Control method for photovoltaic energy power generation

Abstract

The invention relates to the technical field of photovoltaic energy power generation, and discloses a control method for photovoltaic energy power generation. According to the method, the output power of a photovoltaic unit is monitored in real time, and an abnormal unit is identified and processed; establishing a multi-source data fusion model to predict power generation efficiency, generating a dynamic power distribution strategy, and adjusting parameters of energy storage equipment and a grid-connected inverter; acquiring real-time environment data to dynamically adjust the angle and orientation of the photovoltaic panel; constructing a virtual mapping model to optimize system parameters; and the standby energy storage mode is switched when the power grid load fluctuation exceeds the range. According to the invention, the power generation efficiency and stability of the photovoltaic energy power generation system are effectively improved, the fault processing capability is enhanced, the loss of energy storage equipment is reduced, intelligent and efficient control of photovoltaic energy power generation is realized, and the method is of great significance in promoting the development of the photovoltaic energy industry.

Description

Technical Field

[0001] The present invention relates to the technical field of photovoltaic energy generation, and in particular to a control method for photovoltaic energy generation. Background Art

[0002] As the global demand for clean energy grows, photovoltaic energy, as a sustainable green energy, has been widely used in the field of power supply. However, the current photovoltaic energy generation technology still faces many challenges, which limits its further development and efficient utilization.

[0003] During the operation of photovoltaic arrays, performance differences and failures of photovoltaic units are common. Some photovoltaic units may be shielded by foreign objects such as dust and leaves, or have internal faults, resulting in a decrease in output power. However, existing monitoring methods often have difficulty in quickly and accurately identifying these abnormal units, so that the faults cannot be handled in time, which in turn affects the power generation efficiency of the entire photovoltaic array. For example, some traditional monitoring methods rely only on simple power threshold judgments, which are prone to misjudgment, misjudging normally fluctuating photovoltaic units as abnormal, or missing units that actually have problems.

[0004] At the same time, the uncertainty of meteorological conditions has a great impact on photovoltaic energy generation. Meteorological factors such as light intensity and temperature change all the time, and photovoltaic systems have poor adaptability to these changes. Under different light and temperature conditions, the optimal tilt angle and orientation of photovoltaic panels will be different, but most of the existing photovoltaic panel adjustment methods are relatively fixed and cannot be dynamically optimized according to real-time environmental data, resulting in photovoltaic panels not being able to fully receive light and greatly reducing power generation efficiency. For example, in the early morning and evening, the sun's angle is low. If the photovoltaic panel cannot adjust its angle in time, the effective light time and intensity will be reduced.

[0005] In addition, the fluctuation of grid load is also a thorny issue. When the grid load fluctuates greatly, it is difficult for the photovoltaic system to respond quickly and stabilize the power output, which can easily cause an impact on the grid and affect the normal operation of the grid. Moreover, the existing photovoltaic system has deficiencies in working in coordination with energy storage equipment. The charging and discharging management of energy storage equipment is not intelligent enough, and it is impossible to make reasonable plans based on multiple factors such as power generation efficiency and grid load demand, resulting in energy waste and reduced system stability.

[0006] Existing photovoltaic systems lack effective simulation and optimization methods. When faced with different control strategies, it is difficult to intuitively evaluate their impact on energy conversion efficiency, making it impossible to adjust the control strategy in time to achieve optimal performance during system operation. For example, when adjusting the dynamic adjustment parameters of photovoltaic panels and the charge and discharge curves of energy storage devices, there is a lack of scientific basis, and adjustments can only be made through experience, which cannot fully tap the potential of the system. Summary of the invention

[0007] The object of the present invention is to provide a control method for photovoltaic energy generation to solve the problems raised in the above background technology.

