Photovoltaic power generation layout optimization analysis method and system based on artificial intelligence

Abstract

The invention discloses a photovoltaic power generation layout optimization analysis method and system based on artificial intelligence, and belongs to the technical field of photovoltaic layout optimization. According to the system, a prediction model training module trains a prediction model through a gradient descent algorithm according to historical local shadow shielding data to obtain a dynamic shadow prediction model; the local shadow prediction module obtains predicted local shadow shielding data through a dynamic shadow prediction model according to the current local shadow shielding data; the irradiance matrix construction module constructs an irradiance matrix according to the irradiance of the photovoltaic array assembly; the photovoltaic array layout optimization module is used for reconstructing the photovoltaic array through a layout optimization model according to the size of the photovoltaic array, the irradiance matrix and the predicted local shadow shielding data to obtain a photovoltaic array reconstruction optimization matrix; and the photovoltaic array reconstruction optimization module is used for the switch matrix to reconstruct the photovoltaic array according to the photovoltaic array reconstruction optimization matrix. According to the invention, photovoltaic power generation layout optimization based on dynamic shadow analysis is realized.

Description

Technical Field

[0001] The present invention belongs to the technical field of photovoltaic layout optimization, and specifically relates to an artificial intelligence-based photovoltaic power generation layout optimization analysis method and system. Background Art

[0002] As land resources with flat and open terrain and adequate sunshine become increasingly scarce, idle land such as mountains and hills has gradually become a new choice for the construction of photovoltaic power stations. However, due to the limitations of terrain and topography, mountain photovoltaic power stations are more prone to local shadowing, resulting in hot spot failures inside the photovoltaic array. Therefore, adopting a dynamic shadow analysis method and optimizing the layout through photovoltaic array reconstruction can effectively reduce the impact of local shadows and reduce the probability of hot spot failures, thereby improving the stability and efficiency of photovoltaic power generation systems.

[0003] Traditional static reconstruction methods change the physical position of components in the array according to the preset local shadow conditions to achieve dispersed shadows and avoid hot spot effects. However, static reconstruction methods are one-time physical reconstructions and have limitations in the face of complex shadow environments. Existing meta-heuristic dynamic reconstruction methods based on ant colony optimization, gray wolf optimization, particle swarm optimization, and mayfly optimization can collect array information through intelligent monitoring devices, solve the photovoltaic array configuration with the best power output through intelligent algorithms, and drive the switch matrix to change the electrical connection method to optimize the photovoltaic array layout. However, there are still problems such as slow convergence speed, easy to fall into local optimality, and large amount of calculation. Summary of the invention

[0004] In order to solve the above problems existing in the prior art, the present invention provides a photovoltaic power generation layout optimization analysis method and system based on artificial intelligence.

[0005] In order to achieve the above object, the present invention adopts the following technical solution: The photovoltaic power generation layout optimization analysis method based on artificial intelligence includes: Acquire historical local shadow occlusion data, and train a prediction model using a gradient descent algorithm based on the historical local shadow occlusion data to obtain a dynamic shadow prediction model; Acquire current local shadow occlusion data, and obtain predicted local shadow occlusion data through the dynamic shadow prediction model according to the current local shadow occlusion data; Obtain the size of the photovoltaic array, collect the irradiance of the photovoltaic array components through a radiometer, and construct an irradiance matrix according to the irradiance of the photovoltaic array components; Reconstructing the photovoltaic array through a layout optimization model according to the size of the photovoltaic array, the irradiance matrix, and the predicted local shadow occlusion data to obtain a photovoltaic array reconstruction optimization matrix; The switch matrix reconstructs the photovoltaic array according to the photovoltaic array reconstruction optimization matrix.

[0006] A further improvement of the present invention is that the dynamic shadow prediction model is obtained by training the prediction model through a gradient descent algorithm according to the historical local shadow occlusion data, comprising: S101: Fill missing values ​​by linear interpolation according to the historical local shadow occlusion data to obtain historical local shadow occlusion complete data; preset an initial prediction model, the initial prediction model includes a CNN layer, an LSTM layer, and a fully connected layer; S102: performing feature extraction through the CNN layer according to the historical local shadow occlusion completion data to obtain historical local shadow occlusion features; S103: performing time series modeling through the LSTM layer according to the historical local shadow occlusion features to obtain historical local shadow occlusion time series features; S104: Obtaining a historical local shadow occlusion prediction value through the fully connected layer according to the historical local shadow occlusion temporal features; S105: Obtaining a historical local shadow occlusion true value, and calculating a historical local shadow occlusion loss through a mean square error (MSE) loss function according to the historical local shadow occlusion predicted value and the historical local shadow occlusion true value; S106: Obtain model parameters, and obtain model parameter gradients by derivation calculation according to the historical local shadow occlusion loss and the model parameters; S107: updating the parameters of the initial prediction model by using the Adam gradient descent algorithm according to the model parameter gradient to obtain a training dynamic shadow prediction model, wherein the training dynamic shadow prediction model carries a training times label; S108: Preset the number of iterations, and obtain the dynamic shadow prediction model according to the training number label and the number of iterations; The iteration number judgment includes judging whether the training number label and the iteration number are equal, if yes, obtaining the dynamic shadow prediction model; if no, repeating steps S101 to S107 until the training number label and the iteration number are equal.

