracting all m voltage maximum values and n voltage minimum values from the evaluation PV characteristic curve so as to form a voltage maximum value set V_m={v₁,v₂,…,v_mAnd a set of voltage minima U_n={u₁,u₂,…,u_n}; second from the set of voltage maxima V_mDirectly acquiring the voltage maximum value with the maximum value, and defining the voltage maximum value as the first voltage v corresponding to the power global maximum value_mpp1。

S106, carrying out local search on the voltage maximum value set by using a linear search method to obtain all local even-pair data in the voltage maximum value set to form a first even-pair set;

the implementation process of the invention is as follows: set of V values based on the voltage maxima_mThe method comprises the steps of obtaining an ith local optimal track where the ith (i is more than or equal to 1 and less than or equal to m) voltage maximum value in m voltage maximum values is located by utilizing a linear search method, and counting the ith group of local even-pair data passing through the ith local optimal track; and sequentially and circularly executing m times to obtain m groups of local even pair data to form a first even pair set EV.

More specifically, first, a first voltage maximum v is obtained by a linear search method₁The local optimal track is v_1,0→v_1,1→v_1,2→···→v_1,k1And measuring k1 power values corresponding to k1 local voltage points contained in the local optimal trajectory, wherein the power values are respectively represented as p_1,0→p_1,1→p_1,2→···→p_1,k1Therefore, the set of local even pair data passing through the local optimal track can be counted as { v_1,0，p_1,0}→{v_1,1，p_1,1}→{v_1,2，p_1,2}→···→{v_1,k1，p_1,k1}; sequentially carrying out local search and statistics on the rest (m-1) voltage maximum values according to the method to obtain a first even pair set EV = { { v { (v) }_1,0，p_1,0},…,{v_1,k1，p_1,k1},…,{v_m,0，p_m,0},…,{v_m,km，p_m,km}}。

S107, extracting global optimal power from the first even pair set, simultaneously obtaining a second voltage corresponding to the global optimal power, and judging whether the second voltage is equal to the first voltage or not;

the implementation process of the invention is as follows: firstly, obtaining a maximum power value p from the first even pair set EV_mppAnd defined as a global optimum power, and according to the power value p_mppDirectly acquiring the corresponding second voltage v by the local even-pair data_mpp2(ii) a Secondly, judging the second voltage v_mpp2Is equal to the first voltage v_mpp1And the corresponding judgment result comprises: if v is_mpp2=v_mpp1Then, go on to step S108; if v is_mpp2≠v_mpp1Then, the step S109 is skipped.

S108, updating a voltage parameter corresponding to the current maximum output power of the photovoltaic module to the second voltage, and updating the current maximum output power to the global optimal power;

the implementation process of the invention is as follows: according to the parameter value obtained in step S101, the current maximum output power P is obtained_MPP1Corresponding voltage parameter V_MPP0Updated to the second voltage v_mpp2I.e. to say that the fixed voltage parameter v will start from the next moment_mpp2As a new tracking point, while simultaneously setting the current maximum output power P_MPP1Updating to the global optimum power p_mppSo as to realize the maximum power tracking of the photovoltaic assembly under the new environment.

S109, training and updating the original radial basis function neural network model by using the voltage maximum value set and the voltage minimum value set.

The implementation process of the invention comprises the following steps:

(1) based on the set of voltage minima U_nThe method comprises the steps that n voltage minimum values are included, and j-th even-pair data which j (j is more than or equal to 1 and less than or equal to n) th voltage minimum values in the n voltage minimum values belong to are obtained; sequentially and circularly executing n times to obtain n even pair data to form a second even pair set;

specifically, first, a first voltage minimum value u may be obtained according to the estimated PV characteristic curve₁Corresponding power value p_u1Thereby obtaining the voltage minimum u₁Associated pair data { u }₁，p_u1}; sequentially carrying out data statistics on the remaining (n-1) voltage minimum values according to the method, and obtaining a second even pair set EU = { { u { (u) }₁，p_u1},{u₂，p_u2},…,{u_n，p_un}}。

(2) And importing m groups of local even data contained in the first even set EV and n even data contained in the second even set EU into the original radial basis function neural network model, training and updating the original radial basis function neural network model based on a recursive least square method, and returning to re-execute the step S104, wherein the operated original radial basis function neural network model is in an updated state.