[0008] To achieve the above object, the present invention provides the following technical solution: a control method for photovoltaic energy generation, the method comprising: Monitor the output power of each photovoltaic unit in the photovoltaic array in real time and identify abnormal photovoltaic units whose output power is lower than the preset threshold; Establishing a multi-source data fusion model for the photovoltaic system, wherein the multi-source data includes historical power generation data, weather forecast data, and grid load demand data; Predicting power generation efficiency in a future time period according to the multi-source data fusion model, and generating a corresponding dynamic power allocation strategy; Based on the dynamic power allocation strategy, adjusting the charging and discharging parameters of the energy storage device and the output power of the grid-connected inverter; Perform local shading detection on abnormal PV cells, determine the type of shading source, and trigger adaptive cleaning or fault isolation operations based on the type of shading source; Construct a virtual mapping model of the photovoltaic system to simulate the energy conversion efficiency under different control strategies in real time; According to the simulation results of the virtual mapping model, the dynamic adjustment parameters of the photovoltaic panel and the charge and discharge curve of the energy storage device are optimized; When it is detected that the grid load fluctuation exceeds the preset range, it switches to the backup energy storage mode and prioritizes balancing power output through energy storage equipment.

[0009] Preferably, the method further comprises: Acquire real-time environmental data of the photovoltaic system, the environmental data including light intensity, ambient temperature and photovoltaic panel surface temperature; Based on the environmental data, dynamically adjust the tilt angle and orientation of the photovoltaic panel so that the photovoltaic panel receives maximum effective light; the dynamically adjusting the tilt angle and orientation of the photovoltaic panel includes the following steps: Divide the light intensity data into multiple gradient intervals, each gradient interval corresponds to a preset tilt angle adjustment range; According to the difference between the ambient temperature and the surface temperature of the photovoltaic panel, the tilt angle adjustment range is corrected; Dynamic adjustment instructions are generated based on the gradient range and temperature difference, and the angle and direction of the photovoltaic panel are controlled by the servo mechanism.

[0010] Preferably, the step of identifying an abnormal photovoltaic unit whose output power is lower than a preset threshold comprises the following steps: Periodically collect the output current and voltage of each photovoltaic unit and calculate the instantaneous power; Compare the instantaneous power with the power of adjacent photovoltaic units horizontally, and select candidate abnormal units whose deviation rate exceeds a first threshold; A power trend analysis is performed on the candidate abnormal unit for a continuous period of time. If the power continues to decrease and the slope exceeds a second threshold, it is determined to be an abnormal photovoltaic unit.

[0011] Preferably, the steps of establishing a multi-source data fusion model include: The historical power generation data is segmented according to the time series, and the power generation efficiency characteristics of each time period are extracted; Temporally and spatially correlating the meteorological forecast data with the power generation efficiency characteristics to generate a forecast weight matrix; Combined with the priority labels of power grid load demand data, a multi-dimensional data fusion framework is constructed.

[0012] Preferably, the step of generating a dynamic power allocation strategy includes: Divide the power generation efficiency levels according to the prediction weight matrix, and assign corresponding energy storage equipment charging and discharging priorities to each level; Dynamically adjust the charging and discharging priority based on real-time changes in grid load demand data; The power allocation sequence in the future time period is optimized by a sliding window algorithm.

[0013] Preferably, adjusting the charging and discharging parameters of the energy storage device comprises the following steps: Calculate the maximum adjustable power range based on the remaining capacity and charging and discharging efficiency of the current energy storage device; Inserting a buffer zone in the power allocation sequence to smooth power mutations during charging and discharging; The optimal charge and discharge rate is determined through binary search method to minimize the life loss of energy storage equipment.

[0014] Preferably, the local shading detection comprises the following steps: Use infrared thermal imagers to collect surface temperature distribution data of abnormal photovoltaic units; According to the gradient change direction of the temperature distribution data, it is determined whether the shielding source is a static object or a dynamic interference; If it is a static object, the robotic arm is triggered to perform cleaning operations; if it is a dynamic interference, the backup circuit is activated to bypass the abnormal unit.

[0015] Preferably, the step of constructing a virtual mapping model includes: Generate three-dimensional energy distribution map based on the physical parameters of photovoltaic panels and environmental data; Embed the time dimension in the energy distribution diagram to simulate the changes in lighting conditions at different time periods; The control parameters are randomly sampled by the Monte Carlo method to evaluate the system efficiency under various parameter combinations.

[0016] Preferably, the step of optimizing the dynamic adjustment parameters of the photovoltaic panel includes: Extract the top N parameter combinations with the highest efficiency in the virtual mapping model, where N is a preset positive integer constant; Matching the parameter combination with the real-time environmental data to select the combination with the highest degree of adaptability; The control logic of the servo mechanism is updated through an incremental learning algorithm.