[0007] A further improvement of the present invention is that the photovoltaic array is reconstructed through a layout optimization model according to the size of the photovoltaic array, the irradiance matrix, and the predicted local shadow occlusion data to obtain a photovoltaic array reconstruction optimization matrix, which includes: Obtaining a local shadow row current and a local shadow row voltage by current and voltage calculation according to the irradiance matrix and the predicted local shadow shading data; According to the local shadow row currents, a maximum row current and a minimum row current are obtained by sorting; Define an objective function according to the maximum row current, the minimum row current, the local shadow row voltage, and the local shadow row current; The photovoltaic array reconstruction optimization matrix is ​​obtained by using an improved vulture optimization algorithm according to the objective function, the photovoltaic array size, the irradiance matrix, and the predicted local shadow shading data.

[0008] A further improvement of the present invention is that the photovoltaic array reconstruction optimization matrix is ​​obtained by improving the vulture optimization algorithm according to the objective function, the photovoltaic array size, the irradiance matrix, and the predicted local shadow shielding data, including: Obtaining the initialization vulture position according to the irradiance matrix through tent chaotic mapping; Acquire initial fitness, and obtain optimal vultures and suboptimal vultures by sorting according to the initial fitness; Obtaining irradiance fitness through fitness calculation according to the local shadow row voltage and the local shadow row current; According to the optimal vulture and the suboptimal vulture, the distance between the current vulture and the optimal and suboptimal vultures is obtained through an individual memorization search strategy; The photovoltaic array reconstruction optimization matrix is ​​obtained through stage judgment according to the irradiance adaptability.

[0009] A further improvement of the present invention is that the stage judgment includes judging whether the irradiance fitness is greater than or equal to 1. If so, it is in the exploration stage; if not, judging whether the irradiance fitness is greater than 0.5. If so, it is in the first development stage; if not, it is in the late development stage.

[0010] A further improvement of the present invention is that the exploration phase includes obtaining upper and lower boundaries for optimization, the upper and lower boundaries for optimization represent the boundaries of the photovoltaic array, and a new vulture position is obtained through an exploration strategy according to the vulture position, the distance between the current vulture and the optimal and suboptimal vultures, and the upper and lower boundaries for optimization.

[0011] A further improvement of the present invention is that the first stage of development includes obtaining a number of behaviors and a number of behavior determinations in the first stage of development, and obtaining a new vulture position in the first stage of development by determining the behaviors in the first stage of development according to the number of behaviors in the first stage of development and the number of behavior determinations; The behavior judgment of the first stage of development includes judging whether the number of behaviors in the first stage of development is greater than the behavior determination number. If yes, the current vulture position is obtained, and the vulture obtains the new vulture position in the first stage of development by performing circling flight behavior according to the distance between the current vulture and the optimal and suboptimal vultures and the current vulture position; if no, the vulture obtains the new vulture position in the first stage of development by performing siege behavior according to the distance between the current vulture and the optimal and suboptimal vultures and the current vulture position.

[0012] A further improvement of the present invention is that the late stage of development includes obtaining a late stage behavior number and a late stage behavior determination number, and obtaining a new vulture position in the late stage of development by judging the late stage behavior according to the late stage behavior number and the late stage behavior determination number; The late behavior judgment of the development stage includes judging whether the number of late behaviors in the development stage is greater than the late behavior determination number. If yes, the early vulture position is obtained, and the vultures obtain the new vulture position in the late development stage by performing agglomeration behavior according to the distance between the current vulture and the optimal and suboptimal vultures and the early vulture position; if no, the vultures obtain the new vulture position in the late development stage by performing an attack behavior according to the distance between the current vulture and the optimal and suboptimal vultures and the early vulture position.

