Photovoltaic energy storage inverter power coordination control method based on adaptive algorithm

Abstract

The application provides a photovoltaic energy storage inverter power coordination control method based on an adaptive algorithm, and relates to the technical field of data processing.The method comprises the following steps: collecting output power data of a photovoltaic module, energy storage device state data and load demand data of a grid-connected point, and performing normalization processing on the data; calculating preliminary power distribution parameters in a current period; based on load demand data and output power data in historical periods, a load change trend factor and a power fluctuation trend factor are respectively calculated, and the two factors are coupled to form a trend coupling factor; adaptive correction is performed through the trend coupling factor; target power distribution instructions and energy storage charging and discharging adjustment instructions are generated by combining real-time grid-connected point load demand data and energy storage device state data, and are respectively sent to an inverter control unit and an energy storage device control unit; and the application improves the accuracy of the photovoltaic energy storage inverter power coordination control method.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a power coordination control method for photovoltaic energy storage inverters based on adaptive algorithms. Background Technology

[0002] Currently, the mainstream power coordination control methods for photovoltaic energy storage inverters typically employ fixed-weight allocation, fuzzy control, or prediction-based scheduling strategies. These methods dynamically adjust the inverter's output power by real-time monitoring of photovoltaic output, energy storage status, and load demand, combined with a preset control model, to ensure the stability of the system's grid connection and power quality. When the fluctuations in sunlight and load are small, these methods can achieve relatively stable power distribution.

[0003] However, in scenarios with large-scale distributed photovoltaic (PV) grid connections and multiple parallel energy storage systems, numerous controllable parameters exhibit significant fluctuations. Taking predictive control as an example, in an industrial park experiencing peak load concurrency and sudden cloud cover, existing control methods are prone to energy storage response delays due to parameter update lags and insufficient model generalization. This can lead to instantaneous deviations of the inverter output from the setpoint, or even trigger protective shutdowns of the inverter. Summary of the Invention

[0004] The purpose of this invention is to provide a power coordination control method for photovoltaic energy storage inverters based on adaptive algorithms, which aims to solve the problems mentioned in the background art.

[0005] To solve the above-mentioned technical problems, the technical solution of the present invention is as follows:

[0006] A power coordination control method for photovoltaic energy storage inverters based on adaptive algorithms, the method comprising:

[0007] Collect the output power data of photovoltaic modules, the status data of energy storage devices, and the load demand data of grid connection points, and normalize them to obtain the basic data for calculating allocation parameters;

[0008] Based on the basic data, the preliminary power allocation parameters for the current cycle are calculated.

[0009] Based on historical load demand data and output power data, load change trend factor and power fluctuation trend factor are calculated respectively, and the two are coupled to form a trend coupling factor.

[0010] The initial power allocation parameters are adaptively corrected by the trend coupling factor to generate the corrected target power allocation parameters.

[0011] Based on the target power allocation parameters, combined with real-time grid connection point load demand data and energy storage device status data, target power allocation instructions and energy storage charging and discharging adjustment instructions are generated and sent to the inverter control unit and energy storage device control unit, respectively.

[0012] Preferably, based on the basic data, the preliminary power allocation parameters for the current cycle are calculated, including:

[0013] Based on the basic data, the mean of each data item and its corresponding data item in the previous period is differentially processed to obtain the difference value dataset.

[0014] Based on the difference value dataset, the difference values ​​of each data item are weighted and superimposed according to the preset data allocation weight coefficients to obtain a weighted processing value dataset;

[0015] The weighted values ​​are added together to form a composite parameter, and the theoretical load demand is inversely calculated using the composite parameter. The difference between the composite parameter and the actual load demand is then calculated to obtain the power distribution deviation.

[0016] Based on the power allocation deviation, the total power to be allocated is distributed to each power source according to the proportion of the absolute value of the power allocation deviation, thus obtaining their respective preliminary power allocation parameters.

[0017] Preferably, based on historical periodic load demand data and output power data, a load change trend factor and a power fluctuation trend factor are calculated respectively, and the two are coupled to form a trend coupling factor, including:

[0018] Perform time series analysis on a set amount of historical periodic load demand data, calculate the rate of change between load demand data of every two adjacent periods, and average all the rates of change to obtain the mean rate of change.

[0019] The historical cycle load demand data is fitted with a polynomial, the slope of the fitted curve is extracted, and the slope value and the mean rate of change are weighted and superimposed according to the preset trend factor to form the load change trend factor.

[0020] The historical periodic output power data is processed step by step using a sliding window. Within each window, the variance and range of the output power are calculated and multiplied by preset variance weighting coefficients and range weighting coefficients, respectively. The weighted variance value and range value are added together to obtain the initial value of the power fluctuation trend factor.

[0021] The initial value of the power fluctuation trend factor is analyzed for directionality. The duration of the fluctuation rise or fall is statistically analyzed. Based on the correlation between the duration and the initial value of the power fluctuation trend factor, preset duration weight coefficients and initial value weight coefficients are assigned respectively. The two are weighted and superimposed to form the power fluctuation trend factor.

[0022] The load change trend factor and the power fluctuation trend factor are weighted and summed according to the preset coupling weight coefficient. If the two are in the same direction, the coupling weight is increased; if they are in opposite directions, the coupling weight is decreased, thus generating a trend coupling factor.

[0023] Preferably, the initial power allocation parameters are adaptively corrected using a trend coupling factor to generate the corrected target power allocation parameters, including:

[0024] Determine whether the magnitude of the trend coupling factor exceeds the preset trend coupling threshold;

[0025] When the value exceeds the limit, the trend coupling factor is multiplied by the initial power allocation parameter item by item to obtain the first correction value of each power source allocation parameter.

[0026] Based on the first correction value, combined with the current cycle load demand variation and the available capacity of the energy storage device, the two are weighted and superimposed to each first correction value according to the preset feedback adjustment weight coefficient to obtain the second correction value.

[0027] After summing all the second correction values, normalization is performed to ensure that the sum of all allocation parameters equals the current total allocable power, thus generating the target power allocation parameters after multi-level adaptive correction and normalization.

[0028] Preferably, based on the target power allocation parameters and combined with real-time grid connection point load demand data and energy storage device status data, a target power allocation command and an energy storage charge / discharge regulation command are generated, including:

[0029] The system acquires the load demand data of the grid connection point in real time for the current period, calculates the difference between the target power allocation parameters and the load demand data, and allocates the difference to each power source according to the allocation ratio of the allocation parameters to form the target power allocation instruction.

[0030] Simultaneously monitor the current remaining power, upper limit of charging and discharging power, and safe temperature range of the energy storage device. Combine the target power allocation parameters and load demand margin to determine whether the energy storage device needs to be charged or discharged in the current cycle, and set the charging and discharging power value accordingly to generate energy storage charging and discharging adjustment commands.

[0031] The target power allocation command is sent to the inverter control unit, and the energy storage charging and discharging regulation command is sent to the energy storage device control unit, driving the relevant equipment to work together according to their respective commands;

[0032] The inverter output power and the actual charging and discharging status of the energy storage device are monitored and compared with the target command in real time. If a deviation occurs, the system will automatically provide feedback and return to the initial power allocation parameters for recalculation, so as to achieve continuous closed-loop adaptive optimization.

[0033] Preferably, the historical periodic load demand data is subjected to polynomial fitting, the slope of the fitted curve is extracted, and the slope value and the mean rate of change are weighted and superimposed according to a preset trend factor to form a load change trend factor, including:

[0034] Based on historical periodic load demand data, each data item is arranged in chronological order to form a time series sample;

[0035] A multinomial fitting algorithm is applied to the time series samples to obtain a curve reflecting the trend of load change, and the slope value of the curve in the latest period is extracted.

[0036] Based on the preset trend factor, the weight coefficients are synthesized. The slope value and the mean rate of change are multiplied by their respective weight coefficients, and the two multiplications are accumulated and added together to generate the load change trend factor for the current period.

[0037] Preferably, the initial value of the power fluctuation trend factor is subjected to directional analysis, the duration of the fluctuation increase or decrease is statistically analyzed, and based on the correlation between the duration and the initial value of the power fluctuation trend factor, preset duration weight coefficients and initial value weight coefficients are assigned respectively. The two are then weighted and superimposed to form the power fluctuation trend factor, including:

[0038] Based on the initial value of the power fluctuation trend factor, the power output data of the historical cycle is compared with the data cycle by cycle to identify and count the number of cycles with continuous upward or downward changes in the same direction, and the length of the continuous cycle is obtained.

