A photovoltaic module shading optimization system and inverse tracking method

Abstract

The present application relates to photovoltaic tracking control technical field, specifically to a kind of photovoltaic module shelter optimization system and inverse tracking method, comprising: obtaining the local state vector of photovoltaic tracking unit;Receiving the neighborhood state vector of each neighbor unit in the neighborhood of the photovoltaic tracking unit;Calculate the reference angle of the photovoltaic tracking unit;Generate discrete alternative perturbation angle set;Estimate neighborhood influence estimation;Generate synergistic gain;Select the alternative perturbation angle that makes the synergistic gain maximum as optimal perturbation angle;Determine target execution angle in combination with the reference angle and the optimal perturbation angle;According to the target execution angle, output control instruction to execute angle adjustment.Solve the power generation efficiency loss problem caused by traditional inverse tracking algorithm due to dependence on idealized static model.

Description

Technical Field

[0001] This invention relates to the field of photovoltaic tracking control technology, specifically to a photovoltaic module shading optimization system and inverse tracking method. Background Technology

[0002] With the rapid development of the photovoltaic industry, large-scale photovoltaic power plants have become an important energy source. Improving the overall power generation efficiency of photovoltaic arrays is a core technological goal in this field. In densely deployed photovoltaic power plants, mutual shading between photovoltaic modules is a key factor leading to power generation loss. To address this issue, traditional inverse tracking algorithms are widely used; however, such algorithms heavily rely on idealized static geometric models that presuppose flat terrain and unbiased installation. In actual engineering, due to non-ideal factors such as construction errors, terrain undulations, and equipment structural deformation, the deviation between this static model and the physical world causes the algorithm to fail to accurately avoid shading, resulting in significant power generation efficiency losses.

[0003] The limitations of existing methods lie in their open-loop control mode, which lacks the ability to perceive and dynamically compensate for deviations in the real world in real time; the system cannot make adaptive adjustments based on the actual state of each photovoltaic tracking unit, such as the local shading coverage; in addition, traditional algorithms treat each tracking unit as an isolated individual, and its decision-making only aims to maximize its own solar energy capture, ignoring the negative shading effects that adjustment actions may cause to neighboring units; this decision-making mode without a cooperative mechanism leads to the accumulation of local optima, which ultimately damages the overall power generation performance of the power plant.

[0004] In summary, existing technologies lack an intelligent system capable of constructing photovoltaic tracking units with real-time perception, local communication, and collaborative decision-making capabilities. They are unable to dynamically compensate for physical world deviations through distributed computing and quantitatively assess the externalities of decision-making behavior on the neighborhood, thereby maximizing the actual total power generation of the entire photovoltaic array in complex real-world environments. Therefore, designing an optimization control method that can compensate for physical world deviations in real time through distributed collaborative decision-making and aims to maximize the overall power generation of the array is a technical problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0005] To address the aforementioned technical problems, this invention discloses a photovoltaic module shading optimization system and an inverse tracking method. Specifically, the technical solution of this invention is as follows:

[0006] A method for optimizing shading of photovoltaic modules includes:

[0007] Obtain the local state vector of the photovoltaic tracking unit;

[0008] Receive the neighborhood state vectors of each neighboring unit within the neighborhood of the photovoltaic tracking unit;

[0009] The local state vector includes the real-time rotation angle and shadow coverage of the photovoltaic tracking unit;

[0010] The reference angle of the photovoltaic tracking unit is calculated based on the solar position algorithm;

[0011] A discrete set of alternative perturbation angles is generated around the reference angle;

[0012] For any candidate perturbation angle in the set of candidate perturbation angles, calculate the local gain prediction and estimate the neighborhood impact.

[0013] By combining the local gain prediction with the neighborhood influence estimation, a cooperative gain is generated;

[0014] Select the candidate perturbation angle that maximizes the cooperative gain as the optimal perturbation angle;

[0015] By combining the reference angle and the optimal disturbance angle, the target execution angle is determined;

[0016] Based on the target execution angle, output control commands to adjust the execution angle.