[0013] Photovoltaic power generation layout optimization and analysis system based on artificial intelligence, including prediction model training module, local shadow prediction module, irradiance matrix construction module, photovoltaic array layout optimization module and photovoltaic array reconstruction optimization module; The prediction model training module is used to obtain historical local shadow occlusion data, and train the prediction model through a gradient descent algorithm according to the historical local shadow occlusion data to obtain a dynamic shadow prediction model; The local shadow prediction module is used to obtain current local shadow occlusion data, and obtain predicted local shadow occlusion data through the dynamic shadow prediction model according to the current local shadow occlusion data; The irradiance matrix construction module is used to obtain the size of the photovoltaic array, collect the irradiance of the photovoltaic array components through a radiometer, and construct an irradiance matrix according to the irradiance of the photovoltaic array components; The photovoltaic array layout optimization module is used to reconstruct the photovoltaic array through a layout optimization model according to the size of the photovoltaic array, the irradiance matrix, and the predicted local shadow occlusion data to obtain a photovoltaic array reconstruction optimization matrix; The photovoltaic array reconstruction optimization module is used for the switch matrix to reconstruct the photovoltaic array according to the photovoltaic array reconstruction optimization matrix.

[0014] A storage medium containing computer executable instructions, which are used to execute the steps of the photovoltaic power generation layout optimization analysis method based on artificial intelligence when executed by a computer processor.

[0015] Compared with the prior art, the present invention has at least the following beneficial technical effects: The photovoltaic power generation layout optimization analysis method and system based on artificial intelligence provided by the present invention predicts the local shadow occlusion through the trained LSTM model, improves the accuracy of the subsequent photovoltaic array reconstruction, enables the photovoltaic array to dynamically adjust the layout according to the prediction results, minimizes the impact of shadows on power generation efficiency, and reduces the risk of hot spot failure. By improving the vulture optimization algorithm, simulating the foraging behavior and habits of vultures, and adjusting the irradiance matrix of the photovoltaic array through the three stages of exploration stage, first development stage and late development stage, the system can be flexibly adjusted under different lighting, shadow and terrain conditions, has strong environmental adaptability, and improves power generation efficiency. The vulture position is initialized by the Tent chaotic mapping, so that the layout optimization model has better traversal and randomness, and avoids falling into the local optimum; in the process of obtaining the distance between the current vulture and the optimal and suboptimal vultures through the individual memorization search strategy, in addition to referring to the optimal vulture and suboptimal vulture positions, the historical optimal solution of each vulture is also recorded, which improves the accuracy of position update and the accuracy of irradiance matrix optimization, and enhances the optimization ability of photovoltaic power generation layout. By dynamically adjusting the connection mode of photovoltaic array components according to the optimal irradiance matrix through the switch matrix, the photovoltaic power generation layout can be optimized according to dynamic shadow analysis to maximize the light absorption efficiency of the photovoltaic array. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0017] Figure 1 It is a schematic diagram of the flow of the photovoltaic power generation layout optimization analysis method based on artificial intelligence of the present invention.

[0018] Figure 2 This is a structural block diagram of the photovoltaic power generation layout optimization analysis system based on artificial intelligence of the present invention. DETAILED DESCRIPTION

[0019] In the following, only some exemplary embodiments are briefly described. As those skilled in the art will appreciate, the described embodiments may be modified in various ways without departing from the spirit or scope of the present invention. Therefore, the drawings and descriptions are considered to be exemplary and non-restrictive in nature.

[0020] It should be understood that when used in this specification and the appended claims, the terms "include" and "comprises" indicate the presence of described features, integers, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or combinations thereof.

[0021] It should also be understood that the terms used in the present specification are only for the purpose of describing specific embodiments and are not intended to limit the present invention. As used in the present specification and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include plural forms.

[0022] It should be further understood that the term "and / or" used in the present description and the appended claims refers to and includes any and all possible combinations of one or more of the associated listed items.

[0023] Various structural schematic diagrams of the embodiments disclosed in the present invention are shown in the accompanying drawings. These figures are not drawn to scale, and some details are magnified and some details may be omitted for the purpose of clear expression. The shapes of various regions and layers shown in the figures and the relative sizes and positional relationships therebetween are only exemplary, and may deviate in practice due to manufacturing tolerances or technical limitations, and those skilled in the art may additionally design regions / layers with different shapes, sizes, and relative positions according to actual needs.