The training and updating of the original radial basis function neural network model adopts the following recursive least square operation:

P(n)=P(n-1)-[P(n-1)K(n)K^T(n)P(n-1)]/[1+K^T(n)P(n-1)K(n)]

W(n)=W(n-1)+P(n)K(n)[d(n)-W^T(n-1)K(n)

in the formula: n is the number of iterations, d (n) is a data set used in the nth iteration, and is composed of the first even pair set EV and the second even pair set EU, P (n-1) is a variable parameter in the (n-1) th iteration, P (n) is a variable parameter in the nth iteration, and an initial value P (0) is a diagonal matrix, k (n) is a regression coefficient in the nth iteration under the condition of linear regression, W (n-1) is a weight parameter from the hidden layer to the output layer in the (n-1) th iteration, W (n) is a weight parameter from the hidden layer to the output layer in the nth iteration, and T is a transposed symbol.

In the embodiment of the invention, the description of the evaluation PV characteristic curve by the radial basis function neural network model is updated in real time by adopting an iterative correction mode, meanwhile, any local optimal point of the evaluation PV characteristic curve can be rapidly inquired by combining a traditional linear search method, the maximum output power of the photovoltaic module is accurately tracked by utilizing a global comparison mode, and the method is suitable for an application scene of the photovoltaic module with dynamic environment change, so that the fine management of a photovoltaic power station is realized, the operation and maintenance cost is reduced, and the overall power generation efficiency is improved.

Examples

Referring to fig. 3, fig. 3 is a schematic diagram illustrating a composition of a maximum power tracking system of a photovoltaic module according to an embodiment of the invention.

As shown in fig. 3, a photovoltaic module maximum power tracking system, the system comprising:

the voltage component acquiring module 201 is configured to acquire a plurality of voltage components based on a change in a working environment of a photovoltaic module, with an open-circuit voltage of the photovoltaic module as a limiting condition;

the implementation process of the invention is as follows: obtaining the open-circuit voltage V of the photovoltaic module_OCAnd the set V is used as the step value of [0, V_OC]N voltage components are detected in the range, where N = (V)_OC/v+1）。

Before that, the voltage component obtaining module 201 further has an environment pre-judging function, which is specifically represented as:

(1) acquiring the current maximum output power of the photovoltaic module, and calculating the absolute deviation value between the current maximum output power and the maximum output power at the previous moment;

in particular, with a fixed voltage parameter V_MPP0For the tracking point, acquiring the power value P of the photovoltaic module at the tracking point at the current moment_MPP1Simultaneously, the power value P of the photovoltaic module at the tracking point at the last moment is adjusted_MPP0Then, the absolute deviation value between the two power values is calculated as P_A=|P_MPP0-P_MPP1L. It should be noted that the working environment of the photovoltaic module at the previous time is not changed.

(2) And judging whether the working environment of the photovoltaic module changes or not according to the comparison result of the absolute deviation value and a preset threshold value.

In particular, based on the absolute deviation value P_AIf the voltage component is larger than the preset threshold value, the photovoltaic module at the current moment is changed in working environment, and the next step of voltage component acquisition is continuously executed; based on the absolute deviation value P_AAnd if the current maximum output power is less than or equal to the preset threshold, the photovoltaic module is not changed in working environment at the current moment, and then the current maximum output power of the photovoltaic module is returned to be continuously obtained so as to carry out power verification at the next moment.

A characteristic curve generation module 202, configured to introduce the voltage components into an original radial basis function neural network model for operation, so as to generate an estimated PV characteristic curve;

the implementation process of the invention is as follows: firstly, setting a radial basis function based on an original radial basis function neural network model, inputting the N voltage components into the radial basis function for operation, and obtaining N power values associated with the N voltage components, namely, each voltage component in the N voltage components has a corresponding power value; next, an estimated PV characteristic curve is constructed combining the N voltage components and the N power values, and the P-axis on the curve can be used to delineate the N power values, while the V-axis can be used to delineate the N voltage components.