[0017] Preferably, the calculation method for the neighborhood influence estimation includes:

[0018] Obtain the effective irradiance of the photovoltaic tracking unit;

[0019] Calculate the change in shadow area projected onto the neighboring cell due to the alternative perturbation angle;

[0020] By combining the effective irradiance and the change in shadow area, and based on the preset photoelectric conversion efficiency, the estimated power change of the neighboring unit is calculated.

[0021] For each neighboring unit in the neighborhood, the power change estimate is weighted based on a dynamically adaptive collaborative weighting factor, and all weighted results are summed to generate the neighborhood impact estimate.

[0022] Preferably, the determination of the collaborative weighting factor includes:

[0023] Extract the shadow coverage of the neighboring cells;

[0024] The collaborative weighting factor is calculated based on the shadow coverage rate and the preset system cooperation coefficient.

[0025] Preferably, the local state vector further includes:

[0026] The unique identifier and real-time output power of the photovoltaic tracking unit.

[0027] A photovoltaic module shading optimization system includes:

[0028] The state acquisition module is used to acquire the local state vector of the photovoltaic tracking unit and receive the neighborhood state vectors of each neighboring unit in the neighborhood of the photovoltaic tracking unit.

[0029] An angle generation module is used to calculate the reference angle of the photovoltaic tracking unit and generate a discrete set of alternative disturbance angles.

[0030] The gain evaluation module is used to calculate the local gain prediction and the neighborhood impact estimate for any candidate perturbation angle, and combine the two to generate a cooperative gain.

[0031] The decision optimization module is used to select the optimal perturbation angle that maximizes the cooperative gain and to determine the target execution angle;

[0032] The control execution module is used to output control commands based on the target execution angle.

[0033] Preferably, the neighborhood influence estimation in the gain evaluation module is achieved in the following way:

[0034] Obtain the effective irradiance of the photovoltaic tracking unit;

[0035] Calculate the change in shadow area projected onto the neighboring cell due to the alternative perturbation angle;

[0036] The estimated power change for the neighboring unit is calculated.

[0037] For each neighboring unit in the neighborhood, the power change estimate is weighted based on a dynamically adaptive collaborative weighting factor, and all weighted results are summed to generate the neighborhood impact estimate.

[0038] Preferably, the collaborative weighting factor is determined in the following way:

[0039] Extract the shadow coverage of the neighboring cells;

[0040] The collaborative weighting factor is calculated based on the shadow coverage rate and the preset system cooperation coefficient.

[0041] Preferably, the local state vector acquired by the state acquisition module further includes:

[0042] The unique identifier and real-time output power of the photovoltaic tracking unit.

[0043] The present invention provides a photovoltaic module shading optimization system and inverse tracking method, which, compared with the prior art, have the following advantages:

[0044] 1. This invention transforms the photovoltaic array into an adaptive multi-agent system by endowing each photovoltaic tracking unit with real-time status perception and neighborhood communication capabilities. This system can dynamically compensate for physical world deviations such as construction errors and terrain undulations, thus solving the problem of power generation efficiency loss caused by the reliance on idealized static models in traditional inverse tracking algorithms.

[0045] 2. This invention introduces cooperative gain as the decision objective. By combining local gain prediction and neighborhood influence estimation, it quantifies the externality of local decision on the neighborhood, enabling the system decision to shift from pursuing individual optimality to maximizing the overall power generation of the array, thus significantly improving the comprehensive energy output of the power station.

[0046] 3. This invention provides a physically meaningful and computationally efficient neighborhood impact assessment method by combining real-time effective irradiance, geometric shadow change, and photoelectric conversion efficiency. This ensures that the cooperative strategy can operate reliably on resource-constrained edge controllers and has strong engineering applicability.