[0024] The embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0025] Example 1 like Figure 1 As shown, the photovoltaic power generation layout optimization analysis method based on artificial intelligence provided by the present invention includes: Acquire historical local shadow occlusion data, and train a prediction model using a gradient descent algorithm based on the historical local shadow occlusion data to obtain a dynamic shadow prediction model; Acquire current local shadow occlusion data, and obtain predicted local shadow occlusion data through the dynamic shadow prediction model according to the current local shadow occlusion data; Obtain the size of the photovoltaic array, collect the irradiance of the photovoltaic array components through a radiometer, and construct an irradiance matrix according to the irradiance of the photovoltaic array components; Reconstructing the photovoltaic array through a layout optimization model according to the size of the photovoltaic array, the irradiance matrix, and the predicted local shadow occlusion data to obtain a photovoltaic array reconstruction optimization matrix; The switch matrix reconstructs the photovoltaic array according to the photovoltaic array reconstruction optimization matrix.

[0026] Specifically, the gradient descent algorithm training prediction model includes: S101: Fill missing values ​​by linear interpolation according to the historical local shadow occlusion data to obtain historical local shadow occlusion complete data; preset an initial prediction model, the initial prediction model includes a CNN layer, an LSTM layer, and a fully connected layer; S102: performing feature extraction through the CNN layer according to the historical local shadow occlusion completion data to obtain historical local shadow occlusion features; S103: performing time series modeling through the LSTM layer according to the historical local shadow occlusion features to obtain historical local shadow occlusion time series features; S104: Obtaining a historical local shadow occlusion prediction value through the fully connected layer according to the historical local shadow occlusion temporal features; S105: Obtaining a historical local shadow occlusion true value, and calculating a historical local shadow occlusion loss through a mean square error (MSE) loss function according to the historical local shadow occlusion predicted value and the historical local shadow occlusion true value; S106: Obtain model parameters, and obtain model parameter gradients by derivation calculation according to the historical local shadow occlusion loss and the model parameters; S107: updating the parameters of the initial prediction model by using the Adam gradient descent algorithm according to the model parameter gradient to obtain a training dynamic shadow prediction model, wherein the training dynamic shadow prediction model carries a training times label; S108: Preset the number of iterations, and obtain the dynamic shadow prediction model according to the training number label and the number of iterations; The iteration number judgment includes judging whether the training number label and the iteration number are equal, if yes, obtaining the dynamic shadow prediction model; if no, repeating steps S101 to S107 until the training number label and the iteration number are equal.

[0027] Specifically, the irradiance matrix represents the distribution of light intensity received by each point on the photovoltaic array in different directions.

[0028] Specifically, the layout optimization model includes: S401: Obtaining a local shadow row current and a local shadow row voltage through current and voltage calculation according to the irradiance matrix and the predicted local shadow shading data; The current and voltage calculation expression is:

[0029] Among them, I Ri represents the local shadow row current, j represents the photovoltaic module in the jth column, n represents the number of photovoltaic modules, G ij Indicates irradiance, I m is the maximum operating current, V a represents the partial shadow row voltage, i represents the photovoltaic module in the i-th row, b is the number of rows in the photovoltaic array that are bypassed when partial shadow is generated, V mi is the row voltage of the i-th row in the photovoltaic array; S402: Obtaining a maximum row current and a minimum row current by sorting the local shadow row currents; S403: defining an objective function according to the maximum row current, the minimum row current, the local shadow row voltage, and the local shadow row current; The objective function is expressed as:

[0030] Wherein, max(fitness(i)) represents the objective function, which represents the photovoltaic array power obtained by maximizing the minimum deviation between the maximum row current and the minimum row current, P a is the total array power, I Rmax is the maximum line current, I Rmin is the minimum row current, m represents the current and voltage of the mth row of the photovoltaic array, X is the number of rows of the photovoltaic array, I Rm represents the local shaded row current of the mth row, V m represents the local shaded row voltage of the mth row; S404: Obtaining the photovoltaic array reconstruction optimization matrix through an improved vulture optimization algorithm according to the objective function, the photovoltaic array size, the irradiance matrix, and the predicted local shadow occlusion data.

