Distributed photovoltaic intelligent production cluster control method and system

Abstract

The invention discloses a distributed photovoltaic intelligent production cluster control method and system, and the method comprises the following steps: S1, collecting various data of a photovoltaic cluster, and carrying out the Kalman filtering and denoising; s2, performing anomaly detection and normalization processing on the de-noised sensor data; s3, establishing a power model and a cluster power model of the distributed photovoltaic module; s4, establishing an objective function of the distributed photovoltaic intelligent production cluster control model; s5, establishing constraint conditions of the distributed photovoltaic intelligent production cluster control model; s6, solving the established distributed photovoltaic intelligent production cluster control model; and S7, evaluating a solution result of the distributed photovoltaic intelligent production cluster control model. According to the method, the overall operation efficiency of the photovoltaic system is improved, more flexible power service is provided for users, and wide application of renewable energy sources is promoted. Experimental results show that the system remarkably improves the power generation utilization rate and reduces the operation cost.

Description

Technical Field

[0001] The present invention relates to the field of photovoltaic cluster systems, and in particular to a distributed photovoltaic intelligent production cluster control method and system. Background Art

[0002] With the transformation of the global energy structure, distributed photovoltaic power generation has attracted much attention due to its renewability, environmental protection and flexibility. Especially in urban and rural areas, the deployment of distributed photovoltaic systems is increasing rapidly. Compared with traditional centralized power systems, distributed photovoltaic systems can more effectively utilize ground resources, reduce transmission losses and improve energy utilization. However, this system also faces many challenges, including but not limited to:

[0003] Photovoltaic power generation is affected by weather and geographical conditions, resulting in instability in power generation capacity. The access of distributed photovoltaic systems may have an impact on the power grid, which requires effective control and management. Load fluctuations on the user side will affect the utilization efficiency of photovoltaic power generation, which needs to be balanced through intelligent scheduling. Summary of the invention

[0004] The purpose of the present invention is to provide a distributed photovoltaic intelligent production cluster control method and system to solve the problems raised in the above-mentioned background technology. Through the intelligent control method, the operating status of the photovoltaic system can be monitored and adjusted in real time to improve the power generation efficiency and energy utilization. The cluster control system can reduce the impact on the power grid by coordinating the output of each photovoltaic power generation unit, thereby ensuring the stability and security of the power grid. By using advanced algorithms and models to optimize the configuration of distributed photovoltaic resources, more reasonable energy distribution can be achieved and costs can be reduced, and intelligent control technology can further promote its widespread application. Through intelligent scheduling and prediction, users can be provided with more flexible and efficient electricity services, and their sense of participation and satisfaction can be improved. Therefore, it will not only help to meet current technical challenges, but also provide important support for future sustainable energy development.

[0005] In order to achieve the above object, the present invention provides a distributed photovoltaic intelligent production cluster control method, comprising:

[0006] S1: Collect various data of photovoltaic clusters and perform Kalman filtering to denoise;

[0007] S2: Perform anomaly detection and normalization on the denoised sensor data;

[0008] S3: Establish the power model of distributed photovoltaic components and the cluster power model;

[0009] S4: Establish the objective function of the distributed photovoltaic smart production cluster control model;

[0010] S5: Establish constraints for distributed photovoltaic smart production cluster control model;

[0011] S6: Solve the established distributed photovoltaic smart production cluster control model;

[0012] S7: Evaluate the solution results of the distributed photovoltaic smart production cluster control model.

[0013] Furthermore, the step S1 comprises:

[0014] Various sensors are installed on each photovoltaic module to monitor the following parameters: light intensity E i (t), temperature T i (t), voltage V i (t) and current I i (t);

[0015] Assume that each sensor collects data once at time t, and the data collection process is expressed as:

[0016]

[0017] Where: E i (t), T i (t), V i (t), I i (t) are the data of light intensity, temperature, voltage and current collected by the i-th photovoltaic module at time t; S E , S T , S V , S I are the true values ​​of the light intensity sensor, temperature sensor, voltage sensor and current sensor respectively; n E (t), n T (t), n V (t), n I (t) is the corresponding sensor noise;

[0018] Use Kalman filtering to denoise the data of each sensor, and set the state equation as:

[0019] x(t)=Ax(t-1)+Bu(t)+w(t)

[0020] The observation equation is:

[0021] z(t)=Hx(t)+v(t)

[0022] Where: z(t) is the observation data; A is the state transfer matrix; H is the observation matrix; B is the control input matrix; u(t) is the control input vector; w(t) and v(t) are process noise and observation noise, both of which are zero-mean Gaussian noise.