[0047] 4. This invention establishes an adaptive collaborative priority scheduling mechanism by introducing a collaborative weighting factor that is dynamically associated with the neighbor's shadow coverage rate. This enables the system to prioritize helping the neighbor unit with the worst power generation status and guide the entire array to converge to the global optimal working state more quickly in complex shading scenarios such as sunrise, sunset, or drifting clouds. Attached Figure Description

[0048] The present invention will be further explained below with reference to the accompanying drawings and embodiments:

[0049] Figure 1 This is a flowchart of the method of the present invention;

[0050] Figure 2 This is a flowchart of the process for generating neighborhood influence estimation in this invention. Detailed Implementation

[0051] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0052] Example 1:

[0053] Please see Figure 1 A method for optimizing shading of photovoltaic modules, comprising:

[0054] Obtain the local state vector of the photovoltaic tracking unit;

[0055] Receive the neighborhood state vectors of each neighboring unit within the neighborhood of the photovoltaic tracking unit;

[0056] The local state vector includes the real-time rotation angle and shadow coverage of the photovoltaic tracking unit;

[0057] The reference angle of the photovoltaic tracking unit is calculated based on the solar position algorithm;

[0058] A discrete set of alternative perturbation angles is generated around the reference angle;

[0059] For any candidate perturbation angle in the set of candidate perturbation angles, calculate the local gain prediction and estimate the neighborhood impact.

[0060] By combining local gain prediction and neighborhood influence estimation, a cooperative gain is generated.

[0061] Select the alternative perturbation angle that maximizes the cooperative gain as the optimal perturbation angle;

[0062] By combining the baseline angle and the optimal disturbance angle, the target execution angle is determined;

[0063] Based on the target execution angle, output control commands to adjust the execution angle;

[0064] The local state vector further includes: a unique identifier for the photovoltaic tracking unit and its real-time output power.

[0065] A photovoltaic module shading optimization system includes:

[0066] The state acquisition module is used to acquire the local state vector of the photovoltaic tracking unit and receive the neighborhood state vectors of each neighboring unit in the neighborhood of the photovoltaic tracking unit.

[0067] An angle generation module is used to calculate the reference angle of the photovoltaic tracking unit and generate a discrete set of alternative disturbance angles;

[0068] The gain evaluation module is used to calculate the local gain prediction and the neighborhood impact estimate for any candidate perturbation angle, and combine the two to generate a cooperative gain.

[0069] The decision optimization module is used to select the optimal perturbation angle that maximizes the cooperative gain and to determine the target execution angle;

[0070] The control execution module is used to output control commands based on the target execution angle;

[0071] This embodiment discloses a photovoltaic module shading optimization system and corresponding optimization method. The system is applied to a photovoltaic array containing multiple photovoltaic tracking units and aims to overcome the power generation efficiency loss caused by the reliance on idealized static models in traditional inverse tracking algorithms. The present invention constructs each photovoltaic tracking unit as an intelligent agent with perception, communication and decision-making capabilities. Through a distributed collaborative mechanism, it compensates for physical world deviations such as construction errors and terrain undulations in real time, thereby maximizing the overall power generation of the power station.

[0072] Each photovoltaic tracking unit in this system is equipped with a status acquisition module, an angle generation module, a gain evaluation module, a decision optimization module, and a control execution module.

[0073] The state acquisition module aims to provide the necessary self-state data for collaborative decision-making within this unit and to acquire external state information from neighboring units. In this embodiment, this module consists of a high-precision tilt sensor, a power sensor, a photodiode array, and a neighborhood communication module. It is responsible for acquiring and maintaining a local state vector in real time. It periodically receives its neighborhood set through the neighborhood communication module. All neighboring units The neighborhood state vector of the broadcast ;

[0074] To achieve the above collaborative decision-making, a local state vector is defined. As the fundamental data source for all subsequent calculations, its function is to comprehensively characterize the physical and operational state of the tracking unit at any given moment; in this embodiment, this vector Specifically, it includes:

[0075] Unique Identifier This refers to the unique network address of each tracking unit. Its function is to accurately identify the source of information in neighborhood communication and ensure the correctness of data attribution. Its source is assigned during system initialization.

[0076] Real-time rotation angle : refers to the real-time tilt angle of the photovoltaic module plane measured by the tilt sensor. Its function is to serve as a key input parameter for geometric shading estimation of neighboring units. Its source is the real-time reading of the local tilt sensor.

[0077] Real-time output power : refers to the instantaneous output power of the photovoltaic module as measured by the power sensor. Its function is to provide a benchmark for the verification of the local power generation model and to serve as a reference for calculating its own gain. Its source is the real-time reading of the local power sensor.