[0017] Preferably, the switching to the backup energy storage mode comprises the following steps: Monitor the fluctuation frequency and amplitude of power grid load in real time and calculate the fluctuation index; If the fluctuation index exceeds a third threshold, the direct output link of the grid-connected inverter is cut off; The direct current of the energy storage device is converted into alternating current matching the grid through a bidirectional converter and injected into the grid.

[0018] Compared with the prior art, the present invention has the following beneficial effects: In terms of improving power generation efficiency, by real-time monitoring of the output power of each photovoltaic unit in the photovoltaic array, abnormal photovoltaic units with output power lower than the preset threshold can be identified in time. Local shading detection is performed on these abnormal units, and adaptive cleaning or fault isolation operations are triggered according to the type of shading source, effectively reducing the loss of power generation efficiency caused by the shielding. For example, when a static object is detected to be shading, the robotic arm quickly cleans and removes dust, leaves, etc., so that the photovoltaic unit can fully receive light again and restore power generation efficiency. At the same time, real-time environmental data of the photovoltaic system, such as light intensity, ambient temperature and photovoltaic panel surface temperature, are obtained, and the tilt angle and orientation of the photovoltaic panel are dynamically adjusted based on these data to ensure that the photovoltaic panel receives the maximum effective light. The light intensity data is divided into multiple gradient intervals, and the tilt angle adjustment range is corrected by combining the difference between the ambient temperature and the photovoltaic panel surface temperature. This refined adjustment method can make the photovoltaic panel in the best lighting state under different environmental conditions, significantly improving the power generation efficiency of the photovoltaic panel.

[0019] From the perspective of power generation stability, a multi-source data fusion model for the photovoltaic system is established to integrate historical power generation data, meteorological forecast data, and grid load demand data. The power generation efficiency in the future time period is predicted based on the model, and the corresponding dynamic power allocation strategy is generated. This enables the photovoltaic system to predict the power generation situation in advance, and adjust the charging and discharging parameters of the energy storage equipment and the output power of the grid-connected inverter in real time in combination with the grid load demand. When the grid load fluctuates, the system can respond quickly and maintain the stable operation of the grid through reasonable power allocation. For example, when the grid load is at a peak, the energy storage device discharges in time to supplement the power; when the load is at a low point, the excess power is stored in the energy storage device, avoiding grid fluctuations caused by the mismatch between power generation and power consumption, and enhancing the stability of power supply.

[0020] In terms of fault handling and system reliability, the present invention has a complete mechanism. Accurate detection and classification of abnormal photovoltaic units effectively avoids the expansion of faults. When an abnormality caused by dynamic interference is detected, the backup circuit is started to bypass the abnormal unit to ensure the continuous power generation of the entire photovoltaic array. In addition, a virtual mapping model of the photovoltaic system is constructed to simulate the energy conversion efficiency under different control strategies in real time. The dynamic adjustment parameters of the photovoltaic panels and the charging and discharging curves of the energy storage equipment are optimized according to the simulation results, further improving the reliability and adaptability of the system. At the same time, when it is detected that the grid load fluctuation exceeds the preset range, it quickly switches to the backup energy storage mode, and prioritizes the power output through the energy storage device to prevent damage to the photovoltaic system due to grid anomalies, thereby ensuring the stable operation of the photovoltaic system in a complex grid environment.

[0021] In terms of cost control and sustainable development, the charging and discharging parameters of energy storage equipment are optimized, and the life loss of energy storage equipment is minimized by calculating the maximum adjustable power range, inserting buffers to smooth power mutations, and using binary search to determine the optimal charging and discharging rate, thereby reducing the cost of replacing energy storage equipment. In addition, the efficient operation of the entire system reduces energy waste and improves the utilization efficiency of photovoltaic energy, which is in line with the concept of sustainable development and provides strong support for the long-term development of the photovoltaic energy industry. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 This is a working principle diagram of the control method for photovoltaic energy generation according to the present invention; Figure 2 Flow chart for abnormal PV unit identification; Figure 3 A flow chart for adjusting the charging and discharging parameters of energy storage equipment; Figure 4 Flowchart for detection and processing of abnormal photovoltaic unit partial shading. DETAILED DESCRIPTION

[0023] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0024] See also Figure 1-Figure 4 The present invention provides a technical solution: a control method for photovoltaic energy generation, which aims to improve the stability, efficiency and adaptability of photovoltaic energy generation systems to complex working conditions, and achieve efficient utilization and reliable output of photovoltaic energy. The specific implementation steps are as follows: Abnormal photovoltaic unit monitoring and identification: During the operation of the photovoltaic array, the output power of each photovoltaic unit is monitored in real time through professional monitoring equipment. Once the output power of a photovoltaic unit is found to be lower than the preset threshold, it will be preliminarily identified as an abnormal photovoltaic unit for further processing.