[0039] Based on the duration period length and the initial value of the power fluctuation trend factor, and according to their influence in previous adjustments, the duration period weight coefficient and the initial value weight coefficient are set.

[0040] Multiply the duration of the period by the duration weight coefficient, multiply the initial value of the power fluctuation trend factor by the initial value weight coefficient, and add the two weighted results to obtain the power fluctuation trend factor for the current period.

[0041] Preferably, based on the first correction value, and combining the current cycle load demand variation and the available capacity of the energy storage device, the two are weighted and superimposed onto each of the first correction values ​​according to a preset feedback adjustment weighting coefficient to obtain the second correction value, including:

[0042] The current period load demand variation and the available capacity of the energy storage device are obtained, and feedback adjustment weight coefficients are assigned to each of them. The feedback adjustment weight coefficients are dynamically adjusted according to the historical contribution of their respective parameters to the stability of power distribution.

[0043] The weighted feedback value is obtained by multiplying the change in load demand by its feedback adjustment weighting coefficient and by multiplying the available capacity of the energy storage device by its feedback adjustment weighting coefficient.

[0044] Each first correction value is added to the corresponding weighted feedback value to obtain the second correction value for each power source in the current cycle.

[0045] Preferably, the load demand data of the grid connection point for the current period is acquired in real time, the difference between the target power allocation parameters and the load demand data is calculated, and the difference is allocated to each power source according to the allocation ratio of the allocation parameters to form a target power allocation instruction, including:

[0046] The load demand data of the grid connection point in the current period is obtained in real time, and it is compared with the target power allocation parameters of the current period item by item. The difference between the theoretical output power of each power source and the actual load demand is calculated to obtain the power difference that needs to be adjusted for each power source.

[0047] The power differences of each power source are summed to obtain the total power difference.

[0048] Based on the proportion of each power source in the target power allocation parameters, the total power difference is allocated to each power source proportionally to form their respective allocation adjustment amounts;

[0049] By combining the historical output data and dynamic response capabilities of each power source in the previous cycle, the correction coefficient of the allocation adjustment amount for each power source is determined, and the corrected allocation adjustment amount is added to its target power allocation parameter to obtain the final target output value.

[0050] The final target output value is checked for safety. If it exceeds the maximum or minimum power range allowed by the device, it is automatically corrected to the boundary value allowed by the device.

[0051] The final target output value after safety verification is used as the target power allocation command for the current cycle.

[0052] Preferably, the system simultaneously monitors the remaining power, upper limit of charging / discharging power, and safe temperature range of the energy storage device. Combining this with target power allocation parameters and load demand margin, it determines whether the energy storage device needs to charge or discharge in the current cycle, and accordingly sets the charging / discharging power value, generating an energy storage charging / discharging regulation command, including:

[0053] Real-time monitoring of the remaining power, upper limit of charging and discharging power, and actual temperature of the energy storage device in the current cycle forms an energy storage status data input set;

[0054] By comprehensively analyzing the energy storage status data input set with the target power allocation parameters and the load demand margin of the current cycle, it is determined whether the energy storage device should adopt charging mode or discharging mode in the current cycle, and the operation mode determination result is obtained.

[0055] Based on the operation mode determination result, determine the direction of charging or discharging required by the energy storage device in the current cycle, and set the actual executable charging and discharging power value with the remaining power, the upper limit of the equipment's allowable power and the temperature safety range as constraints.

[0056] The charging and discharging power values ​​are checked for safety and rationality against the target power allocation parameters and actual load requirements. If the values ​​exceed the equipment limits or there are safety hazards, they are automatically adjusted to the maximum or minimum values ​​within the allowable range of the equipment, generating corrected charging and discharging power values.

[0057] Based on the revised charge and discharge power values, the energy storage charge and discharge regulation command for the current cycle is generated.

[0058] The above-described solution of the present invention has at least the following beneficial effects:

[0059] Through adaptive algorithms, the system can dynamically reflect the actual operating status of each power source based on real-time normalization of photovoltaic module output power data, energy storage device status data, and grid connection point load demand data, providing a unified and standardized calculation basis for power allocation. Compared to existing schemes using fixed weights or static control models, this invention can not only adapt to the complex changes in multi-source data under different operating conditions, but also enhance the algorithm's sensitivity and response capability to extreme loads or light fluctuation scenarios.

[0060] By combining historical cycle load demand and output power trend factor analysis, this invention utilizes adaptive correction and trend coupling mechanisms to enable target power allocation parameters to actively track actual load and power fluctuations, achieving dynamic optimization and adjustment of the output of each power source. In situations involving large-scale distributed photovoltaic integration, multi-point parallel energy storage, and rapid changes in grid load, it effectively alleviates the energy storage response delay problem caused by model lag and insufficient generalization ability, reducing the risk of inverter protective shutdowns due to power deviations from set values. For example, during peak energy consumption in industrial parks and when sudden cloud cover occurs, it can promptly correct power allocation strategies, ensuring stable inverter output and significantly improving the safety and reliability of grid operation.

[0061] Furthermore, the adaptive multi-level closed-loop adjustment mechanism employed in this invention can adjust power allocation commands and energy storage charging and discharging regulation commands in real time based on current operational feedback, achieving dynamic adaptation and closed-loop optimization throughout the entire process. This mechanism not only improves the accuracy of power allocation in complex scenarios but also significantly enhances the photovoltaic energy storage inverter's resistance to shocks and its stable operation under extreme disturbances. Attached Figure Description

[0062] Figure 1 This is a flowchart of a photovoltaic energy storage inverter power coordination control method based on an adaptive algorithm provided in an embodiment of the present invention. Detailed Implementation

[0063] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0064] like Figure 1 As shown, embodiments of the present invention propose a power coordination control method for photovoltaic energy storage inverters based on an adaptive algorithm, the method comprising:

[0065] Collect the output power data of photovoltaic modules, the status data of energy storage devices, and the load demand data of grid connection points, and normalize them to obtain the basic data for calculating allocation parameters;

[0066] Based on the basic data, the preliminary power allocation parameters for the current cycle are calculated.

[0067] Based on historical load demand data and output power data, load change trend factor and power fluctuation trend factor are calculated respectively, and the two are coupled to form a trend coupling factor.

[0068] The initial power allocation parameters are adaptively corrected by the trend coupling factor to generate the corrected target power allocation parameters.

[0069] Based on the target power allocation parameters, combined with real-time grid connection point load demand data and energy storage device status data, target power allocation instructions and energy storage charging and discharging adjustment instructions are generated and sent to the inverter control unit and energy storage device control unit, respectively.

[0070] In this embodiment of the invention, by collecting output power data of photovoltaic modules, status data of energy storage devices, and load demand data of grid-connected points, and normalizing them, a unified dimensional basis can be provided for subsequent power allocation calculations. This eliminates uncertainties caused by differences in units and amplitudes between different data sources, ensuring that various types of data can be efficiently integrated and processed within the same calculation system. The system utilizes the basic data, combined with historical period load demand and output power data, to extract load change trend factors and power fluctuation trend factors, respectively. Through a trend coupling mechanism, it dynamically reflects the combined impact of load changes and power fluctuations within the current period, effectively improving the perception and adaptive capabilities to complex grid conditions.

[0071] By adaptively correcting the initial power allocation parameters using a trend coupling factor, dynamic adjustment of the coordinated output of multiple power sources can be achieved. This enables the target power allocation parameters to have greater flexibility and response speed when facing scenarios such as large-scale distributed photovoltaic access, multi-point parallel connection of energy storage systems, and frequent load fluctuations. For example, in situations such as a sudden increase in load or a sudden drop in sunlight in an industrial park, the system can automatically adjust the coordinated output ratio of the inverter and energy storage device, significantly reducing the risk of output deviation and protective shutdown, and improving the operational safety and reliability of the system.

[0072] Furthermore, by combining target power allocation parameters with real-time grid-connected load demand data and energy storage device status data, target power allocation commands and energy storage charging and discharging adjustment commands can be generated and issued, enabling closed-loop dynamic optimization throughout the entire process. With real-time monitoring and adaptive correction, the entire system can continuously optimize the output of each power source, significantly improving the accuracy of power allocation and enhancing the overall utilization efficiency and disturbance resistance of the new energy system.