[0023] Furthermore, the step S2 comprises:

[0024] The data of each sensor is normalized as follows:

[0025]

[0026] Among them, max(x i (t)) and min(x i (t)) represent the maximum and minimum values ​​of the sensor data sequence respectively; represents the normalized sensor data, x i (t) is the sensor data sequence;

[0027] Then define the anomaly detection threshold to determine whether it is abnormal:

[0028]

[0029] Where: Z(t) is the outlier judgment index, μ is the average value of sensor data, and σ is the standard deviation of sensor data; when Z(t) is greater than or equal to 3 or Z(t) is less than or equal to -3, it is judged as abnormal; the detected abnormal data is corrected or discarded, and the nearest neighbor interpolation method (KNN) or mean replacement method is used for data filling; Represents the normalized sensor data.

[0030] Furthermore, the following steps are also included to establish a power model of distributed photovoltaic components and a cluster power model:

[0031] The power output of each photovoltaic module is not only affected by the light intensity, but also related to factors such as temperature and radiation angle. Suppose the output power of the photovoltaic module is:

[0032] P i (t) = E i (t)·M i ·η i (T i (t),θ i (t))

[0033] Where: M i is the area of ​​the photovoltaic module, θ i (t) is the radiation angle, η i (T i (t),θ i (t)) is the efficiency corrected for temperature and radiation angle, specifically:

[0034] η i (T i (t),θ i (t)) = η i,0·(1-β(T i -T ref ))

[0035] Where: η i,0 is the nominal efficiency, β is the temperature coefficient, T ref is the reference temperature;

[0036] Calculate the total power of the photovoltaic cluster:

[0037]

[0038] Where n is the number of photovoltaic modules.

[0039] Furthermore, the step S4 comprises:

[0040] The objective function of the distributed photovoltaic smart production cluster control model includes maximizing total power output, minimizing photovoltaic module losses, and optimizing photovoltaic module efficiency;

[0041] The maximum total power output is expressed as:

[0042]

[0043] Minimizing the photovoltaic module loss is expressed as:

[0044]

[0045] Among them, L loss (t) is the loss of the photovoltaic module, R i is the resistance of the ith component, I i is the current of the i-th component;

[0046] The optimized photovoltaic module efficiency is expressed as:

[0047]

[0048] Among them, P in (t) is the total input power at time t;

[0049] Finally, the objective function of the distributed photovoltaic intelligent production cluster control model is:

[0050] minJ=w2J2-w1J1-w3J3

[0051] Among them, w1, w2 and w3 are weight coefficients used to balance the relationship between different objectives.

[0052] Furthermore, the step S5 comprises:

[0053] The constraints of the distributed photovoltaic smart production cluster control model include: total power balance constraints, component power constraints and dynamic constraints;

[0054] The total power balance constraint is:

[0055] P total (t) = P in (t)+L loss (t)

[0056] Among them, L loss (t) is the photovoltaic module loss, expressed by the loss function;

[0057] The component power constraints are:

[0058] P i,min ≤P i (t)≤P i,max (i=1,2,...,n)

[0059] Among them, P i,max is the maximum output power of the ith PV module, P i,min is the minimum power output of the ith PV module;

[0060] The dynamic constraints include the change rate constraint of the power output of the i-th photovoltaic module:

[0061] |P i (t)-P i (t-1)|≤ΔP max

[0062] Among them, P i (t-1) is the photovoltaic module power value at time t-1, ΔP max is the maximum power change value at two moments.