[0025] Establishment of multi-source data fusion model: Collect historical power generation data of photovoltaic systems, meteorological forecast data and grid load demand data. Use these data to establish a multi-source data fusion model to provide strong support for subsequent power generation efficiency prediction and power allocation strategy formulation.

[0026] Power generation efficiency prediction and dynamic power allocation strategy generation: With the help of the established multi-source data fusion model, the power generation efficiency in the future time period is predicted. Based on the prediction results, the corresponding dynamic power allocation strategy is generated to ensure that the photovoltaic system can reasonably allocate power under different working conditions.

[0027] Adjustment of energy storage equipment and grid-connected inverter parameters: According to the dynamic power allocation strategy, the charging and discharging parameters of the energy storage equipment and the output power of the grid-connected inverter are accurately adjusted. This step is crucial for balancing the power output of the photovoltaic system and improving energy utilization.

[0028] Abnormal PV unit processing: For the identified abnormal PV units, local shading detection is performed to determine the type of shading source. According to the type of shading source, adaptive cleaning or fault isolation operations are automatically triggered to ensure the normal operation of the PV system.

[0029] Virtual mapping model construction and simulation: Construct a virtual mapping model of the photovoltaic system and use it to simulate the energy conversion efficiency under different control strategies in real time. The simulation results provide a reference for optimizing the operating parameters of the photovoltaic system.

[0030] Operation parameter optimization: Based on the simulation results of the virtual mapping model, the dynamic adjustment parameters of the photovoltaic panels and the charge and discharge curves of the energy storage equipment are optimized. By continuously optimizing these parameters, the overall performance of the photovoltaic system is improved.

[0031] Switching to backup energy storage mode: Real-time monitoring of grid load fluctuations. When the grid load fluctuations exceed the preset range, the system will quickly switch to backup energy storage mode. In backup energy storage mode, power output is balanced through energy storage equipment to ensure stable operation of the grid.

[0032] The present invention will be further described below in conjunction with Examples 1 to 6:

[0033] Embodiment 1: In actual application scenarios, this control method has a detailed operation process for the acquisition and processing of environmental data of the photovoltaic system operation and the identification of abnormal photovoltaic units.

[0034] The first step is to obtain real-time environmental data of the photovoltaic system, which is achieved by installing light intensity sensors, ambient temperature sensors, and photovoltaic panel surface temperature sensors around the photovoltaic panels. The light intensity sensor uses a high-precision silicon photocell sensor, which can accurately measure the light intensity received within a specific area in lux (lx). The ambient temperature sensor uses a thermistor sensor, which can monitor the temperature of the surrounding air in real time in degrees Celsius (℃). The photovoltaic panel surface temperature sensor uses a thermocouple sensor to directly measure the temperature of the photovoltaic panel surface, also in degrees Celsius (℃). These sensors transmit the collected data to the data processing center in real time.

[0035] Dynamically adjusting the tilt angle and orientation of the photovoltaic panel based on environmental data is of great significance to improving the efficiency of the photovoltaic panel in receiving light. First, the light intensity data is divided into multiple gradient intervals, for example, 0-200lx is divided into a low light interval, 201-800lx is divided into a medium light interval, and 801lx and above is divided into a high light interval. Each gradient interval corresponds to a preset tilt angle adjustment range. In the low light interval, the tilt angle adjustment range is set to 0°-15°, the purpose is to make the photovoltaic panel as perpendicular to the incident direction of light as possible to increase the amount of light received; the adjustment range of the medium light interval is 10°-30°. Considering that the light intensity is moderate at this time, it is necessary to balance the amount of light received and the stability of the equipment; the high light interval is 20°-40°, which avoids excessive light from damaging the photovoltaic panel while ensuring sufficient light reception.