[0073] This involves collecting output power data from photovoltaic modules, status data from energy storage devices, and load demand data from grid connection points, and then normalizing these data to obtain the basic data for calculating allocation parameters. Specifically, this includes:

[0074] In actual system operation, power acquisition sensors acquire output power data of photovoltaic modules in each sampling cycle, status monitoring units acquire status data such as remaining power and charging / discharging status of energy storage devices, and power consumption monitoring devices acquire real-time load demand data of grid-connected points. For different types and physical quantities of data, to facilitate subsequent unified modeling and analysis, all collected data are first normalized according to preset benchmarks such as maximum, minimum, or historical average values, converting each data item into a dimensionless standard value. For example, the method of subtracting the minimum historical value of each data item from each acquired value and then dividing by its range can be used to normalize all data to between 0 and 1. After normalization, a basic dataset containing normalized values ​​from multiple sources, including photovoltaic modules, energy storage devices, and grid-connected point loads, is obtained, providing a unified and comparable input data foundation for subsequent power allocation parameter calculations.

[0075] The inverter control unit specifically includes:

[0076] The inverter control unit receives target power allocation commands from the system's host computer or dispatch center and adjusts the inverter's output power in real time according to the commands. This unit includes a digital signal processing module, an inverter main control board, and communication interfaces. In actual implementation, the inverter control unit first receives the target power allocation command via a communication interface (such as CAN, Modbus, Ethernet, etc.). Upon receiving the command, the digital signal processing module automatically adjusts the conduction state of the inverter bridge arms and the PWM control signal based on the target output power specified in the command, combined with currently acquired grid parameters (such as voltage, current, frequency, etc.), to achieve dynamic energy flow between the DC and AC sides of the inverter.

[0077] To ensure the accuracy and safety of command execution, the inverter control unit also monitors the synchronization status of the output power with the grid in real time. If a deviation is found between the actual output power and the command, or if the system operating parameters exceed the safety threshold, the output strategy is automatically corrected through the feedback loop, or an alarm message is issued to the system. The entire control process requires no manual intervention, enabling the inverter to respond quickly, stably, and in a closed-loop manner to system scheduling commands.

[0078] The energy storage device control unit specifically includes:

[0079] The energy storage device control unit receives energy storage charge and discharge regulation commands and precisely controls the charging and discharging process and power of the energy storage device according to the commands. This unit consists of a Battery Management System (BMS), a power regulation module, and a communication interface. During implementation, the energy storage device control unit receives charging and discharging power and operating mode (charging or discharging) commands from the dispatch center via wired or wireless communication. Upon receiving the command, the BMS module first determines the current energy storage unit's power level, temperature, and safety range to confirm that the device is within an adjustable range. If all conditions are met, the power regulation module switches the energy storage device to the corresponding charging or discharging mode according to the command and precisely adjusts the actual charging and discharging power to ensure that it does not exceed the device's maximum allowable power and maximum / minimum SOC limits.

[0080] During operation, the energy storage device control unit continuously monitors parameters such as current, voltage, and temperature of the battery pack. If a significant deviation is detected between the actual charging / discharging process and the adjustment command, or if a parameter exceeds the safe range, the system will automatically trigger a protection mechanism to adjust or terminate the charging / discharging process and feed the abnormal information back to the upper-level system, thereby achieving dynamic closed-loop management and multi-level protection of the energy storage device's charging / discharging process.

[0081] In a preferred embodiment of the present invention, preliminary power allocation parameters for the current cycle are calculated based on basic data, including:

[0082] Based on the basic data, the mean of each data item and its corresponding data item in the previous period is differentially processed to obtain the difference value dataset.

[0083] Based on the difference value dataset, the difference values ​​of each data item are weighted and superimposed according to the preset data allocation weight coefficients to obtain a weighted processing value dataset;

[0084] The weighted values ​​are added together to form a composite parameter, and the theoretical load demand is inversely calculated using the composite parameter. The difference between the composite parameter and the actual load demand is then calculated to obtain the power distribution deviation.

[0085] Based on the power allocation deviation, the total power to be allocated is distributed to each power source according to the proportion of the absolute value of the power allocation deviation, thus obtaining their respective preliminary power allocation parameters.

[0086] In this embodiment of the invention, preliminary power allocation parameters for the current cycle are calculated based on basic data, enabling dynamic allocation optimization of different power sources. By differentially processing the basic data with the average data of the previous cycle, the changing trends of each power source can be reflected in real time. Furthermore, the differential results are weighted using preset data allocation weighting coefficients, fully considering the power allocation contribution of various parameters in historical cycles, resulting in more scientifically weighted values. These weighted values ​​are used to infer theoretical load demand and compare it with actual load demand, thereby obtaining a power allocation deviation that more closely reflects actual operating conditions. Subsequently, the total power is allocated according to the proportion of the absolute value of the power allocation deviation, allowing the preliminary allocation parameters of each power source to flexibly reflect the current supply and demand relationship of the system. For example, if the state data of the energy storage device is relatively favorable in a certain cycle, its allocation parameters are automatically increased, improving the power supply elasticity of the entire system.

[0087] The preset data allocation weighting coefficients specifically include:

[0088] To reasonably reflect the influence of each data item (such as photovoltaic output power, energy storage status, and grid-connected load) in power allocation, the system assigns a weighting coefficient to each data item. This weighting coefficient can be initially set based on historical data analysis, system structure design, or practical operating experience during the initial system deployment phase. For example, energy storage units with fast response speeds and high energy density can be assigned a higher weight, while photovoltaic output, which is more volatile and susceptible to external influences, can be assigned a lower weight. During subsequent operation, the weighting coefficients can be dynamically adjusted based on indicators such as historical power allocation errors and the difference between target and actual responses, ensuring that each weight better reflects the contribution of the current system operating status to allocation accuracy. Ultimately, each weighting coefficient is used to weight each data item in the differential value dataset, ensuring that the allocation result takes into account both the actual capacity and current status of each power source.

[0089] Among these methods, the theoretical load demand is inversely calculated using synthesized parameters, and the difference between this and the actual load demand is calculated to obtain the power distribution deviation, which specifically includes:

[0090] After weighting and summing each normalized data item, the system obtains the composite parameter for the current period. This parameter can be understood as the theoretical combined output capacity of each power source after weighting. Subsequently, the system compares the theoretical load demand corresponding to this composite parameter with the actual grid-connected load demand data monitored in this period using a reverse calculation method. Specifically, the composite parameter is normalized and inversely transformed back to the physical magnitude of the theoretical load demand. Then, the difference between the actual load demand value and the theoretical value is processed to obtain the power allocation deviation, which reflects the difference between actual demand and theoretical supply. This deviation is used to guide the priority and adjustment range of subsequent total power allocation among the power sources, thereby dynamically balancing the matching degree between the output of each source and the actual energy consumption of the system.

[0091] The total power to be allocated specifically includes:

[0092] The total power to be allocated refers to the maximum effective power that the system can allocate or distribute within the current cycle, based on load demand and the actual available capacity of each power source. In practice, the total available power budget for the current cycle is first calculated based on the actual load demand and the maximum available output power of each power source (such as photovoltaics and energy storage). It is important to note that the total power to be allocated cannot exceed the total system load demand, nor can it exceed the physical limits of the output capacity of each power source. For example, when load demand is low and energy storage is sufficient, the total power to be allocated is determined by the load; while during peak load periods and when some power sources are current-limited, the total power to be allocated is determined by the maximum capacity of the power sources, ensuring allocation safety and system stability.

[0093] In a preferred embodiment of the present invention, based on historical periodic load demand data and output power data, a load change trend factor and a power fluctuation trend factor are calculated respectively, and the two are coupled to form a trend coupling factor, including:

[0094] Perform time series analysis on a set amount of historical periodic load demand data, calculate the rate of change between load demand data of every two adjacent periods, and average all the rates of change to obtain the mean rate of change.

[0095] The historical cycle load demand data is fitted with a polynomial, the slope of the fitted curve is extracted, and the slope value and the mean rate of change are weighted and superimposed according to the preset trend factor to form the load change trend factor.