[0063] Furthermore, the step S6 comprises:

[0064] Calculate the minimization objective function J(u),

[0065]

[0066] Where N represents the total number of calculation cycles, k is the time delay value, x(t+k) is the control variable at time t+k, u(t+k) is the control input at time t+k, and L(x(t+k),u(t+k)) is the loss function, which is:

[0067]

[0068] Where: α is the weight factor, P total(t+k) is the total power at time t+k, L(t+k) is the loss function at time t+k, u i (t+k) is the control input at time t+k;

[0069] Using the gradient descent method, calculate the gradient of the minimized objective function J(u) with respect to the control input u:

[0070]

[0071] Then the gradient of the minimized objective function J(u) is:

[0072]

[0073] in, is the partial differential symbol, is the objective function for the input variable u n Partial differential of ; computed using the chain rule:

[0074]

[0075] Where k represents the delay time;

[0076] The gradient descent method is used to update the control input, specifically:

[0077]

[0078] Where η is the learning rate;

[0079] Set the learning rate η and the maximum number of iterations, and iteratively update the control input until convergence or the maximum number of iterations is reached; set the convergence condition as:

[0080]

[0081] Among them, ║˙║ is the norm, ε a A positive number indicates the optimization accuracy.

[0082] Furthermore, the step S7 includes: evaluating the solution result of the distributed photovoltaic smart production cluster control model, and evaluating the control effect through the following performance indicators:

[0083] The total output efficiency is:

[0084]

[0085] The control accuracy is expressed as:

[0086]

[0087] Among them, P target is the target power.

[0088] The present invention also discloses a distributed photovoltaic intelligent production cluster control system, including:

[0089] Photovoltaic power generation module: including photovoltaic modules and inverters;

[0090] Data acquisition and monitoring module: collects data from photovoltaic modules, temperature sensors, light intensity sensors, and humidity sensors for real-time monitoring and historical data analysis;

[0091] Communication and network module: establish communication networks between each module and the system and external environment network and transmit data;

[0092] Optimization Control Module: Develop and implement cluster control strategies to optimize the scheduling between generation, loads, and storage to achieve balanced and stable power supply.

[0093] Compared with the prior art, the beneficial effects of the present invention are as follows: the present invention proposes a distributed photovoltaic intelligent production cluster control method and system, which aims to improve the efficiency of distributed photovoltaic power generation and the stability of the power grid. The system combines advanced sensor technology, data analysis and intelligent control algorithms to monitor the operating status of each photovoltaic unit in real time, predict power generation capacity and load demand, and thus achieve optimized scheduling and resource allocation. Through cluster control, the system can coordinate the output of each photovoltaic power generation unit and balance the instability caused by weather changes and load fluctuations. At the same time, the machine learning algorithm is used to analyze historical data to further improve the accuracy of power generation prediction, reduce the impact on the power grid, and enhance the system's adaptive ability. This method not only improves the overall operating efficiency of the photovoltaic system, but also provides users with more flexible power services and promotes the widespread application of renewable energy. The experimental results show that the system significantly improves the power generation utilization rate, reduces operating costs, and provides important technical support for the development of distributed photovoltaics. BRIEF DESCRIPTION OF THE DRAWINGS

[0094] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0095] Figure 1 This is a flow chart of the distributed photovoltaic intelligent production cluster control method studied in the present invention;

[0096] Figure 2 The light intensity variation curve studied in the present invention;

[0097] Figure 3 The photovoltaic module power generation prediction curve studied in the present invention;

[0098] Figure 4This is a comparison curve of power generation before and after control studied in the present invention. DETAILED DESCRIPTION

[0099] The principles and features of the present invention are described below in conjunction with the accompanying drawings. The examples given are only used to explain the present invention and are not used to limit the scope of the present invention. The present invention is described in more detail by way of example with reference to the accompanying drawings in the following paragraphs. The advantages and features of the present invention will become clearer according to the following description and claims. It should be noted that the drawings are all in a very simplified form and are not in precise proportions, and are only used to conveniently and clearly assist in explaining the purpose of the embodiments of the present invention.

[0100] Unless otherwise defined, all technical and scientific terms used in the present invention have the same meaning as those commonly understood by those skilled in the art of the present invention. The terms used in the specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention.