An extreme value set extraction module 203, configured to extract a voltage maximum value set and a voltage minimum value set from the estimated PV characteristic curve, and obtain a first voltage corresponding to a power global maximum value from the voltage maximum value set;

in the embodiment of the present invention, according to the extreme value definition diagram on the curve shown in fig. 2, the extreme value extraction criteria are set as follows: the value of the solid squares is defined as the maximum value on the curve and the value of the solid origin is defined as the minimum value on the curve, without considering the two end points of the curve.

The implementation process of the invention is as follows: firstly, based on the extreme value extraction standard, respectively extracting all m voltage maximum values and n voltage minimum values from the evaluation PV characteristic curve so as to form a voltage maximum value set V_m={v₁,v₂,…,v_mAnd a set of voltage minima U_n={u₁,u₂,…,u_n}; second from the set of voltage maxima V_mDirectly acquiring the voltage maximum value with the maximum value, and defining the voltage maximum value as the first voltage v corresponding to the power global maximum value_mpp1。

The data searching module 204 is configured to perform local search on the voltage maximum value set by using a linear search method, acquire all local even-pair data in the voltage maximum value set, and form a first even-pair set;

the implementation process of the invention is as follows: set of V values based on the voltage maxima_mThe method comprises the steps of obtaining an ith local optimal track where the ith (i is more than or equal to 1 and less than or equal to m) voltage maximum value in m voltage maximum values is located by utilizing a linear search method, and counting the ith group of local even-pair data passing through the ith local optimal track; and sequentially and circularly executing m times to obtain m groups of local even pair data to form a first even pair set EV.

More specifically, first, a first voltage maximum v is obtained by a linear search method₁The local optimal track is v_1,0→v_1,1→v_1,2→···→v_1,k1And measuring k1 power values corresponding to k1 local voltage points contained in the local optimal trajectory, and respectively tabulatingIs shown as p_1,0→p_1,1→p_1,2→···→p_1,k1Therefore, the set of local even pair data passing through the local optimal track can be counted as { v_1,0，p_1,0}→{v_1,1，p_1,1}→{v_1,2，p_1,2}→···→{v_1,k1，p_1,k1}; sequentially carrying out local search and statistics on the rest (m-1) voltage maximum values according to the method to obtain a first even pair set EV = { { v { (v) }_1,0，p_1,0},…,{v_1,k1，p_1,k1},…,{v_m,0，p_m,0},…,{v_m,km，p_m,km}}。

A voltage determining module 205, configured to extract a global optimal power from the first even-pair set, obtain a second voltage corresponding to the global optimal power, and determine whether the second voltage is equal to the first voltage;

the implementation process of the invention is as follows: firstly, obtaining a maximum power value p from the first even pair set EV_mppAnd defined as a global optimum power, and according to the power value p_mppDirectly acquiring the corresponding second voltage v by the local even-pair data_mpp2(ii) a Secondly, judging the second voltage v_mpp2Is equal to the first voltage v_mpp1The judgment result comprises: if v is_mpp2=v_mpp1Then the parameter update module 206 continues to run; if v is_mpp2≠v_mpp1Then the run model training module 207 is skipped.

A parameter updating module 206, configured to update a voltage parameter corresponding to the current maximum output power of the photovoltaic module to the second voltage after determining that the second voltage is equal to the first voltage, and update the current maximum output power to the global optimal power;

the implementation process of the invention is as follows: according to the parameter value mentioned by the voltage component obtaining module 201, the current maximum output power P is obtained_MPP1Corresponding voltage parameter V_MPP0Updated to the second voltage v_mpp2I.e. to say that the next moment will be at a fixed potentialPressure parameter v_mpp2As a new tracking point, while simultaneously setting the current maximum output power P_MPP1Updating to the global optimum power p_mppSo as to realize the maximum power tracking of the photovoltaic assembly under the new environment.