[0096] The historical periodic output power data is processed step by step using a sliding window. Within each window, the variance and range of the output power are calculated and multiplied by preset variance weighting coefficients and range weighting coefficients, respectively. The weighted variance value and range value are added together to obtain the initial value of the power fluctuation trend factor.

[0097] The initial value of the power fluctuation trend factor is analyzed for directionality. The duration of the fluctuation rise or fall is statistically analyzed. Based on the correlation between the duration and the initial value of the power fluctuation trend factor, preset duration weight coefficients and initial value weight coefficients are assigned respectively. The two are weighted and superimposed to form the power fluctuation trend factor.

[0098] The load change trend factor and the power fluctuation trend factor are weighted and summed according to the preset coupling weight coefficient. If the two are in the same direction, the coupling weight is increased; if they are in opposite directions, the coupling weight is decreased, thus generating a trend coupling factor.

[0099] In this embodiment of the invention, load change trend factors and power fluctuation trend factors are calculated based on historical periodic load demand data and output power data, respectively. These two factors are then coupled to form a trend coupling factor, which significantly enhances the system's adaptive perception and response capabilities to load changes and power fluctuations. Specifically, the load change trend factor is extracted using historical data time series analysis and polynomial fitting, which can promptly capture the overall direction and rate of load change. Simultaneously, the power fluctuation trend factor is extracted using statistical methods such as sliding windows, effectively reflecting the short-term fluctuation characteristics of output power. After weighted coupling, a trend coupling factor that considers both global and local changes is formed, serving as an important reference for subsequent adaptive correction. For example, if historical data shows a trend of continuously increasing load and intensified power fluctuations within a certain period, the trend coupling factor will amplify the correction weight, making the allocation scheme more robust and effectively reducing practical risks such as inverter overload and frequent switching of energy storage devices.

[0100] This includes performing time series analysis on a set amount of historical periodic load demand data to calculate the rate of change between load demand data from two adjacent periods, specifically including:

[0101] First, set up a historical analysis window in the system and determine the number of historical periods to be statistically analyzed, such as the last ten sampling periods. Arrange the load demand data within these periods in chronological order to obtain a set of continuous load demand data points. Next, starting from the second data point, subtract the data of the previous period from the data of the current period, and then ratio the difference with the data of the previous period to obtain the rate of change between each pair of adjacent period load demand data. Finally, compile all the rates of change into a sequence to provide basic data for subsequent trend analysis and factor extraction. For example, if the load data for the last five periods are A, B, C, D, and E, then when calculating the rate of change, subtract A from B and divide by A, subtract B from C and divide by B, and so on, to obtain four sets of rate of change results.

[0102] The preset trend factor synthesis weight coefficients specifically include:

[0103] The trend factor synthesis weighting coefficient is used to weight and superimpose trend indicators from different sources (such as the slope of the fitted curve, the mean rate of change, etc.) according to their importance in the overall trend judgment. During the initial system deployment, the weighting coefficient for each trend indicator can be empirically set based on historical operating conditions and data fluctuation characteristics. For example, for scenarios where the load changes relatively steadily over a long period but experiences occasional large fluctuations, a higher weight can be assigned to the slope of the fitted curve, while a lower weight can be assigned to the mean rate of change. During operation, these weighting coefficients can be dynamically adjusted based on the actual prediction error and response effect of the system to more accurately reflect the contribution of trend factors to power allocation optimization. Ultimately, the role of the trend factor synthesis weighting coefficient is to reasonably integrate the influence of each individual trend parameter into a single overall trend judgment result.

[0104] The preset variance weighting coefficient and range weighting coefficient specifically include:

[0105] The variance weighting coefficient and range weighting coefficient are primarily used for multi-dimensional evaluation of the fluctuation characteristics in historical periodic output power data. The system first performs sliding window processing on the historical periodic output power data within a set window, calculating the variance (reflecting the degree of dispersion of fluctuations) and range (reflecting the fluctuation range between the maximum and minimum values) of the output power within each window. The variance weighting coefficient measures the contribution of data dispersion in judging power fluctuation trends; the range weighting coefficient measures the influence of the maximum fluctuation amplitude. Both weights can be configured by the system administrator or through an adaptive algorithm, depending on the system's sensitivity to stationarity or extreme fluctuations. By multiplying the variance and range of each window by their corresponding weights and then summing them, a more comprehensive and objective foundation of power fluctuation trend factors is formed. For example, if the system is particularly concerned about sudden large fluctuations, a higher value can be assigned to the range weight.

[0106] The preset coupling weight coefficients specifically include:

[0107] The coupling weight coefficient is used to weight and synthesize the load change trend factor and the power fluctuation trend factor to form a more comprehensive trend coupling factor. During the system setup phase, the coupling weight coefficient can be set to a base value based on actual operating experience and the performance of each factor in system power allocation optimization. For example, for scenarios with high volatility but slow load changes, a higher coupling weight can be assigned to the power fluctuation trend factor. The coupling weight coefficient can also dynamically change with the system's operating state. For instance, when the two trend factors change in the same direction, the coupling weight is automatically increased to highlight the system's rapid response to drastic changes; conversely, when they change in opposite directions, the coupling weight is appropriately decreased to prevent over-correction. Through this dynamic adjustment, the system can be ensured to respond agilely to drastic changes while avoiding unnecessary adjustment interference, thus improving the overall stability and reliability of power allocation.

[0108] In a preferred embodiment of the present invention, the initial power allocation parameters are adaptively corrected using a trend coupling factor to generate corrected target power allocation parameters, including:

[0109] Determine whether the magnitude of the trend coupling factor exceeds the preset trend coupling threshold;

[0110] When the value exceeds the limit, the trend coupling factor is multiplied by the initial power allocation parameter item by item to obtain the first correction value of each power source allocation parameter.

[0111] Based on the first correction value, combined with the current cycle load demand variation and the available capacity of the energy storage device, the two are weighted and superimposed to each first correction value according to the preset feedback adjustment weight coefficient to obtain the second correction value.

[0112] After summing all the second correction values, normalization is performed to ensure that the sum of all allocation parameters equals the current total allocable power, thus generating the target power allocation parameters after multi-level adaptive correction and normalization.

[0113] In this embodiment of the invention, the initial power allocation parameters are adaptively corrected using a trend coupling factor. This enables the power allocation process to dynamically adapt to complex fluctuations in grid load and multi-source inputs such as photovoltaics and energy storage, achieving more refined and robust power coordination control. When the trend coupling factor exceeds a preset threshold, the system can automatically identify significant changes within the current cycle and initiate the adaptive correction process promptly without manual intervention. After adjustment by the trend coupling factor, the power source allocation parameters are further weighted by combining the current cycle's load demand changes and the available capacity of the energy storage device, and finally normalized to ensure a reasonable total power output. This significantly improves the inverter's response speed and stability in high-frequency fluctuation scenarios. For example, in scenarios with sudden load increases and rapid changes in sunlight, this embodiment can promptly adjust the output of each power source to prevent system instability caused by excessive fluctuations in a single source.

[0114] The preset trend coupling threshold specifically includes:

[0115] In practical engineering applications, before the system performs adaptive power allocation correction, a trend coupling threshold must first be set. This threshold is used to determine whether the trend coupling factor in the current period has reached the condition requiring adaptive correction. The trend coupling threshold can be set by referring to historical operating data, system adjustment experience, or preset according to the system operating environment. For example, the typical fluctuation range of the trend coupling factor over a period of time can be statistically analyzed, and the mean or percentile can be selected as the initial threshold. Alternatively, a maximum allowable variation value can be manually set based on the safety tolerance of actual power fluctuations. During system operation, the threshold can also be dynamically fine-tuned based on the actual response of the equipment to ensure that it always matches the equipment adjustment capability and system safety requirements. In actual operation, when the absolute value of the trend coupling factor is greater than or equal to the threshold, the system automatically triggers subsequent adaptive correction steps. If it is less than the threshold, the current allocation strategy remains unchanged to avoid frequent adjustments by the system to minor changes.