[0101] Embodiment 1:

[0102] Please see attached Figure 1 , a distributed photovoltaic intelligent production cluster control method, comprising:

[0103] S1: Collect various data of photovoltaic clusters and perform Kalman filtering to denoise;

[0104] Various sensors are installed on each photovoltaic module to monitor the following parameters: light intensity E i (t), temperature T i (t), voltage V i (t) and current I i (t);

[0105] Assume that each sensor collects data once at time t, and the data collection process is expressed as:

[0106]

[0107] Where: E i (t), T i (t), V i (t), I i (t) are the data of light intensity, temperature, voltage and current collected by the i-th photovoltaic module at time t; S E , S T , S V , S I are the true values ​​of the light intensity sensor, temperature sensor, voltage sensor and current sensor respectively; n E (t), n T (t), n V (t), n I(t) is the corresponding sensor noise;

[0108] Use Kalman filtering to denoise the data of each sensor, and set the state equation as:

[0109] x(t)=Ax(t-1)+Bu(t)+w(t)

[0110] The observation equation is:

[0111] z(t)=Hx(t)+v(t)

[0112] Where: z(t) is the observation data; A is the state transfer matrix; H is the observation matrix; B is the control input matrix; u(t) is the control input vector; w(t) and v(t) are process noise and observation noise, both of which are zero-mean Gaussian noise.

[0113] The combination of photovoltaic cluster data collection and Kalman filtering denoising can significantly improve the quality and accuracy of data. Kalman filtering can effectively extract signals from noise and improve data accuracy.

[0114] S2: Perform anomaly detection and normalization on the denoised sensor data;

[0115] Normalize the data of each sensor for subsequent processing, specifically:

[0116]

[0117] Among them, max(x i (t)) and min(x i (t)) represent the maximum and minimum values ​​of the sensor data sequence respectively; represents the normalized sensor data, x i (t) is the sensor data sequence;

[0118] Then define the anomaly detection threshold to determine whether it is abnormal:

[0119]

[0120] Where: Z(t) is the outlier judgment index, μ is the average value of sensor data, and σ is the standard deviation of sensor data; when Z(t) is greater than or equal to 3 or Z(t) is less than or equal to -3, it is judged as abnormal; the detected abnormal data is corrected or discarded, and the nearest neighbor interpolation method (KNN) or mean replacement method is used for data filling; Represents the normalized sensor data.

[0121] Combining anomaly detection and normalization processing can effectively improve the quality and availability of sensor data, and provide a more reliable information basis for the management and optimization of photovoltaic clusters. This not only helps to improve the performance of the system, but also reduces operation and maintenance costs and ensures the long-term stable operation of the system.

[0122] S3: Establish the power model of distributed photovoltaic components and the cluster power model;

[0123] Assume the output power of the photovoltaic module is:

[0124] P i (t) = E i (t)·M i ·η i (T i (t),θ i (t))

[0125] Where: M i is the area of ​​the photovoltaic module, θ i (t) is the radiation angle, η i (T i (t),θ i (t)) is the efficiency corrected for temperature and radiation angle, specifically:

[0126] η i (T i (t),θ i (t)) = η i,0 ·(1-β(T i -T ref ))

[0127] Where: η i,0 is the nominal efficiency, β is the temperature coefficient, T ref is the reference temperature;

[0128] Calculate the total power of the photovoltaic cluster:

[0129]

[0130] Where n is the number of photovoltaic modules.

[0131] S4: Establish the objective function of the distributed photovoltaic smart production cluster control model;

[0132] The objective functions of the distributed photovoltaic smart production cluster control model mainly include maximizing total power output, minimizing system losses and optimizing system efficiency;

[0133] The maximum total power output is expressed as:

[0134]

[0135] Minimizing the photovoltaic module loss is expressed as:

[0136]

[0137] Among them, L loss (t) is the loss of the photovoltaic module, R i is the resistance of the ith component, I i is the current of the i-th component;

[0138] The optimized photovoltaic module efficiency is expressed as:

[0139]

[0140] Among them, P in (t) is the total input power at time t;

[0141] Finally, the objective function of the distributed photovoltaic intelligent production cluster control model is:

[0142] minJ=w2J2-w1J1-w3J3

[0143] Among them, w1, w2 and w3 are weight coefficients used to balance the relationship between different objectives.

[0144] The objective function plays an optimization role in the control model of distributed photovoltaic smart production clusters, ensuring that the system can reach the optimal state in terms of economy, technology and other aspects. It not only affects the operating efficiency, but is also crucial to the realization of continuous optimization.