[0031] Specifically, the improved vulture optimization algorithm includes: S404-1: Obtaining an initialized vulture position according to the irradiance matrix through tent chaotic mapping; The tent chaos mapping expression is:

[0032] Among them, x t+1represents the initialization vulture position, tent represents the tent chaos map, x t Indicates the position of the vulture before the chaotic mapping, u is 0.45; S404-2: Obtaining initial fitness, and obtaining the best vulture and the second-best vulture by sorting according to the initial fitness; S404-3: Obtaining irradiance fitness through fitness calculation according to the local shadow row voltage and the local shadow row current; The fitness calculation expression is:

[0033] in, represents the irradiance adaptation, is a random number, In order to avoid falling into the local optimal adjustment parameters, z t is a random number, t represents the tth iteration, T represents the total number of iterations, and h t is a random number, ω is a stage parameter, the stage parameter represents the exploration stage and the development stage, sin is a sine function, and cos is a cosine function; S404-4: Obtaining the distance between the current vulture and the optimal and suboptimal vultures through an individual memorization search strategy according to the optimal vulture and the suboptimal vulture; The individual memorization search strategy expression is:

[0034] in, represents the distance between the current vulture and the optimal and suboptimal vultures, , is the coefficient, for the optimal vulture and the suboptimal vulture, is the position of the ith vulture at the tth iteration, C1 and C2 are random numbers, P i is the optimal position reached by the i-th vulture in history, t represents the t-th iteration, and T represents the total number of iterations; S404-5: Obtaining the photovoltaic array reconstruction optimization matrix through stage judgment according to the irradiance adaptability; The stage judgment includes judging whether the irradiance fitness is greater than or equal to 1. If yes, it is in the exploration stage; if no, judging whether the irradiance fitness is greater than 0.5. If yes, it is in the first stage of development; if no, it is in the late stage of development.

[0035] Specifically, the exploration phase includes obtaining upper and lower boundaries for optimization, wherein the upper and lower boundaries for optimization represent boundaries of the photovoltaic array, and obtaining a new vulture position through an exploration strategy according to the vulture position, the distance between the current vulture and the optimal and suboptimal vultures, and the upper and lower boundaries for optimization; The exploration strategy is expressed as:

[0036] in, represents the position of the i-th vulture at the t+1 iteration, represents the distance between the current vulture and the optimal and suboptimal vultures, represents the irradiance adaptation, , 、rand i3 Represents a random number ranging from 0 to 1, R t (i) represents the optimal vulture and the suboptimal vulture, i represents the i-th vulture, P1 is the strategy selection probability, I b represents the lower boundary of optimization, u b Represents the upper boundary of optimization.

[0037] Specifically, the first stage of development includes: Obtaining the number of behaviors and the number of behavior determinations in the first stage of development, and obtaining the new vulture position in the first stage of development by judging the behaviors in the first stage of development according to the number of behaviors in the first stage of development and the number of behavior determinations; The behavior judgment of the first stage of development includes judging whether the number of behaviors in the first stage of development is greater than the behavior determination number. If yes, the current vulture position is obtained, and the vulture obtains the new vulture position in the first stage of development by performing circling flight behavior according to the distance between the current vulture and the optimal and suboptimal vultures and the current vulture position; if no, the vulture obtains the new vulture position in the first stage of development by performing siege behavior according to the distance between the current vulture and the optimal and suboptimal vultures and the current vulture position.

[0038] Specifically, the hovering flight behavior is expressed as:

[0039] in, Indicates the new vulture location for the first phase of the development, represents the current vulture position, R t (i) represents the optimal vulture and the suboptimal vulture, i represents the i-th vulture, S1 and S2 represent the circling position parameters, rand i5 、rand i6 is a random number ranging from 0 to 1, cos represents the cosine function, represents the distance between the current vulture and the optimal and suboptimal vultures, Pi is the optimal position that the i-th vulture has ever reached in history.

[0040] Specifically, the siege behavior is represented by:

[0041] in, Indicates the new vulture location for the first phase of the development, represents the current vulture position, is a variable, represents the distance between the current vulture and the optimal and suboptimal vultures, represents the irradiance adaptation, It is a random number ranging from 0 to 1.

[0042] Specifically, the late stage of development includes obtaining a late stage behavior number and a late stage behavior determination number, and obtaining a new vulture position in the late stage of development by late stage behavior judgment according to the late stage behavior number and the late stage behavior determination number; The late behavior judgment of the development stage includes judging whether the number of late behaviors in the development stage is greater than the late behavior determination number. If yes, the early vulture position is obtained, and the vultures obtain the new vulture position in the late development stage by performing agglomeration behavior according to the distance between the current vulture and the optimal and suboptimal vultures and the early vulture position; if no, the vultures obtain the new vulture position in the late development stage by performing an attack behavior according to the distance between the current vulture and the optimal and suboptimal vultures and the early vulture position.

[0043] Specifically, the aggregation behavior is expressed as:

[0044] in, represents the new vulture position in the later stage of the development, A1 and A2 are intermediate variables used to calculate the new vulture position, represents the optimal vulture position at time t, represents the second-best vulture position at time t, represents the previous vulture position, Represents the irradiance adaptation.