[0116] Specifically, the trend coupling factor is multiplied term by term by the initial power allocation parameters to obtain the first correction value for each power source allocation parameter, which includes:

[0117] When the system determines that the trend coupling factor has reached or exceeded a preset threshold, it enters the adaptive correction phase. At this point, the system uses the initial power allocation parameters of each power source as input and multiplies them sequentially with the trend coupling factor of the same period. In other words, for each power source, its initial allocation parameters are directly multiplied by the value of the trend coupling factor to obtain the first correction value for that power source. Thus, the changing trend of the trend coupling factor (such as positive amplification or negative weakening) is directly reflected in the allocation correction magnitude of each power source, achieving adaptive adjustment of power allocation. For example, if the trend coupling factor is greater than 1, the correction value of the corresponding power source will increase accordingly; if it is less than 1, the correction value will decrease; when the trend coupling factor is negative, the allocation direction is switched. All first correction values ​​serve as the basis for subsequent multiple corrections and normalization processes, ensuring that the power allocation results closely follow the system's operating trend, achieving more efficient and flexible multi-source coordinated adjustment.

[0118] In a preferred embodiment of the present invention, a target power allocation command and an energy storage charging and discharging adjustment command are generated based on the target power allocation parameters, combined with real-time grid connection point load demand data and energy storage device status data, including:

[0119] The system acquires the load demand data of the grid connection point in real time for the current period, calculates the difference between the target power allocation parameters and the load demand data, and allocates the difference to each power source according to the allocation ratio of the allocation parameters to form the target power allocation instruction.

[0120] Simultaneously monitor the current remaining power, upper limit of charging and discharging power, and safe temperature range of the energy storage device. Combine the target power allocation parameters and load demand margin to determine whether the energy storage device needs to be charged or discharged in the current cycle, and set the charging and discharging power value accordingly to generate energy storage charging and discharging adjustment commands.

[0121] The target power allocation command is sent to the inverter control unit, and the energy storage charging and discharging regulation command is sent to the energy storage device control unit, driving the relevant equipment to work together according to their respective commands;

[0122] The inverter output power and the actual charging and discharging status of the energy storage device are monitored and compared with the target command in real time. If a deviation occurs, the system will automatically provide feedback and return to the initial power allocation parameters for recalculation, so as to achieve continuous closed-loop adaptive optimization.

[0123] In this embodiment of the invention, by performing polynomial fitting on historical periodic load demand data, extracting the slope of the fitted curve, and then weighting and superimposing the slope value and the mean rate of change according to a preset trend factor, a load change trend factor is formed, which can more accurately reflect the dynamic evolution of the overall system load. Polynomial fitting can fully utilize the continuity of historical data changes, effectively combining short-term fluctuations with long-term trends, and extracting the slope to further quantify the rate of load growth or decrease. Then, through weighted synthesis with the mean rate of change, a more comprehensive evaluation of the current load change trend is formed. For example, when the grid load shows a slow increase but occasional sudden changes, the fitted slope can reflect the main trend, while the mean rate of change captures sudden fluctuations. The synthesis of the two can provide more forward-looking parameter support for subsequent power allocation corrections.

[0124] This includes monitoring the inverter's output power and the actual charging and discharging status of the energy storage device, comparing them in real time with the target command. If a deviation occurs, the system automatically provides feedback and returns to the initial power allocation parameters for recalculation, achieving continuous closed-loop adaptive optimization. Specifically, this includes:

[0125] During system operation, a real-time monitoring module is set up to continuously collect the actual output power data of the inverter and the actual charging and discharging status information of the energy storage device. The collected data is compared item by item with the target power allocation command and energy storage charging and discharging adjustment command issued by the system in the current cycle to determine the difference between each actual output or operating status and the command requirements. When it is found that the actual output of a power source or the charging / discharging status of an energy storage device deviates from the target command by more than the set tolerance range, the system immediately inputs the deviation information as a feedback signal to the dispatch control center. After receiving the deviation feedback, the dispatch center automatically triggers the recalculation process of the allocation algorithm, directly returning to the calculation stage of the initial power allocation parameters, and re-completes the entire power allocation process based on the latest collected real-time data, status, and system feedback. This design realizes the continuous closed-loop adaptive optimization of the system, ensuring that the power allocation result can continuously self-correct according to the actual operating status, and improving the operational stability and responsiveness of the entire photovoltaic energy storage inverter system in complex dynamic environments.

[0126] In a preferred embodiment of the present invention, historical periodic load demand data is subjected to polynomial fitting to extract the slope of the fitted curve, and the slope value and the mean rate of change are weighted and superimposed according to a preset trend factor to form a load change trend factor, including:

[0127] Based on historical periodic load demand data, each data item is arranged in chronological order to form a time series sample;

[0128] A multinomial fitting algorithm is applied to the time series samples to obtain a curve reflecting the trend of load change, and the slope value of the curve in the latest period is extracted.

[0129] Based on the preset trend factor, the weight coefficients are synthesized. The slope value and the mean rate of change are multiplied by their respective weight coefficients, and the two multiplications are accumulated and added together to generate the load change trend factor for the current period.

[0130] In this embodiment of the invention, by performing directional analysis on the initial value of the power fluctuation trend factor, the duration of the fluctuation increase or decrease is statistically analyzed. Based on the correlation between the duration and the initial value of the power fluctuation trend factor, preset weight coefficients for the duration and the initial value are assigned respectively. The two are then weighted and superimposed to form the power fluctuation trend factor, which can effectively enhance the system's ability to identify short-term fluctuation trends and continuous fluctuation characteristics. The directional analysis not only considers the amplitude of power fluctuations but also focuses on the impact of fluctuation duration. Through a weight adjustment mechanism, the sensitivity of power allocation under complex fluctuation conditions is improved. For example, when the output power shows a continuous increase in the same direction for multiple cycles in a certain stage, the system can promptly increase the adjustment amplitude of the corresponding power source; while if the fluctuation direction changes frequently, the weight is adjusted to reduce over-response to a single fluctuation signal, thereby enhancing the smoothness and robustness of the overall operation.

[0131] Among these methods, a polynomial fitting algorithm is applied to the time series samples to obtain a curve reflecting the trend of load changes, specifically including:

[0132] When processing historical periodic load demand data, the system first arranges a set number of historical periodic load demand data points in chronological order to form a complete time series sample set. The system selects a fitting order (such as a first-order, second-order, or higher-order polynomial), then uses the time series data as the independent variable and the load demand value as the dependent variable, and models these data as a polynomial function using a minimum-error fitting method. During the fitting process, the system adjusts the coefficients of each order of the polynomial to ensure that the fitted curve fits all historical data points as closely as possible overall. After fitting, a continuous and smooth mathematical curve is obtained, which accurately reflects the overall trend and local variation characteristics of historical load demand. The system can then further extract parameters such as the slope or local rate of change of this curve at the latest period point for subsequent calculation of load trend factors. In this way, polynomial fitting not only improves the ability to capture the evolution of load demand but also provides a solid data foundation and trend criteria for subsequent dynamic optimization of power allocation.

[0133] In a preferred embodiment of the present invention, the initial value of the power fluctuation trend factor is subjected to directional analysis, the duration of the fluctuation increase or decrease is statistically analyzed, and based on the correlation between the duration and the initial value of the power fluctuation trend factor, preset duration weight coefficients and initial value weight coefficients are assigned respectively. The two are then weighted and superimposed to form the power fluctuation trend factor, including:

[0134] Based on the initial value of the power fluctuation trend factor, the power output data of the historical cycle is compared with the data cycle by cycle to identify and count the number of cycles with continuous upward or downward changes in the same direction, and the length of the continuous cycle is obtained.

[0135] Based on the duration period length and the initial value of the power fluctuation trend factor, and according to their influence in previous adjustments, the duration period weight coefficient and the initial value weight coefficient are set.

[0136] Multiply the duration of the period by the duration weight coefficient, multiply the initial value of the power fluctuation trend factor by the initial value weight coefficient, and add the two weighted results to obtain the power fluctuation trend factor for the current period.

[0137] In this embodiment of the invention, by performing directional analysis on the initial value of the power fluctuation trend factor, the duration of the fluctuation increase or decrease is statistically analyzed. Based on the correlation between the duration and the initial value of the power fluctuation trend factor, preset duration weight coefficients and initial value weight coefficients are assigned respectively. The two are then weighted and superimposed to form the power fluctuation trend factor, which can improve the ability to identify short-term fluctuations and continuous changes in power output. The statistical analysis of the duration allows the system to not only focus on the amplitude of the fluctuation, but also to reflect the continuity and periodicity of power changes in a timely manner. This is of great significance for dealing with continuous abnormal load or lighting conditions. In practical applications, when the output power is detected to increase in the same direction for several consecutive cycles, the system can enhance the response speed to the trend by increasing the duration weight coefficient, and realize the forward-looking correction of energy storage charging and discharging or inverter adjustment, thereby reducing frequent small-scale adjustments and resource waste, and improving the overall energy efficiency and service life of the system.