[0036] Then, the tilt angle adjustment range is corrected according to the difference between the ambient temperature and the surface temperature of the photovoltaic panel. , the surface temperature of the photovoltaic panel is ,when When , it means that the temperature of the photovoltaic panel is too high. To reduce the temperature, the tilt angle is appropriately increased on the basis of the original adjustment range to increase air circulation and heat dissipation. For example, in the high light range, the adjustment range becomes 25° - 45°; when When the light is low, the tilt angle can be appropriately reduced to enhance the absorption of light by the photovoltaic panels. For example, the adjustment range in the low light range can be changed to 0° - 10°.

[0037] Finally, a dynamic adjustment command is generated based on the gradient interval and temperature difference, and the angle and direction of the photovoltaic panel are controlled by the servo mechanism. After receiving the command, the servo mechanism uses the motor and transmission device to accurately adjust the tilt angle and rotation direction of the photovoltaic panel so that the photovoltaic panel receives the maximum effective light.

[0038] In terms of abnormal photovoltaic unit identification, the output current and voltage of each photovoltaic unit are collected periodically. The current sensor and voltage sensor are used for collection. The current sensor uses a Hall current sensor, which can accurately measure the current value in the circuit in amperes (A); the voltage sensor uses a resistor divider voltage sensor to measure the voltage in volts (V). The collection cycle is set to 5 minutes, and the instantaneous power is calculated after collection. ,in is the collected voltage value, is the collected current value.

[0039] Compare the instantaneous power with the power of adjacent photovoltaic units horizontally, and select candidate abnormal units whose deviation rate exceeds the first threshold (set to 15%). , the average power of adjacent photovoltaic units is , deviation rate ,when , the unit is listed as a candidate abnormal unit.

[0040] The power trend analysis of the candidate abnormal unit is carried out for a continuous period of time (set to 1 hour). If the power continues to decrease and the slope exceeds the second threshold (set to -5W / min), it is determined to be an abnormal photovoltaic unit. By performing linear fitting on the power data, the power change slope is calculated. If the slope is less than -5W / min, the unit can be determined to be an abnormal photovoltaic unit for subsequent timely processing.

[0041] Embodiment 2: The multi-source data fusion model is a key step to achieve accurate power generation efficiency prediction and reasonable power allocation. In actual operation, the historical power generation data is segmented according to the time series. The historical power generation data of the past year is divided into 1 hour time periods. In each time period, the power generation efficiency features are extracted, such as the average power generation power, maximum power, power generation duration, etc. in the time period.

[0042] The meteorological forecast data is temporally and spatially associated with the power generation efficiency characteristics. The meteorological forecast data includes the light intensity forecast value, temperature forecast value, wind speed forecast value, etc. Taking the light intensity forecast value as an example, if the light intensity is predicted to be high in a certain period in the future, and the historical data shows that the power generation efficiency is high under similar light intensity, then when the time and space are correlated, the power generation efficiency characteristic weight of this period is correspondingly increased. In this way, a prediction weight matrix is ​​generated, which reflects the degree of influence of different meteorological conditions on power generation efficiency.

[0043] A multi-dimensional data fusion framework is constructed by combining the priority tags of power grid load demand data. Power grid load demand data is divided into multiple priorities according to different user types and power consumption periods. For example, industrial power consumption is high priority during production peaks, and residential power consumption is low priority during nighttime valleys. Integrating these priority tags into the data fusion framework allows the model to fully consider changes in power grid load demand when predicting power generation efficiency and formulating power allocation strategies.

[0044] In terms of generating dynamic power allocation strategies, the power generation efficiency levels are divided according to the prediction weight matrix. The power generation efficiency is divided into three levels: high, medium, and low. For example, power generation efficiency above 80% is high, 50% - 80% is medium, and below 50% is low. The corresponding energy storage device charging and discharging priority is assigned to each level. When the power generation efficiency is high, energy storage charging is prioritized to store excess electric energy; when it is medium, charging and discharging are dynamically adjusted according to the load demand of the power grid; when it is low, discharge is prioritized to meet the load demand.

[0045] Based on the real-time changes in grid load demand data, the charging and discharging priority is adjusted dynamically. When the grid load is at peak times, even if the power generation efficiency is at a high level, the energy storage charging priority is appropriately reduced to give priority to power supply to the grid; when the load is at low times, the energy storage charging priority is increased.