[0145] S5: Establish constraints for distributed photovoltaic smart production cluster control model;

[0146] Includes: total power balance constraints, component power constraints, and dynamic constraints;

[0147] The total power balance constraint is:

[0148] P total (t) = P in (t)+L loss (t)

[0149] Among them, L loss (t) is the photovoltaic module loss, expressed by the loss function;

[0150] The component power constraints are:

[0151] P i,min ≤P i (t)≤P i,max (i=1,2,...,n)

[0152] Among them, P i,maxis the maximum output power of the ith PV module, P i,min is the minimum power output of the ith PV module;

[0153] The dynamic constraints include the change rate constraint of the power output of the i-th photovoltaic module:

[0154] |P i (t)-P i (t-1)|≤ΔP max

[0155] Among them, P i (t-1) is the photovoltaic module power value at time t-1, ΔP max is the maximum power change value at two moments.

[0156] S6: Solve the established distributed photovoltaic smart production cluster control model;

[0157] Calculate the minimization objective function J(u),

[0158]

[0159] Where N represents the total number of calculation cycles, k is the time delay value, x(t+k) is the control variable at time t+k, u(t+k) is the control input at time t+k, and L(x(t+k),u(t+k)) is the loss function, which is:

[0160]

[0161] Where: α is the weight factor, P total (t+k) is the total power at time t+k, L(t+k) is the loss function at time t+k, u i (t+k) is the control input at time t+k;

[0162] Using the gradient descent method, calculate the gradient of the minimized objective function J(u) with respect to the control input u:

[0163]

[0164] Then the gradient of the minimized objective function J(u) is:

[0165]

[0166] in, is the partial differential symbol, is the objective function for the input variable u n Partial differential of ; computed using the chain rule:

[0167]

[0168] Where k represents the delay time;

[0169] The gradient descent method is used to update the control input, specifically:

[0170]

[0171] Where η is the learning rate;

[0172] Set the learning rate η and the maximum number of iterations, and iteratively update the control input until convergence or the maximum number of iterations is reached; set the convergence condition as:

[0173]

[0174] Among them, ║˙║ is the norm, ε a A positive number indicates the optimization accuracy.

[0175] In the distributed photovoltaic smart production cluster control model, it is possible to effectively solve the optimization target, improve the system's operating efficiency and economic benefits, and promote the realization of intelligent management. Through continuous iteration and adjustment, it is possible to ensure that the system can operate stably and efficiently in a complex environment.

[0176] S7: Evaluate the solution results of the distributed photovoltaic smart production cluster control model;

[0177] The evaluation of the distributed photovoltaic smart production cluster control model solution results, and the control effect is evaluated through the following performance indicators:

[0178] The total output efficiency is:

[0179]

[0180] The control accuracy is expressed as:

[0181]

[0182] Among them, P target is the target power.

[0183] In order to verify the effectiveness of the method proposed in this invention, a distributed photovoltaic intelligent production cluster control system based on MATLAB is designed. The system consists of: Photovoltaic unit: set to 5 photovoltaic modules, located in different geographical locations. Control strategy: a hybrid strategy based on centralized control and distributed control. Environmental variables: consider the impact of meteorological data (light intensity, temperature, etc.) on power generation.

[0184] Parameter settings include: rated power of each photovoltaic module: 300W; module light intensity range: 0-1000W / m 2; Temperature influence coefficient of the module: -0.5% / ℃; Experimental period: 24 hours, time step is 1 hour;

[0185] The experimental steps include: data collection: simulate and obtain light intensity and temperature data within 24 hours; power generation prediction: use historical data and real-time monitoring data to predict the power generation capacity of each module; control strategy execution: apply control algorithms to adjust the output of each photovoltaic unit; performance evaluation: compare the power generation efficiency and total power generation of each module before and after control.

[0186] Figure 2 The light intensity changes over a 24-hour period. At 6 a.m., the light intensity starts to increase and reaches a peak value (1000 W / m 2 ) at 12 noon, and then gradually weakened until the evening. This map provides basic data for subsequent power generation forecasts. Figure 3 The power generation of each photovoltaic module predicted according to the light intensity is shown. It can be seen that the power generation is the highest at noon, about 300W. This result verifies the positive correlation between light intensity and power generation. Figure 4 The power generation of each module before and after the control is compared. By implementing the control strategy, the power generation increased by about 10%. This shows the effectiveness of the intelligent control strategy in optimizing the efficiency of photovoltaic power generation.