[0138] Specifically, based on the duration period length and the initial value of the power fluctuation trend factor, and considering their influence in past adjustments, the duration period weighting coefficient and the initial value weighting coefficient are set, including:

[0139] When calculating the power fluctuation trend factor, the system first counts the number of consecutive periods of rising or falling changes (i.e., the duration of the period) and the initial value of the power fluctuation trend factor for the current period. To determine the importance of each factor in comprehensively judging the trend, the system reviews and analyzes historical data to assess the impact of the duration of the period and the initial value on the stability and accuracy of the system in actual allocation and adjustment. For example, if historical operating results indicate that the persistence of power fluctuations has a significant impact on allocation accuracy, the duration of the period weight coefficient is set to a higher value; if the initial value of the power fluctuation trend factor can effectively reflect the fluctuation amplitude, the initial value weight coefficient is increased. This assessment can be completed by statistically analyzing indicators such as allocation errors and system fluctuation frequencies corresponding to different coefficient settings under historical operating conditions, or by manually setting the initial value and then having the system adaptively adjust it based on the results. Finally, the duration of the period weight coefficient and the initial value weight coefficient will be used as parameters for the weighted calculation of the fluctuation trend factor for the current period, so that the weighted result can reflect both the impact of the fluctuation persistence and the reference value of the fluctuation amplitude in a single period, thereby improving the accuracy of power allocation trend judgment.

[0140] In a preferred embodiment of the present invention, based on a first correction value, and combining the current periodic load demand variation and the available capacity of the energy storage device, the two are weighted and superimposed onto each of the first correction values ​​according to a preset feedback adjustment weighting coefficient to obtain a second correction value, including:

[0141] The current period load demand variation and the available capacity of the energy storage device are obtained, and feedback adjustment weight coefficients are assigned to each of them. The feedback adjustment weight coefficients are dynamically adjusted according to the historical contribution of their respective parameters to the stability of power distribution.

[0142] The weighted feedback value is obtained by multiplying the change in load demand by its feedback adjustment weighting coefficient and by multiplying the available capacity of the energy storage device by its feedback adjustment weighting coefficient.

[0143] Each first correction value is added to the corresponding weighted feedback value to obtain the second correction value for each power source in the current cycle.

[0144] In this embodiment of the invention, by combining the current periodic load demand fluctuation and the available capacity of the energy storage device with a first correction value, and weighting both according to a preset feedback adjustment weighting coefficient, a second correction value is obtained. This allows the power allocation correction process to fully integrate real-time demand and energy storage capacity, improving the dynamic adaptive level. Specifically, the load demand fluctuation reflects the fluctuation characteristics of actual energy consumption on the user side, while the available capacity of the energy storage device reflects the response redundancy of the current energy storage system. The two influence the final correction result through weighted superposition. Thus, under high demand and sufficient energy storage capacity, the allocation parameters are automatically relaxed, which helps to improve system output; when energy storage is close to its limit, the parameters are tightened to prioritize equipment safety and operational reliability. For example, during peak electricity consumption periods, the system automatically allocates more power to the load side, while during off-peak periods, it prioritizes energy storage, effectively coping with variable operating conditions and extreme events.

[0145] This includes obtaining the current cycle load demand variation and the available capacity of energy storage devices, and assigning feedback adjustment weighting coefficients to both, specifically including:

[0146] Within the current cycle, the system collects and calculates the load demand fluctuation in real time, i.e., the change in load demand value between the current cycle and the previous cycle or a set baseline value. Simultaneously, it monitors the available capacity of the energy storage device, including its available power and dispatchable charging / discharging power margin. For these two key parameters, the system assigns a feedback adjustment weight coefficient. The initial setting of the weight coefficient can be based on equipment characteristics, operating conditions, and historical operating data. For example, if load demand fluctuates greatly and has a significant impact on system operation, the load fluctuation weight coefficient is set to a high value; if the energy storage device capacity is tight or plays a critical role in system stability, its weight coefficient is appropriately increased. In this way, the role of each input parameter in subsequent power allocation corrections can be flexibly adjusted, thereby better adapting to the current actual system operating conditions.

[0147] The feedback adjustment weight coefficients are dynamically adjusted based on their respective parameters to assess the historical contribution of power allocation stability, specifically including:

[0148] To achieve more precise and adaptive power allocation optimization, the system not only adjusts the weighting coefficients for load demand fluctuations and available energy storage capacity allocation during the initial setup phase, but also dynamically optimizes the weighting settings based on the historical contribution of each parameter to improving allocation stability during long-term system operation. Specifically, the system periodically backtracks and analyzes historical operating data to assess the impact of fluctuations in each parameter on key indicators such as power allocation error, output stability, and the number of equipment protection triggers. For example, if historical data shows that increasing the weighting coefficient significantly reduces system allocation error, the weight of that parameter is maintained or further increased; conversely, if increasing the weight of a parameter does not improve allocation performance, the weight value is appropriately reduced. This dynamic adjustment can be achieved through rolling evaluation algorithms, feedback calibration logic, or manual intervention, ensuring that the parameter weights involved in each cycle always reflect their actual contribution to system stability, achieving closed-loop adaptive optimization.

[0149] In a preferred embodiment of the present invention, the load demand data of the grid connection point for the current period is acquired in real time, the difference between the target power allocation parameters and the load demand data is calculated, and the difference is allocated to each power source according to the allocation ratio of the allocation parameters to form a target power allocation instruction, including:

[0150] The load demand data of the grid connection point in the current period is obtained in real time, and it is compared with the target power allocation parameters of the current period item by item. The difference between the theoretical output power of each power source and the actual load demand is calculated to obtain the power difference that needs to be adjusted for each power source.

[0151] The power differences of each power source are summed to obtain the total power difference.

[0152] Based on the proportion of each power source in the target power allocation parameters, the total power difference is allocated to each power source proportionally to form their respective allocation adjustment amounts;

[0153] By combining the historical output data and dynamic response capabilities of each power source in the previous cycle, the correction coefficient of the allocation adjustment amount for each power source is determined, and the corrected allocation adjustment amount is added to its target power allocation parameter to obtain the final target output value.

[0154] The final target output value is checked for safety. If it exceeds the maximum or minimum power range allowed by the device, it is automatically corrected to the boundary value allowed by the device.

[0155] The final target output value after safety verification is used as the target power allocation command for the current cycle.

[0156] In this embodiment of the invention, by acquiring the grid-connected load demand data for the current period in real time and comparing it item by item with the target power allocation parameters, the real-time difference between the theoretical output of each power source and the actual load can be accurately reflected, effectively identifying system deviations. The differences between each power source are summarized and allocated to each power source as needed according to the allocation ratio of the target allocation parameters, making the power adjustment of the entire system more holistic and coordinated. Furthermore, by combining the historical output data and dynamic response capabilities of each power source from the previous period, a correction coefficient for the allocation adjustment amount of each power source is set, realizing dynamic adaptive optimization of the allocation adjustment process. This correction coefficient can flexibly adjust the adjustment range of each power source based on factors such as the actual response of the equipment and historical fluctuation characteristics, effectively avoiding continuous deviations or output oscillations caused by the lag or inertia of a certain power source. Subsequently, the corrected allocation adjustment amount is added to the target power allocation parameters to obtain the final target output value, and a safety check and boundary correction are performed to ensure that the output of all power sources is within the safe allowable range of the equipment. Ultimately, this invention enables each power source to respond quickly, accurately, and dynamically to changes in grid load, significantly improving the system's adjustment flexibility, safety, and operational stability. For example, in actual power grid operation, if a power source has a lagging response in the previous cycle, the system will automatically reduce the increase in the amount of adjustment allocated to it in the current cycle, thereby avoiding output overshoot and improving the overall energy efficiency and equipment lifespan of the system.

[0157] Specifically, by combining the historical output data and dynamic response capabilities of each power source from the previous cycle, a correction coefficient for the adjustment amount allocated to each power source is determined, including:

[0158] During actual operation, the system continuously records the historical output data of each power source (such as photovoltaic, energy storage, etc.), including the actual output power in the previous cycle and its deviation from the target output value. At the same time, the system also obtains the dynamic response capability of each power source based on the equipment's factory parameters or operation evaluation results, such as the maximum adjustment speed, the allowable output change range, and its inertial response.