[0046] The power allocation sequence in the future time period is optimized through the sliding window algorithm. The sliding window size is set to 1 hour, and the window contains power allocation data at multiple time points. At each time point, the optimal power allocation plan is calculated based on the current power generation efficiency, energy storage status and grid load demand. The window is continuously sliding and the power allocation sequence is updated in real time to ensure the rationality and efficiency of power allocation.

[0047] Embodiment 3: Adjusting the charging and discharging parameters of energy storage equipment is an important part of ensuring the stable operation of photovoltaic systems and improving energy utilization. In actual operation, the maximum adjustable power range is first calculated based on the remaining capacity and charging and discharging efficiency of the current energy storage equipment. Assume that the rated capacity of the energy storage equipment is (Unit: kilowatt-hour, kWh), the remaining capacity is (Unit: kWh), the charging and discharging efficiency is (dimensionless), then the maximum charging power , maximum discharge power ,in and They are the set charging and discharging time (unit: hours, h) respectively.

[0048] Insert a buffer zone in the power allocation sequence to smooth out power mutations during the charging and discharging process. For example, insert a 10-minute buffer zone between two adjacent power allocation points. In the buffer zone, the power changes linearly from one allocation value to the next, avoiding shocks to energy storage devices and the power grid caused by power mutations.

[0049] The optimal charge and discharge rate is determined by binary search to minimize the life loss of energy storage equipment. The life loss of energy storage equipment is closely related to the charge and discharge rate. A life loss function is set ,in is the charge and discharge rate (unit: kilowatt, kW). Within the maximum adjustable power range, use the binary search method to continuously try different charge and discharge rates and calculate the corresponding life loss value. Each time the power range is halved, the half with smaller life loss is selected to continue searching until the charge and discharge rate with the smallest life loss is found. For example, the initial power range is , first attempt ,calculate ,like Less than , then the next time Search within the range, otherwise Search within the range, and repeat this cycle until the optimal charge and discharge rate is found.

[0050] Embodiment 4: Detecting partial shading of abnormal photovoltaic units and taking corresponding measures is an important means to ensure the power generation efficiency of photovoltaic systems. In practical applications, infrared thermal imagers are used to collect surface temperature distribution data of abnormal photovoltaic units. Infrared thermal imagers generate temperature distribution images by detecting infrared radiation emitted by objects, with a resolution of up to 0.1°C.

[0051] According to the gradient change direction of the temperature distribution data, it is judged that the shielding source is a static object or dynamic interference. If the temperature distribution shows a local low temperature and the gradient change is relatively stable, it may be a static object shielding, such as leaves, dust, etc.; if the temperature distribution changes rapidly and there is no obvious pattern, it may be a dynamic interference, such as flying birds, clouds, etc.

[0052] If it is a static object, the robot arm is triggered to perform cleaning operations. The robot arm is installed near the photovoltaic panel and has flexible joints and cleaning devices such as brushes or nozzles. After receiving the cleaning command, the robot arm moves to the abnormal photovoltaic unit according to the pre-set path planning, removes the obstruction by brushing or spraying water, and restores the normal power generation capacity of the photovoltaic unit.

[0053] If it is a dynamic interference, the backup circuit is started to bypass the abnormal unit. The backup circuit has been pre-laid when the photovoltaic system is designed. When dynamic interference is detected and the abnormal unit cannot work normally, the control system automatically switches to the backup circuit to bypass the abnormal unit, ensuring the normal power output of the entire photovoltaic array and reducing the power generation loss caused by dynamic interference.

[0054] Embodiment 5: Building a virtual mapping model and optimizing the dynamic adjustment parameters based on its simulation results can help improve the overall performance of the photovoltaic system. When building the virtual mapping model, a three-dimensional energy distribution diagram is generated based on the physical parameters and environmental data of the photovoltaic panel. The physical parameters of the photovoltaic panel include the size, material properties, conversion efficiency, etc. of the photovoltaic panel, and environmental data such as light intensity and ambient temperature. Taking the length, width and height of the photovoltaic panel as the three dimensions, combined with the distribution of light intensity at different locations, a three-dimensional energy distribution diagram is generated to intuitively display the energy reception of the photovoltaic panel at different locations.