[0078] Shadow coverage : refers to a range characterized by an array of photodiodes. The dimensionless parameter within the interval is used to quantitatively assess the degree to which the surface of the photovoltaic module is shaded. This parameter is a key custom parameter of this invention. It is not only used to assess its own state, but more importantly, after being broadcast to the neighbors, it will become the direct input for the collaborative weight calculation in the neighbor decision-making model, thereby dynamically influencing the behavior decisions of the neighbors. Its source is calculated by processing the readings of the photodiode array deployed on the surface of the module.

[0079] Specifically, the calculation method includes: calibrating the photodiode array under ideal illumination conditions without any obstruction to obtain the reference reading value of each diode; setting an activation threshold, such as 80% of the reference reading value; counting in real time the number of diodes in the array whose reading value is lower than the activation threshold, and using the ratio of this number to the total number of diodes in the array as the shadow coverage rate. .

[0080] Based on the above system architecture, the detailed process of the photovoltaic module shading optimization method in this embodiment is as follows:

[0081] The initial step of this method involves the state acquisition module performing state acquisition and broadcasting; each unit Obtain its own local state vector And through the neighborhood communication module to its neighborhood All neighboring units within Broadcast the vector, and simultaneously receive neighborhood state vectors from these neighbors. This step constructs a dynamic distributed information network, which is a prerequisite for subsequent collaborative decision-making.

[0082] The angle generation module calculates the reference angle and generates a set of alternative disturbance angles; reference angle This refers to the unit under ideal conditions without considering any obstruction. To achieve the theoretically optimal tracking angle for maximizing solar energy capture, its function is to provide a calculation starting point for subsequent compensatory fine-tuning. In this embodiment, it is calculated using a standard solar position algorithm based on local precise time and geographic location information. This module generates a finite, discrete set of alternative perturbation angles around the reference angle. ,For example This set constitutes the space of all possible adjustment actions within the current decision-making cycle, and its range of values ​​is based on balancing the precision of the adjustment with the computational load.

[0083] The core step involves the gain evaluation module evaluating the set of candidate perturbation angles. Any of the alternative perturbation angles Calculate its cooperative gain; the cooperative gain here ( () is a quantitative indicator whose purpose is to evaluate the contribution of an alternative action to the overall power generation gain of the unit and its neighborhood; according to the limitations of Example 1, this gain consists of two parts: calculating the local gain prediction. This value is based on the local photoelectric conversion model and predicts the execution. The change in its own power can be expressed by the complete mathematical model as follows:

[0084] ;

[0085] The estimation of neighborhood influence will be detailed later; specifically, the local gain prediction can be calculated using the following model:

[0086] ;

[0087] in, From the local state vector The current real-time output power obtained from the data; The predicted output power after performing the perturbation angle is calculated as follows:

[0088] ;

[0089] In this model, The preset photoelectric conversion efficiency, The effective light-receiving area of ​​a photovoltaic module. The effective irradiance is measured locally in real time, and This represents the rotation angle when executed. Then, the angle of incidence between the sunlight and the normal of the photovoltaic module; this angle of incidence can be determined based on the solar position algorithm and the target execution angle. The model, which combines geometric angle changes with real-time environmental and equipment parameters, provides a quantitative physical basis for local decision-making.

[0090] The optimal decision is made by the decision optimization module; this module iterates through all candidate perturbation angles and selects the one that maximizes the cooperative gain. The largest alternative perturbation angle is taken as the optimal perturbation angle. By combining the baseline angle and the optimal disturbance angle, the final target execution angle is determined. ;

[0091] The control execution module executes the target angle. It outputs control commands to actuators such as servo motors to achieve precise adjustment of the rotation angle of photovoltaic modules;

[0092] This embodiment transforms the entire photovoltaic array from a group of isolated individuals passively executing preset trajectories into an adaptive multi-agent system capable of local information interaction and collaborative decision-making through the aforementioned system and method. This approach no longer relies on an idealized geometric model that fails to reflect reality. Instead, through real-time state perception and distributed computing, it enables each tracking unit's decisions to quantify the externalities of its behavior on the neighborhood. This allows the system to dynamically and precisely compensate for non-ideal shading introduced by factors such as construction errors, terrain undulations, and structural deformation, guiding the entire array to dynamically converge to a better overall operating state. This overcomes the bottlenecks of existing technologies and significantly improves the actual total power generation of dense photovoltaic power plants in complex real-world environments.