[0159] To determine the adjustment correction coefficient for the current cycle, the system first analyzes the deviation between the actual output and the target output of the previous cycle. If a power source lagged behind or had a large deviation in the previous cycle, the adjustment allocated to it in the current cycle should be appropriately reduced to avoid continuous error accumulation and system oscillation. Conversely, if its response is good and its output is stable, the adjustment can be maintained or appropriately increased to improve the overall dynamics of the system. Simultaneously, considering the dynamic response capability of the equipment, if the equipment's adjustment speed is slow or its range of change is limited, even if the algorithm allocates a large adjustment, it must be limited by a correction coefficient to ensure that the allocation result does not exceed the equipment's tolerance range. In practice, preset rules, rolling average statistics, weight mapping, and other methods can be used to calculate the correction coefficient for each power source, combining historical output and response capability. This coefficient applies to the adjustment allocation in the current cycle, ensuring that all adjustments have both historical correction logic and physical feasibility, effectively guaranteeing the safety and stability of system operation.

[0160] The security verification of the final target output value specifically includes:

[0161] After all the final target output values ​​of all power sources have been corrected and superimposed, the system needs to perform a comprehensive safety check on each target output value to prevent risks such as equipment over-limit operation, misoperation, or failure. The first step of the safety check is to compare the final target output value of each power source with the maximum and minimum allowable power range of its equipment to check whether it exceeds the safety range specified in the equipment's technical parameters. If a power source's target output is found to exceed the maximum power limit, the output value is automatically corrected to the maximum allowable value; if it is lower than the minimum safe output, it is adjusted to the minimum allowable value. Secondly, the system can further check the slope of the target output change to ensure that the power change per unit time does not exceed the equipment's maximum adjustment rate to prevent impact on the equipment. In addition, the total output power can be reasonably allocated according to the overall system operation strategy and grid stability requirements to prevent the overall system from being affected by the overload of a single power source. Only after all safety checks have passed will these target output values ​​be issued as actual power allocation commands, thereby ensuring that the system can operate efficiently and guarantee the safety and stability of equipment and the grid under various extreme conditions.

[0162] In a preferred embodiment of the present invention, the remaining power, upper limit of charging / discharging power, and safe temperature range of the energy storage device are monitored simultaneously. Combined with target power allocation parameters and load demand margin, it is determined whether the energy storage device needs to charge or discharge in the current cycle. Based on this, the charging / discharging power value is set, and an energy storage charging / discharging adjustment command is generated, including:

[0163] Real-time monitoring of the remaining power, upper limit of charging and discharging power, and actual temperature of the energy storage device in the current cycle forms an energy storage status data input set;

[0164] By comprehensively analyzing the energy storage status data input set with the target power allocation parameters and the load demand margin of the current cycle, it is determined whether the energy storage device should adopt charging mode or discharging mode in the current cycle, and the operation mode determination result is obtained.

[0165] Based on the operation mode determination result, determine the direction of charging or discharging required by the energy storage device in the current cycle, and set the actual executable charging and discharging power value with the remaining power, the upper limit of the equipment's allowable power and the temperature safety range as constraints.

[0166] The charging and discharging power values ​​are checked for safety and rationality against the target power allocation parameters and actual load requirements. If the values ​​exceed the equipment limits or there are safety hazards, they are automatically adjusted to the maximum or minimum values ​​within the allowable range of the equipment, generating corrected charging and discharging power values.

[0167] Based on the revised charge and discharge power values, the energy storage charge and discharge regulation command for the current cycle is generated.

[0168] In this embodiment of the invention, by simultaneously monitoring the remaining power, upper limit of charging / discharging power, and safe temperature range of the energy storage device, and combining the target power allocation parameters with the load demand margin, it is determined whether the energy storage device needs to be charged or discharged in the current cycle. Based on this, the charging / discharging power value is set, and energy storage charging / discharging adjustment commands are generated. This enables the energy storage system to respond sensitively and proactively to grid fluctuations and changes in energy consumption. Real-time acquisition of energy storage status data ensures that the charging / discharging control of the energy storage device in each cycle is based on the most accurate equipment condition and environmental constraints. Through comprehensive analysis with the target power allocation parameters and actual load margin, the optimal strategy for prioritizing charging or discharging the energy storage device can be dynamically determined, and appropriate charging / discharging power can be allocated accordingly, avoiding risks such as overcharging, over-discharging, or high temperature of the energy storage device. For example, when the grid-connected load drops significantly and the energy storage device's power is low, the system automatically enters charging mode and allocates appropriate charging power; conversely, when the load suddenly increases and the energy storage is in a high power range, it automatically switches to discharging mode to reasonably compensate for grid shortages. All set power values ​​undergo dynamic verification against equipment limits and safety constraints to ensure that the energy storage charging and discharging behavior supports stable grid operation while maximizing the safety and lifespan of the energy storage system itself. Ultimately, the energy storage device receives adjustment commands based on equipment status, demand changes, and safety constraints, achieving precise and efficient energy management and distribution.

[0169] Among them, the actual executable charging and discharging power values ​​are set based on the remaining power, the upper limit of the equipment's allowable power, and the safe temperature range. Specifically, these include:

[0170] Within each cycle, the system first acquires the remaining power information of the energy storage device in real time to understand the specific amount of available energy. Next, it queries the charging and discharging power limits specified by the energy storage device manufacturer or system configuration file to obtain the maximum allowable charging and discharging power under the current operating conditions. Simultaneously, the system also monitors the operating temperature of the energy storage device using temperature sensors or the device's built-in status monitoring function to ensure that its current temperature is within the specified safe operating range. If the temperature exceeds the safe range, the system automatically reduces the chargeable and discharge power or suspends operation.

[0171] After considering the above three constraints, the system adjusts the theoretical charging or discharging power calculated for this cycle: if the power exceeds the charging capacity allowed by the current remaining power or the upper limit of the device power, the power value is limited to the maximum allowable value; if the temperature is close to the limit, the power value will also be reduced according to a predetermined ratio. The final actual executable charging and discharging power value satisfies the current scheduling requirements without causing overcharging, over-discharging, battery life damage, or safety accidents, thus achieving safe and reliable energy storage participation in energy regulation.

[0172] This includes verifying the safety and rationality of the charging and discharging power values ​​against the target power allocation parameters and actual load requirements, specifically including:

[0173] After the system determines the actual executable charging and discharging power value for the current cycle, it further compares this value with the target power allocation parameters item by item to confirm whether it is consistent with the overall system scheduling target. For example, if the target allocation parameters require charging operations to be performed in the current cycle, but the actual calculated safe power value is negative or zero, the system will prioritize equipment safety and automatically adjust the actual power to the minimum allowable safe value or suspend the charging operation. At the same time, the system also needs to synchronously verify the actual load demand to ensure that the charging and discharging behavior of the energy storage device does not cause excessive fluctuations in parameters such as voltage and current in the local power grid, affecting the normal operation of the power consumption side or other equipment.

[0174] The process also includes determining whether the charging and discharging behavior may trigger the protection actions of other linked equipment, such as inverter overcurrent or bus voltage over-limit. After all verification steps are completed, if the actual power value passes all safety and rationality checks, it is directly used as the energy storage charging and discharging adjustment command issued for this cycle; if it fails, the power value is further adjusted to within all safety constraints, or the charging and discharging task for this cycle is temporarily suspended.

[0175] This series of rigorous safety and rationality checks ensures that the energy storage device always participates in grid energy regulation within a controllable and predictable range, maximizing the system's operational safety and economic benefits.

[0176] Embodiments of the present invention also provide a photovoltaic energy storage inverter power coordination control system based on an adaptive algorithm, the system comprising:

[0177] The data acquisition module is used to collect output power data of photovoltaic modules, status data of energy storage devices and load demand data of grid connection points, and normalize them to obtain the basic data for calculating allocation parameters.

[0178] The allocation parameter calculation module is used to calculate the preliminary power allocation parameters for the current cycle based on the basic data.

[0179] The trend analysis module is used to calculate the load change trend factor and the power fluctuation trend factor based on historical period load demand data and output power data, and couple the two to form a trend coupling factor.