And the model training module 207 is configured to train and update the original radial basis function neural network model by using the voltage maximum value set and the voltage minimum value set after determining that the second voltage is not equal to the first voltage.

The implementation process of the invention comprises the following steps:

(1) based on the set of voltage minima U_nThe method comprises the steps that n voltage minimum values are included, and j-th even-pair data which j (j is more than or equal to 1 and less than or equal to n) th voltage minimum values in the n voltage minimum values belong to are obtained; sequentially and circularly executing n times to obtain n even pair data to form a second even pair set;

specifically, first, a first voltage minimum value u may be obtained according to the estimated PV characteristic curve₁Corresponding power value p_u1Thereby obtaining the voltage minimum u₁Associated pair data { u }₁，p_u1}; sequentially carrying out data statistics on the remaining (n-1) voltage minimum values according to the method, and obtaining a second even pair set EU = { { u { (u) }₁，p_u1},{u₂，p_u2},…,{u_n，p_un}}。

(2) And importing m groups of local even data contained in the first even set EV and n even data contained in the second even set EU into the original radial basis function neural network model, training and updating the original radial basis function neural network model based on a recursive least square method, and returning to re-execute the step S104, wherein the operated original radial basis function neural network model is in an updated state.

The training and updating of the original radial basis function neural network model adopts the following recursive least square operation:

P(n)=P(n-1)-[P(n-1)K(n)K^T(n)P(n-1)]/[1+K^T(n)P(n-1)K(n)]

W(n)=W(n-1)+P(n)K(n)[d(n)-W^T(n-1)K(n)

in the formula: n is the number of iterations, d (n) is a data set used in the nth iteration, and is composed of the first even pair set EV and the second even pair set EU, P (n-1) is a variable parameter in the (n-1) th iteration, P (n) is a variable parameter in the nth iteration, and an initial value P (0) is a diagonal matrix, k (n) is a regression coefficient in the nth iteration under the condition of linear regression, W (n-1) is a weight parameter from the hidden layer to the output layer in the (n-1) th iteration, W (n) is a weight parameter from the hidden layer to the output layer in the nth iteration, and T is a transposed symbol.

In the embodiment of the invention, the description of the evaluation PV characteristic curve by the radial basis function neural network model is updated in real time by adopting an iterative correction mode, meanwhile, any local optimal point of the evaluation PV characteristic curve can be rapidly inquired by combining a traditional linear search method, the maximum output power of the photovoltaic module is accurately tracked by utilizing a global comparison mode, and the method is suitable for an application scene of the photovoltaic module with dynamic environment change, so that the fine management of a photovoltaic power station is realized, the operation and maintenance cost is reduced, and the overall power generation efficiency is improved.

The computer-readable storage medium stores an executable computer program, and when the program is executed by a processor, the method for tracking the maximum power of the photovoltaic module according to the embodiment of the present invention is implemented. The computer-readable storage medium includes, but is not limited to, any type of disk including floppy disks, hard disks, optical disks, CD-ROMs, and magneto-optical disks, ROMs (Read-Only memories), RAMs (Random AcceSS memories), EPROMs (EraSable Programmable Read-Only memories), EEPROMs (Electrically EraSable Programmable Read-Only memories), flash memories, magnetic cards, or optical cards. That is, a storage device includes any medium that stores or transmits information in a form readable by a device (e.g., a computer, a mobile phone, etc.), and may be a read-only memory, a magnetic or optical disk, or the like.

The method, the system and the storage medium for tracking the maximum power of the photovoltaic module provided by the embodiment of the invention are described in detail, a specific example is adopted in the description to explain the principle and the implementation mode of the invention, and the description of the embodiment is only used for helping to understand the method and the core idea of the invention; meanwhile, for a person skilled in the art, according to the idea of the present invention, there may be variations in the specific embodiments and the application scope, and in summary, the content of the present specification should not be construed as a limitation to the present invention.