[0045] Specifically, the attack behavior is represented as follows:

[0046] in, represents the new vulture position at the later stage of the development, R t (i) represents the optimal vulture and the suboptimal vulture, is a variable, represents the irradiance fitness, Levy(d) is the Levy flight function, u and v are random numbers ranging from 0 to 1, σ is a variable, and ρ is 1.5. is the gamma function and sin is the sine function.

[0047] In this embodiment, the improved vulture algorithm uses the photovoltaic array as the upper and lower boundaries of the vulture optimization, regards the photovoltaic array component irradiance matrix as the vulture, and the photovoltaic array irradiance matrix as the population. Through the exploration stage, the first development stage, and the late development stage, the vulture position is continuously updated to obtain the optimal irradiance matrix, thereby achieving the optimal reconstruction of the photovoltaic array.

[0048] In this embodiment, the photovoltaic array is a fully cross-connected structure, and photovoltaic array information is collected through intelligent monitoring. The photovoltaic array information includes row current, row current difference, row power, and photovoltaic array component irradiance. The local shadow shading situation is predicted based on LSTM to obtain the predicted local shadow shading situation. According to the predicted local shadow shading situation, the photovoltaic array is dynamically reconstructed by improving the vulture optimization algorithm to obtain a photovoltaic array reconstruction optimization matrix. The switching matrix is ​​driven to change the electrical connection method according to the photovoltaic array reconstruction optimization matrix. When the array is running, the photovoltaic array can be dynamically reconstructed according to the changes in local shadows, thereby dispersing local shadows.

[0049] Example 2 like Figure 2 As shown, the photovoltaic power generation layout optimization and analysis system based on artificial intelligence provided by the present invention includes a prediction model training module, a local shadow prediction module, an irradiance matrix construction module, a photovoltaic array layout optimization module and a photovoltaic array reconstruction optimization module; The prediction model training module is used to obtain historical local shadow occlusion data, and train the prediction model through a gradient descent algorithm according to the historical local shadow occlusion data to obtain a dynamic shadow prediction model; The local shadow prediction module is used to obtain current local shadow occlusion data, and obtain predicted local shadow occlusion data through the dynamic shadow prediction model according to the current local shadow occlusion data; The irradiance matrix construction module is used to obtain the size of the photovoltaic array, collect the irradiance of the photovoltaic array components through a radiometer, and construct an irradiance matrix according to the irradiance of the photovoltaic array components; The photovoltaic array layout optimization module is used to reconstruct the photovoltaic array through a layout optimization model according to the size of the photovoltaic array, the irradiance matrix, and the predicted local shadow occlusion data to obtain a photovoltaic array reconstruction optimization matrix; The photovoltaic array reconstruction optimization module is used for the switch matrix to reconstruct the photovoltaic array according to the photovoltaic array reconstruction optimization matrix.

[0050] Example 3 The present invention provides a storage medium containing computer executable instructions, which, when executed by a computer processor, are used to execute the steps of the photovoltaic power generation layout optimization analysis method based on artificial intelligence.

[0051] The computer storage medium of the present embodiment can adopt any combination of one or more computer-readable media. Computer-readable media can be computer-readable signal media or computer-readable storage media. Computer-readable storage media can be, for example, but not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or devices, or any combination of the above. More specific examples (non-exhaustive list) of computer-readable storage media include: electrical connections with one or more wires, portable computer disks, hard disks, random access memories (RAM), read-only memories (ROM), erasable programmable read-only memories (EPROM or flash memory), optical fibers, portable compact disk read-only memories (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the above. In this document, computer-readable storage media can be any tangible medium containing or storing a program, which can be used by an instruction execution system, device or device or used in combination with it.

[0052] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, which carry computer-readable program code. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer-readable signal media may also be any computer-readable medium other than a computer-readable storage medium, which may send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0053] The program code included on the computer readable medium can be transmitted with any appropriate medium, including but not limited to wireless, electric wire, optical cable, RF, etc., or any suitable combination of the above. The computer program code for performing the operation of the present invention can be written in one or more programming languages ​​or their combinations, and the programming language includes object-oriented programming languages-such as Java, Smalltalk, C++, and also includes conventional procedural programming languages-such as "C" language or similar programming languages. The program code can be executed completely on the user's computer, partially on the user's computer, as an independent software package, partially on the user's computer and partially on the remote computer, or completely on the remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computer (for example, using an Internet service provider to connect through the Internet).