[0055] The time dimension is embedded in the energy distribution diagram to simulate the changes in lighting conditions at different time periods. For example, from morning to evening, the light intensity and angle change continuously. By updating the lighting conditions in the energy distribution diagram in chronological order, the energy conversion process of photovoltaic panels at different times of the day can be simulated.

[0056] The control parameters are randomly sampled by the Monte Carlo method to evaluate the system efficiency under each parameter combination. The control parameters include the tilt angle and orientation of the photovoltaic panel, the charging and discharging strategy of the energy storage device, etc. The sampling number is set to 1000 times, and a set of control parameters is randomly generated each time, and input into the virtual mapping model to calculate the system efficiency. For example, a random tilt angle of 25°, an orientation of 10° east of due south, and an energy storage charging priority of high level are generated, and the system power generation efficiency at this time is calculated.

[0057] In terms of optimizing dynamic adjustment parameters, the top N efficiency rankings in the virtual mapping model are extracted (set ) parameter combinations. The 10 most efficient parameter combinations were selected from the 1000 sampling results.

[0058] These parameter combinations are matched with the real-time environmental data to select the combination with the highest adaptability. For example, if the real-time environmental data shows that the light intensity is high and the temperature is low, the parameter combination with higher efficiency in similar environments is preferred.

[0059] The control logic of the servo mechanism is updated through the incremental learning algorithm. The parameter combination with the highest adaptability is used as a new learning sample and input into the incremental learning algorithm. The algorithm adjusts and optimizes the control logic of the servo mechanism based on the new sample, so that the servo mechanism can more accurately adjust the angle and direction of the photovoltaic panel according to the real-time environment and system status, thereby improving the power generation efficiency of the photovoltaic system.

[0060] Embodiment 6: When the grid load fluctuation exceeds the preset range, switching to the backup energy storage mode is crucial to ensure the stable operation of the grid. In actual operation, the fluctuation frequency and amplitude of the grid load are monitored in real time, and the fluctuation index is calculated. The voltage and current data of the grid load are collected through the grid monitoring equipment to calculate the change of the load power. Assume that the grid load power changes at the time interval The change in , the initial power is , the fluctuation frequency is , then the volatility index .

[0061] If the fluctuation index exceeds the third threshold (set to 0.2), the direct output link of the grid-connected inverter is cut off. This operation is achieved through the relay in the control circuit. When the fluctuation index is detected to be excessive, a control signal is quickly sent to activate the relay to disconnect the grid-connected inverter from the grid to prevent unstable power from being input into the grid.

[0062] The bidirectional converter converts the DC power of the energy storage device into AC power that matches the grid and injects it into the grid. The bidirectional converter has rectification and inversion functions. In the backup energy storage mode, it converts the DC power in the energy storage device into AC power that matches the grid voltage, frequency and phase. The conversion process is precisely adjusted by the control circuit to ensure that the output power quality meets the grid requirements, stably supplies power to the grid, balances the power output, and ensures the normal operation of the grid.

[0063] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0064] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A control method for photovoltaic energy generation, characterized in that: include: Monitor the output power of each photovoltaic unit in the photovoltaic array in real time and identify abnormal photovoltaic units whose output power is lower than the preset threshold; Establishing a multi-source data fusion model for the photovoltaic system, wherein the multi-source data includes historical power generation data, weather forecast data, and grid load demand data; Predicting power generation efficiency in a future time period according to the multi-source data fusion model, and generating a corresponding dynamic power allocation strategy; Based on the dynamic power allocation strategy, adjusting the charging and discharging parameters of the energy storage device and the output power of the grid-connected inverter; Perform local shading detection on abnormal PV cells, determine the type of shading source, and trigger adaptive cleaning or fault isolation operations based on the type of shading source; Construct a virtual mapping model of the photovoltaic system to simulate the energy conversion efficiency under different control strategies in real time; According to the simulation results of the virtual mapping model, the dynamic adjustment parameters of the photovoltaic panel and the charge and discharge curve of the energy storage device are optimized; When it is detected that the grid load fluctuation exceeds the preset range, it switches to the backup energy storage mode and prioritizes balancing power output through energy storage equipment.