[0093] Example 2:

[0094] Please see Figure 2 The calculation method for the neighborhood influence estimation includes:

[0095] Obtain the effective irradiance of the photovoltaic tracking unit;

[0096] Calculate the change in shadow area projected onto neighboring cells due to alternative disturbance angles;

[0097] By combining the effective irradiance and the change in shadow area, and based on the preset photoelectric conversion efficiency, the estimated power change of the neighboring cell is calculated.

[0098] For each neighboring cell in the neighborhood, the power change estimate is weighted based on a dynamic adaptive collaborative weighting factor, and all weighted results are summed to generate a neighborhood impact estimate.

[0099] The neighborhood influence estimation in the gain evaluation module is achieved in the following way:

[0100] Obtain the effective irradiance of the photovoltaic tracking unit;

[0101] Calculate the change in shadow area projected onto neighboring cells due to alternative disturbance angles;

[0102] Calculate the estimated power change for neighboring cells;

[0103] For each neighboring cell in the neighborhood, the power change estimate is weighted based on a dynamic adaptive collaborative weighting factor, and all weighted results are summed to generate a neighborhood impact estimate.

[0104] Based on the previous embodiment, this embodiment specifies the calculation method for neighborhood influence estimation in the gain evaluation module. Its core purpose is to provide the decision-making unit with a tool that has clear physical meaning and can be efficiently calculated on the edge device to evaluate the actual power impact of its behavior on neighboring units.

[0105] To achieve this estimation, the underlying calculation logic is as follows:

[0106] This estimation process obtains the local effective irradiance. As a premise; effective irradiance This refers to the solar energy density that is currently and actually irradiated onto the surface of the photovoltaic module, as measured in real time by the light intensity sensor in the status acquisition module. Its unit is watts per square meter (W / m²). Its function is to correlate subsequent geometric occlusion calculations with the current real-time environmental energy intensity, ensuring the physical accuracy of the power estimation;

[0107] Based on this, calculate the change in shaded area. This parameter refers to the condition when this unit... Implement an alternative perturbation angle This will cause projection onto neighboring units. The change in the shaded area of ​​the surface, expressed in square meters (m²). Its function is to quantitatively describe the shading effect of a rotational action on its neighbors in geometric space. In this embodiment, this value is calculated by constructing a simplified ray tracing model. The simplification of this ray tracing model lies in: abstracting the light-receiving surface of each photovoltaic module as a two-dimensional rectangular plane and ignoring the thickness of the module; treating sunlight as parallel light without diffraction effects; and combining the solar altitude angle and azimuth angle obtained based on the solar position algorithm with the values ​​of this unit. with neighboring units Given the relative three-dimensional spatial positions, the element can be directly calculated using the geometric projection method. Rotate alternative disturbance angle At that time, it was related to the neighboring unit. The change in the area of ​​shading caused by the surface;

[0108] The model's input includes units with neighbors Known relative geometric position, from neighbor state vectors Real-time rotation angle obtained from and alternative perturbation angles ;

[0109] Based on the above parameters and the preset photoelectric conversion efficiency, the estimated power change of the neighboring cell is calculated. To transform the abstract geometric occlusion change into a power estimate with clear physical meaning, this embodiment introduces the following calculation model:

[0110] ;

[0111] in, :unit Execute corner Expected impact on neighbors The resulting change in output power; Watts (W); calculated using this formula;

[0112] Photovoltaic module photoelectric conversion efficiency is an inherent parameter characterizing module performance; it is dimensionless; this parameter is a known equipment specification parameter when the photovoltaic module leaves the factory, and it is determined and preset in the system during the power plant design stage.