[0054] The above shows and describes the basic principles and main features of the present invention and the advantages of the present invention. It is obvious to those skilled in the art that the present invention is not limited to the details of the above exemplary embodiments, and the present invention can be implemented in other specific forms without departing from the spirit or basic features of the present invention. Therefore, no matter from which point of view, the embodiments should be regarded as exemplary and non-restrictive. The scope of the present invention is defined by the attached claims rather than the above description, and it is intended that all changes falling within the meaning and scope of the equivalent elements of the claims are included in the present invention. Any figure mark in the claims should not be regarded as limiting the claims involved.

[0055] In addition, it should be understood that although this specification is described in accordance with the implementation modes, not every implementation mode contains only one independent technical solution. This description of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment can also be appropriately combined to form other implementation modes that can be understood by those skilled in the art. The above content is only to illustrate the technical idea of ​​the present invention, and cannot be used to limit the protection scope of the present invention. Any changes made on the basis of the technical solution according to the technical idea proposed by the present invention shall fall within the protection scope of the claims of the present invention.

Claims

1. Photovoltaic power generation layout optimization analysis method based on artificial intelligence, characterized in that: include: Acquire historical local shadow occlusion data, and train a prediction model using a gradient descent algorithm based on the historical local shadow occlusion data to obtain a dynamic shadow prediction model; Acquire current local shadow occlusion data, and obtain predicted local shadow occlusion data through the dynamic shadow prediction model according to the current local shadow occlusion data; Obtain the size of the photovoltaic array, collect the irradiance of the photovoltaic array components through a radiometer, and construct an irradiance matrix according to the irradiance of the photovoltaic array components; Reconstructing the photovoltaic array through a layout optimization model according to the size of the photovoltaic array, the irradiance matrix, and the predicted local shadow occlusion data to obtain a photovoltaic array reconstruction optimization matrix; The switch matrix reconstructs the photovoltaic array according to the photovoltaic array reconstruction optimization matrix.

2. The photovoltaic power generation layout optimization analysis method based on artificial intelligence according to claim 1 is characterized in that: The step of obtaining a dynamic shadow prediction model by training a prediction model through a gradient descent algorithm according to the historical local shadow occlusion data comprises: S101: Fill missing values ​​by linear interpolation according to the historical local shadow occlusion data to obtain historical local shadow occlusion complete data; preset an initial prediction model, the initial prediction model includes a CNN layer, an LSTM layer, and a fully connected layer; S102: performing feature extraction through the CNN layer according to the historical local shadow occlusion completion data to obtain historical local shadow occlusion features; S103: performing time series modeling through the LSTM layer according to the historical local shadow occlusion features to obtain historical local shadow occlusion time series features; S104: Obtaining a historical local shadow occlusion prediction value through the fully connected layer according to the historical local shadow occlusion temporal features; S105: Obtaining a historical local shadow occlusion true value, and calculating a historical local shadow occlusion loss through a mean square error (MSE) loss function according to the historical local shadow occlusion predicted value and the historical local shadow occlusion true value; S106: Obtain model parameters, and obtain model parameter gradients by derivation calculation according to the historical local shadow occlusion loss and the model parameters; S107: updating the parameters of the initial prediction model by using the Adam gradient descent algorithm according to the model parameter gradient to obtain a training dynamic shadow prediction model, wherein the training dynamic shadow prediction model carries a training times label; S108: Preset the number of iterations, and obtain the dynamic shadow prediction model according to the training number label and the number of iterations; The iteration number judgment includes judging whether the training number label and the iteration number are equal, if yes, obtaining the dynamic shadow prediction model; if no, repeating steps S101 to S107 until the training number label and the iteration number are equal.

3. The photovoltaic power generation layout optimization analysis method based on artificial intelligence according to claim 1 is characterized in that: The reconstructing the photovoltaic array through the layout optimization model according to the photovoltaic array size, the irradiance matrix, and the predicted local shadow shielding data to obtain the photovoltaic array reconstruction optimization matrix includes: Obtaining a local shadow row current and a local shadow row voltage by current and voltage calculation according to the irradiance matrix and the predicted local shadow shading data; According to the local shadow row currents, a maximum row current and a minimum row current are obtained by sorting; Define an objective function according to the maximum row current, the minimum row current, the local shadow row voltage, and the local shadow row current; The photovoltaic array reconstruction optimization matrix is ​​obtained by using an improved vulture optimization algorithm according to the objective function, the photovoltaic array size, the irradiance matrix, and the predicted local shadow shading data.