2. A control method for photovoltaic energy generation according to claim 1, characterized in that: The method further comprises: Acquire real-time environmental data of the photovoltaic system, the environmental data including light intensity, ambient temperature and photovoltaic panel surface temperature; Based on the environmental data, dynamically adjust the tilt angle and orientation of the photovoltaic panel so that the photovoltaic panel receives maximum effective light; the dynamically adjusting the tilt angle and orientation of the photovoltaic panel includes the following steps: Divide the light intensity data into multiple gradient intervals, each gradient interval corresponds to a preset tilt angle adjustment range; According to the difference between the ambient temperature and the surface temperature of the photovoltaic panel, the tilt angle adjustment range is corrected; Dynamic adjustment instructions are generated based on the gradient range and temperature difference, and the angle and direction of the photovoltaic panel are controlled by the servo mechanism.

3. A control method for photovoltaic energy generation according to claim 2, characterized in that: The method of identifying an abnormal photovoltaic unit whose output power is lower than a preset threshold comprises the following steps: Periodically collect the output current and voltage of each photovoltaic unit and calculate the instantaneous power; Compare the instantaneous power with the power of adjacent photovoltaic units horizontally, and select candidate abnormal units whose deviation rate exceeds a first threshold; A power trend analysis is performed on the candidate abnormal unit for a continuous period of time. If the power continues to decrease and the slope exceeds a second threshold, it is determined to be an abnormal photovoltaic unit.

4. A control method for photovoltaic energy generation according to claim 3, characterized in that: The steps to establish a multi-source data fusion model include: The historical power generation data is segmented according to the time series, and the power generation efficiency characteristics of each time period are extracted; Temporally and spatially correlating the meteorological forecast data with the power generation efficiency characteristics to generate a forecast weight matrix; Combined with the priority labels of power grid load demand data, a multi-dimensional data fusion framework is constructed.

5. A control method for photovoltaic energy generation according to claim 4, characterized in that: The steps to generate a dynamic power allocation strategy include: Divide the power generation efficiency levels according to the prediction weight matrix, and assign corresponding energy storage equipment charging and discharging priorities to each level; Dynamically adjust the charging and discharging priority based on real-time changes in grid load demand data; The power allocation sequence in the future time period is optimized by a sliding window algorithm.

6. A control method for photovoltaic energy generation according to claim 5, characterized in that: The step of adjusting the charge and discharge parameters of the energy storage device comprises the following steps: Calculate the maximum adjustable power range based on the remaining capacity and charging and discharging efficiency of the current energy storage device; Inserting a buffer zone in the power allocation sequence to smooth power mutations during charging and discharging; The optimal charge and discharge rate is determined through binary search method to minimize the life loss of energy storage equipment.

7. A control method for photovoltaic energy generation according to claim 6, characterized in that: The local occlusion detection comprises the following steps: Use infrared thermal imagers to collect surface temperature distribution data of abnormal photovoltaic units; According to the gradient change direction of the temperature distribution data, it is determined whether the shielding source is a static object or a dynamic interference; If it is a static object, the robotic arm is triggered to perform cleaning operations; if it is a dynamic interference, the backup circuit is activated to bypass the abnormal unit.

8. A control method for photovoltaic energy generation according to claim 7, characterized in that: The steps to build a virtual mapping model include: Generate three-dimensional energy distribution map based on the physical parameters of photovoltaic panels and environmental data; Embed the time dimension in the energy distribution diagram to simulate the changes in lighting conditions at different time periods; The control parameters are randomly sampled by the Monte Carlo method to evaluate the system efficiency under various parameter combinations.

9. A control method for photovoltaic energy generation according to claim 8, characterized in that: The steps for optimizing the dynamic adjustment parameters of photovoltaic panels include: Extract the top N parameter combinations with the highest efficiency in the virtual mapping model, where N is a preset positive integer constant; Matching the parameter combination with the real-time environmental data to select the combination with the highest degree of adaptability; The control logic of the servo mechanism is updated through an incremental learning algorithm.

10. A control method for photovoltaic energy generation according to claim 9, characterized in that: The switching to the backup energy storage mode comprises the following steps: Monitor the fluctuation frequency and amplitude of power grid load in real time and calculate the fluctuation index; If the fluctuation index exceeds a third threshold, the direct output link of the grid-connected inverter is cut off; The direct current of the energy storage device is converted into alternating current matching the grid through a bidirectional converter and injected into the grid.

Images