[0187] Through the above experiments, the effectiveness of the distributed photovoltaic smart production cluster control method was verified. The results show that the system can optimize power generation efficiency under different environmental conditions and effectively meet load requirements, providing strong support for future distributed photovoltaic applications.

[0188] Embodiment 2:

[0189] The present invention also discloses a distributed photovoltaic intelligent production cluster control system, including:

[0190] Photovoltaic power generation module: including photovoltaic modules and inverters;

[0191] Data acquisition and monitoring module: collects data from photovoltaic modules, temperature sensors, light intensity sensors, and humidity sensors for real-time monitoring and historical data analysis;

[0192] Communication and network module: establish communication networks between each module and the system and external environment network and transmit data;

[0193] Optimization Control Module: Develop and implement cluster control strategies to optimize the scheduling between generation, loads, and storage to achieve balanced and stable power supply.

[0194] The above description is only a preferred embodiment of the present invention and does not limit the present invention in any form. Any ordinary technician in the industry can smoothly implement the present invention as shown in the drawings and described above. However, any equivalent changes, modifications and evolutions made by technicians familiar with the profession without departing from the scope of the technical solution of the present invention using the technical content disclosed above are all equivalent embodiments of the present invention. At the same time, any equivalent changes, modifications and evolutions made to the above embodiments based on the essential technology of the present invention are still within the protection scope of the technical solution of the present invention.

Claims

1. A distributed photovoltaic intelligent production cluster control method, characterized in that: include: S1: Collect various data of photovoltaic clusters and perform Kalman filtering to denoise; S2: Perform anomaly detection and normalization on the denoised sensor data; S3: Establish the power model of distributed photovoltaic components and the cluster power model; S4: Establish the objective function of the distributed photovoltaic smart production cluster control model; S5: Establish constraints for distributed photovoltaic smart production cluster control model; S6: Solve the established distributed photovoltaic smart production cluster control model; S7: Evaluate the solution results of the distributed photovoltaic smart production cluster control model.

2. The distributed photovoltaic intelligent production cluster control method according to claim 1 is characterized in that: The step S1 comprises: Various sensors are installed on each photovoltaic module to monitor the following parameters: light intensity E i (t), temperature T i (t), voltage V i (t) and current I i (t); Assume that each sensor collects data once at time t, and the data collection process is expressed as: Where: E i (t), T i (t), V i (t), I i (t) are the data of light intensity, temperature, voltage and current collected by the i-th photovoltaic module at time t; S E , S T , S V , S I are the true values ​​of the light intensity sensor, temperature sensor, voltage sensor and current sensor respectively; n E (t), n T (t), n V (t), n I (t) is the corresponding sensor noise; Use Kalman filtering to denoise the data of each sensor, and set the state equation as: x(t)=Ax(t-1)+Bu(t)+w(t) The observation equation is: z(t)=Hx(t)+v(t) Where: z(t) is the observation data; A is the state transfer matrix; H is the observation matrix; B is the control input matrix; u(t) is the control input vector; w(t) and v(t) are the process noise and observation noise.

3. The distributed photovoltaic intelligent production cluster control method according to claim 2 is characterized in that: The step S2 comprises: The data of each sensor is normalized as follows: Among them, max(x i (t)) and min(x i (t)) represent the maximum and minimum values ​​of the sensor data sequence respectively; represents the normalized sensor data, x i (t) is the sensor data sequence; Then define the anomaly detection threshold to determine whether it is abnormal: Where: Z(t) is the outlier judgment index, μ is the average value of sensor data, and σ is the standard deviation of sensor data; when Z(t) is greater than or equal to 3 or Z(t) is less than or equal to -3, it is judged as abnormal; the detected abnormal data is corrected or discarded, and the nearest neighbor interpolation method or mean replacement method is used for data filling; Represents the normalized sensor data.