[0113] :unit Effective irradiance measured locally; dimensionless in watts per square meter (W / m²) (), is collected in real time by the light intensity sensor in the local state perception module;

[0114] :unit Rotation To the neighbors The resulting change in shaded area; the dimension is square meters ( ), calculated from the simplified ray tracing model in the preceding steps;

[0115] The introduction of this computational model, by using real-time physical environment energy intensity With geometric occlusion Multiply, and combine with known equipment efficiency. This enables reliable quantification of the impact on neighboring power;

[0116] Based on a dynamically adaptive collaborative weighting factor The estimated power changes of all neighbors are weighted and then aggregated to generate a neighborhood impact estimate. ;

[0117] This embodiment provides a neighborhood impact assessment method that combines physical realism with engineering feasibility. By anchoring to locally measured physical parameters rather than relying on complex global environment models, and combining simplified geometric calculations, this method significantly reduces the computational complexity of collaborative decision-making, enabling it to run efficiently on resource-constrained edge controllers. This design ensures that the collaborative strategy is not only theoretically superior, but also robust, efficient, and low-cost in practical applications, thus successfully realizing the concept of distributed collaborative optimization into a large-scale deployable engineering solution.

[0118] Example 3:

[0119] The determination of the synergistic weighting factor includes:

[0120] Extract the shadow coverage of neighboring cells;

[0121] Calculate the collaborative weighting factor based on the shadow coverage rate and the preset system cooperation coefficient;

[0122] The collaborative weighting factor is determined in the following way:

[0123] Extract the shadow coverage of neighboring cells;

[0124] Calculate the collaborative weighting factor based on the shadow coverage rate and the preset system cooperation coefficient;

[0125] This embodiment, based on the previous embodiment, adjusts the collaborative weighting factor. The determination method has been specified. The purpose of designing this weighting factor is to enable the system's cooperative behavior to have the ability to dynamically and adaptively adjust priorities, and to give priority to helping the neighboring unit with the worst current power generation status.

[0126] Determine the collaborative weighting factor The process is as follows:

[0127] From neighboring units The neighborhood state vector of the broadcast Extract its shadow coverage. This step ensures that the weighting factor calculation is based on the latest, true occlusion status of the neighbors.

[0128] Based on extracted shadow coverage With a preset system cooperation coefficient The collaborative weighting factor is calculated; to achieve the above-mentioned adaptive priority adjustment goal, the following weight calculation model is constructed in this embodiment:

[0129] ;

[0130] in, Collaborative weighting factor, representing the unit In its decision-making, the neighbors The degree of importance attached to interests; dimensionless; calculated by this formula;

[0131] The system cooperation coefficient is a globally adjustable non-negative parameter used to adjust the overall altruistic tendency of the entire photovoltaic array. It is dimensionless. This parameter can be manually set according to the specific characteristics of the power station during the system commissioning phase, or the optimal value can be obtained through simulation optimization of the digital twin model of the power station.

[0132] Neighboring Units Real-time shadow coverage, dimensionless; from neighbors broadcast state vector Extract from;

[0133] This model reflects the neighbor dilemma. This is dynamically introduced into the decision-making process of this unit, realizing the intelligent allocation of decision priorities; when a neighbor... When severely obstructed, its The higher the value, the higher the calculated weight. It will also increase accordingly;

[0134] In the cooperative gain calculation of the previous embodiment, a larger weight This unit will be enlarged. A certain action of the neighbor The positive or negative impact, thereby guiding the unit Prefer to choose those that can help neighbors more effectively To overcome the obstruction of the rotation angle, even if it might mean sacrificing a small amount of its own power generation gain;

[0135] when hour, The value is always equal to 1, and the model degenerates into a basic collaborative mode that treats all neighbors equally; The larger the value, the more the system tends to sacrifice local interests to help severely occluded units, thereby pursuing a higher level of global optimal solution;

[0136] This embodiment injects an intelligent adaptive priority scheduling mechanism into the entire collaborative system. Compared with fixed or indiscriminate collaborative strategies, this method enables the system's mutual assistance behavior to accurately and dynamically focus on the units that need help the most at the moment. This decision-making tendency guided by the overall interests of the neighborhood can effectively avoid global suboptimal results caused by the accumulation of local optimal decisions. Especially in scenarios with complex and non-uniform shading such as sunrise, sunset, or cloud drift, it can more quickly guide the entire array to converge to a working state with higher overall power generation, thereby achieving a more extreme utilization of solar energy resources.