4. The photovoltaic power generation layout optimization analysis method based on artificial intelligence according to claim 3 is characterized in that: The step of obtaining the photovoltaic array reconstruction optimization matrix by using an improved vulture optimization algorithm according to the objective function, the photovoltaic array size, the irradiance matrix, and the predicted local shadow shielding data comprises: Obtaining the initialization vulture position according to the irradiance matrix through tent chaotic mapping; Acquire initial fitness, and obtain optimal vultures and suboptimal vultures by sorting according to the initial fitness; Obtaining irradiance fitness through fitness calculation according to the local shadow row voltage and the local shadow row current; According to the optimal vulture and the suboptimal vulture, the distance between the current vulture and the optimal and suboptimal vultures is obtained through an individual memorization search strategy; The photovoltaic array reconstruction optimization matrix is ​​obtained through stage judgment according to the irradiance adaptability.

5. The photovoltaic power generation layout optimization analysis method based on artificial intelligence according to claim 4 is characterized in that: The stage judgment includes judging whether the irradiance fitness is greater than or equal to 1. If yes, it is in the exploration stage; if no, judging whether the irradiance fitness is greater than 0.

5. If yes, it is in the first stage of development; if no, it is in the late stage of development.

6. The photovoltaic power generation layout optimization analysis method based on artificial intelligence according to claim 5 is characterized in that: The exploration phase includes obtaining the upper and lower boundaries of the optimization, which represent the boundaries of the photovoltaic array, and obtaining a new vulture position through an exploration strategy according to the vulture position, the distance between the current vulture and the optimal and suboptimal vultures, and the upper and lower boundaries of the optimization.

7. The photovoltaic power generation layout optimization analysis method based on artificial intelligence according to claim 5 is characterized in that: The first stage of development includes obtaining the number of behaviors and the number of behavior determinations in the first stage of development, and obtaining the new vulture position in the first stage of development by judging the behavior in the first stage of development according to the number of behaviors in the first stage of development and the number of behavior determinations; The behavior judgment of the first stage of development includes judging whether the number of behaviors in the first stage of development is greater than the behavior determination number. If yes, the current vulture position is obtained, and the vulture obtains the new vulture position in the first stage of development by performing circling flight behavior according to the distance between the current vulture and the optimal and suboptimal vultures and the current vulture position; if no, the vulture obtains the new vulture position in the first stage of development by performing siege behavior according to the distance between the current vulture and the optimal and suboptimal vultures and the current vulture position.

8. The photovoltaic power generation layout optimization analysis method based on artificial intelligence according to claim 5 is characterized in that: The late stage of development includes obtaining a late stage behavior number and a late stage behavior determination number, and obtaining a new vulture position in the late stage of development by late stage behavior judgment according to the late stage behavior number and the late stage behavior determination number; The late behavior judgment of the development stage includes judging whether the number of late behaviors in the development stage is greater than the late behavior determination number. If yes, the early vulture position is obtained, and the vultures obtain the new vulture position in the late development stage by performing agglomeration behavior according to the distance between the current vulture and the optimal and suboptimal vultures and the early vulture position; if no, the vultures obtain the new vulture position in the late development stage by performing an attack behavior according to the distance between the current vulture and the optimal and suboptimal vultures and the early vulture position.

9. Photovoltaic power generation layout optimization analysis system based on artificial intelligence, characterized by: It includes prediction model training module, local shadow prediction module, irradiance matrix construction module, photovoltaic array layout optimization module and photovoltaic array reconstruction optimization module; The prediction model training module is used to obtain historical local shadow occlusion data, and train the prediction model through a gradient descent algorithm according to the historical local shadow occlusion data to obtain a dynamic shadow prediction model; The local shadow prediction module is used to obtain current local shadow occlusion data, and obtain predicted local shadow occlusion data through the dynamic shadow prediction model according to the current local shadow occlusion data; The irradiance matrix construction module is used to obtain the size of the photovoltaic array, collect the irradiance of the photovoltaic array components through a radiometer, and construct an irradiance matrix according to the irradiance of the photovoltaic array components; The photovoltaic array layout optimization module is used to reconstruct the photovoltaic array through a layout optimization model according to the size of the photovoltaic array, the irradiance matrix, and the predicted local shadow occlusion data to obtain a photovoltaic array reconstruction optimization matrix; The photovoltaic array reconstruction optimization module is used for the switch matrix to reconstruct the photovoltaic array according to the photovoltaic array reconstruction optimization matrix.

10. A storage medium containing computer executable instructions, characterized in that: The computer executable instructions are used to execute the steps of the photovoltaic power generation layout optimization analysis method based on artificial intelligence as described in any one of claims 1 to 8 when executed by a computer processor.

Images