4. The distributed photovoltaic intelligent production cluster control method according to claim 2 is characterized in that: The step S3 comprises: Assume the output power of the photovoltaic module is: P i (t)=E i (t)·M i ·η i (T i (t),θ i (t)) Where: M i is the area of ​​the photovoltaic module, θ i (t) is the radiation angle, η i (T i (t),θ i (t)) is the efficiency corrected for temperature and radiation angle, specifically: or i (T i (t),θ i (t))=η i,0 ·(1-β(T i -T ref )) Where: η i,0 is the nominal efficiency, β is the temperature coefficient, T ref is the reference temperature; Calculate the total power of the photovoltaic cluster: Where n is the number of photovoltaic modules.

5. The distributed photovoltaic intelligent production cluster control method according to claim 4 is characterized in that: The step S4 comprises: The objective function of the distributed photovoltaic smart production cluster control model includes maximizing total power output, minimizing photovoltaic module losses, and optimizing photovoltaic module efficiency; The maximum total power output is expressed as: Minimizing the photovoltaic module loss is expressed as: Among them, L loss (t) is the loss of the photovoltaic module, R i is the resistance of the i-th photovoltaic module, I i is the current of the i-th component; The optimized photovoltaic module efficiency is expressed as: Among them, P in (t) is the total input power at time t; Finally, the objective function of the distributed photovoltaic intelligent production cluster control model is: minJ=w2J2-w1J1-w3J3 Among them, w1, w2 and w3 are weight coefficients used to balance the relationship between different objectives.

6. The distributed photovoltaic intelligent production cluster control method according to claim 5 is characterized in that: The step S5 comprises: The constraints of the distributed photovoltaic smart production cluster control model include: total power balance constraints, component power constraints and dynamic constraints; The total power balance constraint is: P total (t)=P in (t)+L loss (t) Among them, L loss (t) is the photovoltaic module loss, expressed by the loss function; The component power constraints are: P i,min ≤P i (t)≤P i,max (i=1,2,...,n) Among them, P i,max is the maximum output power of the ith PV module, P i,min is the minimum power output of the ith PV module; The dynamic constraints include the change rate constraint of the power output of the i-th photovoltaic module: |P i (t)-P i (t-1)|≤ΔP max Among them, P i (t-1) is the photovoltaic module power value at time t-1, ΔP max is the maximum power change value at two moments.

7. The distributed photovoltaic intelligent production cluster control method according to claim 6 is characterized in that: The step S6 comprises: Calculate the minimization objective function J(u), Where N represents the total number of calculation cycles, k is the time delay value, x(t+k) is the control variable at time t+k, u(t+k) is the control input at time t+k, and L(x(t+k),u(t+k)) is the loss function, which is: Where: α is the weight factor, P total (t+k) is the total power at time t+k, L(t+k) is the loss function at time t+k, u i (t+k) is the control input at time t+k; Using the gradient descent method, calculate the gradient of the minimized objective function J(u) with respect to the control input u: Then the gradient of the minimized objective function J(u) is: in, is the partial differential symbol, is the objective function for the input variable u n Partial differential of ; computed using the chain rule: Where k represents the delay time; The control input is updated using the gradient descent method, specifically: Where η is the learning rate; Set the learning rate η and the maximum number of iterations, and iteratively update the control input until convergence or the maximum number of iterations is reached; set the convergence condition as: Among them, ║˙║ is the norm, ε a A positive number indicates the optimization accuracy.

8. The distributed photovoltaic intelligent production cluster control method according to claim 6 is characterized in that: The step S7 includes: evaluating the solution result of the distributed photovoltaic smart production cluster control model, and evaluating the control effect through the following performance indicators: The total output efficiency is: The control accuracy is expressed as: Among them, P target is the target power.

9. A distributed photovoltaic intelligent production cluster control system, characterized in that: include: Photovoltaic power generation module: including photovoltaic modules and inverters; Data acquisition and monitoring module: collects data from photovoltaic modules, temperature sensors, light intensity sensors, and humidity sensors for real-time monitoring and historical data analysis; Communication and network module: establish communication networks between each module and the system and external environment network and transmit data; Optimization control module: formulate and implement cluster control strategies to optimize the scheduling between generation, load and storage to achieve balanced and stable power supply; To realize a distributed photovoltaic intelligent production cluster control method as described in any one of claims 1-8.

Images