[0137] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for optimizing shading of photovoltaic modules, characterized in that, include: Obtain the local state vector of the photovoltaic tracking unit; Receive the neighborhood state vectors of each neighboring unit within the neighborhood of the photovoltaic tracking unit; The local state vector includes the real-time rotation angle and shadow coverage of the photovoltaic tracking unit; The reference angle of the photovoltaic tracking unit is calculated based on the solar position algorithm; A discrete set of alternative perturbation angles is generated around the reference angle; For any candidate perturbation angle in the set of candidate perturbation angles, calculate the local gain prediction and estimate the neighborhood impact. By combining the local gain prediction with the neighborhood influence estimation, a cooperative gain is generated; Select the candidate perturbation angle that maximizes the cooperative gain as the optimal perturbation angle; By combining the reference angle and the optimal disturbance angle, the target execution angle is determined; Based on the target execution angle, output control commands to adjust the execution angle; The calculation method for the neighborhood influence estimation includes: Obtain the effective irradiance of the photovoltaic tracking unit; Calculate the change in shadow area projected onto the neighboring cell due to the alternative perturbation angle; By combining the effective irradiance and the change in shadow area, and based on the preset photoelectric conversion efficiency, the estimated power change of the neighboring unit is calculated. For each neighboring unit in the neighborhood, the power change estimate is weighted based on a dynamic adaptive collaborative weighting factor, and all weighted results are summed to generate the neighborhood impact estimate. Determine the collaborative weighting factor The calculation model is as follows: ； in, Collaborative weighting factor; System cooperation coefficient; Neighboring Units Real-time shadow coverage; when a neighbor When severely obstructed, its Higher values ​​result in higher calculated weights. It will also increase accordingly.

2. The photovoltaic module shading optimization method according to claim 1, characterized in that, The determination of the collaborative weighting factor includes: Extract the shadow coverage of the neighboring cells; The collaborative weighting factor is calculated based on the shadow coverage rate and the preset system cooperation coefficient.

3. The photovoltaic module shading optimization method according to claim 1, characterized in that, The local state vector further includes: The unique identifier and real-time output power of the photovoltaic tracking unit.

4. A photovoltaic module shading optimization system, applied to the photovoltaic module shading optimization method according to any one of claims 1-3, characterized in that, include: The state acquisition module is used to acquire the local state vector of the photovoltaic tracking unit and receive the neighborhood state vectors of each neighboring unit in the neighborhood of the photovoltaic tracking unit. An angle generation module is used to calculate the reference angle of the photovoltaic tracking unit and generate a discrete set of alternative disturbance angles. The gain evaluation module is used to calculate the local gain prediction and the neighborhood impact estimate for any candidate perturbation angle, and combine the two to generate a cooperative gain. The decision optimization module is used to select the optimal perturbation angle that maximizes the cooperative gain and to determine the target execution angle; The control execution module is used to output control commands based on the target execution angle.

5. A photovoltaic module shading optimization system according to claim 4, characterized in that, The neighborhood influence estimation in the gain evaluation module is achieved in the following way: Obtain the effective irradiance of the photovoltaic tracking unit; Calculate the change in shadow area projected onto the neighboring cell due to the alternative perturbation angle; The estimated power change for the neighboring unit is calculated. For each neighboring unit in the neighborhood, the power change estimate is weighted based on a dynamically adaptive collaborative weighting factor, and all weighted results are summed to generate the neighborhood impact estimate.

6. A photovoltaic module shading optimization system according to claim 5, characterized in that, The determination of the collaborative weighting factor is achieved in the following manner: Extract the shadow coverage of the neighboring cells; The collaborative weighting factor is calculated based on the shadow coverage rate and the preset system cooperation coefficient.

7. A photovoltaic module shading optimization system according to claim 6, characterized in that, The local state vector acquired by the state acquisition module further includes: The unique identifier and real-time output power of the photovoltaic tracking unit.

Images