[0180] The adaptive correction module is used to adaptively correct the initial power allocation parameters through the trend coupling factor, and generate the corrected target power allocation parameters.

[0181] The instruction generation and distribution module is used to generate target power distribution instructions and energy storage charging and discharging adjustment instructions based on the target power allocation parameters, combined with real-time grid connection point load demand data and energy storage device status data, and then distribute them to the inverter control unit and the energy storage device control unit, respectively.

[0182] It should be noted that this system is a system corresponding to the above method. All implementation methods in the above method embodiments are applicable to this embodiment and can achieve the same technical effect.

[0183] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A power coordination control method for photovoltaic energy storage inverters based on adaptive algorithms, characterized in that, The method includes: Collect the output power data of photovoltaic modules, the status data of energy storage devices, and the load demand data of grid connection points, and normalize them to obtain the basic data for calculating allocation parameters; Based on the basic data, the preliminary power allocation parameters for the current cycle are calculated, including: Determine whether the magnitude of the trend coupling factor exceeds the preset trend coupling threshold; When the value exceeds the limit, the trend coupling factor is multiplied by the initial power allocation parameter item by item to obtain the first correction value of each power source allocation parameter. Based on the first correction value, combined with the current cycle load demand variation and the available capacity of the energy storage device, the two are weighted and superimposed to each first correction value according to the preset feedback adjustment weight coefficient to obtain the second correction value. After summing all the second correction values, normalization is performed to ensure that the sum of all allocation parameters equals the current total allocable power, thus generating the target power allocation parameters after multi-level adaptive correction and normalization. Based on historical load demand data and output power data, load change trend factor and power fluctuation trend factor are calculated respectively, and the two are coupled to form a trend coupling factor. The initial power allocation parameters are adaptively corrected by the trend coupling factor to generate the corrected target power allocation parameters. Based on the target power allocation parameters, combined with real-time grid connection point load demand data and energy storage device status data, target power allocation instructions and energy storage charge / discharge adjustment instructions are generated, including: The system acquires the load demand data of the grid connection point in real time for the current period, calculates the difference between the target power allocation parameters and the load demand data, and allocates the difference to each power source according to the allocation ratio of the allocation parameters to form the target power allocation instruction. Simultaneously monitor the current remaining power, upper limit of charging and discharging power, and safe temperature range of the energy storage device. Combine the target power allocation parameters and load demand margin to determine whether the energy storage device needs to be charged or discharged in the current cycle, and set the charging and discharging power value accordingly to generate energy storage charging and discharging adjustment commands. The target power allocation command is sent to the inverter control unit, and the energy storage charging and discharging regulation command is sent to the energy storage device control unit, driving the relevant equipment to work together according to their respective commands; Monitor the inverter output power and the actual charging and discharging status of the energy storage device, and compare them with the target command in real time. If a deviation occurs, the system will automatically provide feedback and return to the initial power allocation parameters for recalculation, so as to achieve continuous closed-loop adaptive optimization. The target power allocation command and the energy storage charging and discharging regulation command are sent to the inverter control unit and the energy storage device control unit, respectively.

2. The photovoltaic energy storage inverter power coordination control method based on adaptive algorithm according to claim 1, characterized in that, Based on the basic data, the preliminary power allocation parameters for the current cycle are calculated, including: Based on the basic data, the mean of each data item and its corresponding data item in the previous period is differentially processed to obtain the difference value dataset. Based on the difference value dataset, the difference values ​​of each data item are weighted and superimposed according to the preset data allocation weight coefficients to obtain a weighted processing value dataset; The weighted values ​​are added together to form a composite parameter, and the theoretical load demand is inversely calculated using the composite parameter. The difference between the composite parameter and the actual load demand is then calculated to obtain the power distribution deviation. Based on the power allocation deviation, the total power to be allocated is distributed to each power source according to the proportion of the absolute value of the power allocation deviation, thus obtaining their respective preliminary power allocation parameters.

3. The photovoltaic energy storage inverter power coordination control method based on adaptive algorithm according to claim 1, characterized in that, Based on historical load demand and output power data, load change trend factors and power fluctuation trend factors are calculated separately, and these two are coupled to form a trend coupling factor, including: Perform time series analysis on a set amount of historical periodic load demand data, calculate the rate of change between load demand data of every two adjacent periods, and average all the rates of change to obtain the mean rate of change. The historical cycle load demand data is fitted with a polynomial, the slope of the fitted curve is extracted, and the slope value and the mean rate of change are weighted and superimposed according to the preset trend factor to form the load change trend factor. The historical periodic output power data is processed step by step using a sliding window. Within each window, the variance and range of the output power are calculated and multiplied by preset variance weighting coefficients and range weighting coefficients, respectively. The weighted variance value and range value are added together to obtain the initial value of the power fluctuation trend factor. The initial value of the power fluctuation trend factor is analyzed for directionality. The duration of the fluctuation rise or fall is statistically analyzed. Based on the correlation between the duration and the initial value of the power fluctuation trend factor, preset duration weight coefficients and initial value weight coefficients are assigned respectively. The two are weighted and superimposed to form the power fluctuation trend factor. The load change trend factor and the power fluctuation trend factor are weighted and summed according to the preset coupling weight coefficient. If the two are in the same direction, the coupling weight is increased; if they are in opposite directions, the coupling weight is decreased, thus generating a trend coupling factor.

4. The photovoltaic energy storage inverter power coordination control method based on adaptive algorithm according to claim 3, characterized in that, Historical periodic load demand data is fitted using a polynomial model to extract the slope of the fitted curve. The slope value and the mean rate of change are then weighted and superimposed according to a preset trend factor to form a load change trend factor, including: Based on historical periodic load demand data, each data item is arranged in chronological order to form a time series sample; A multinomial fitting algorithm is applied to the time series samples to obtain a curve reflecting the trend of load change, and the slope value of the curve in the latest period is extracted. Based on the preset trend factor, the weight coefficients are synthesized. The slope value and the mean rate of change are multiplied by their respective weight coefficients, and the two multiplications are accumulated and added together to generate the load change trend factor for the current period.

5. The photovoltaic energy storage inverter power coordination control method based on adaptive algorithm according to claim 3, characterized in that, A directional analysis is performed on the initial value of the power fluctuation trend factor, and the duration of the fluctuation increase or decrease is statistically analyzed. Based on the correlation between the duration and the initial value of the power fluctuation trend factor, preset weighting coefficients for the duration and the initial value are assigned respectively. The two are then weighted and superimposed to form the power fluctuation trend factor, including: Based on the initial value of the power fluctuation trend factor, the power output data of the historical cycle is compared with the data cycle by cycle to identify and count the number of cycles with continuous upward or downward changes in the same direction, and the length of the continuous cycle is obtained. Based on the duration period length and the initial value of the power fluctuation trend factor, and according to their influence in previous adjustments, the duration period weight coefficient and the initial value weight coefficient are set. Multiply the duration of the period by the duration weight coefficient, multiply the initial value of the power fluctuation trend factor by the initial value weight coefficient, and add the two weighted results to obtain the power fluctuation trend factor for the current period.

6. The photovoltaic energy storage inverter power coordination control method based on adaptive algorithm according to claim 1, characterized in that, Based on the first correction value, and combined with the current cycle load demand variation and the available capacity of the energy storage device, the two are weighted and superimposed onto each of the first correction values ​​according to a preset feedback adjustment weighting coefficient to obtain the second correction value, including: The current period load demand variation and the available capacity of the energy storage device are obtained, and feedback adjustment weight coefficients are assigned to each of them. The feedback adjustment weight coefficients are dynamically adjusted according to the historical contribution of their respective parameters to the stability of power distribution. The weighted feedback value is obtained by multiplying the change in load demand by its feedback adjustment weighting coefficient and by multiplying the available capacity of the energy storage device by its feedback adjustment weighting coefficient. Each first correction value is added to the corresponding weighted feedback value to obtain the second correction value for each power source in the current cycle.

7. The photovoltaic energy storage inverter power coordination control method based on adaptive algorithm according to claim 1, characterized in that, The system acquires real-time load demand data for the current period at grid-connected points, calculates the difference between the target power allocation parameters and the load demand data, and distributes this difference to each power source according to the allocation ratio of the allocation parameters, forming a target power allocation instruction, including: The load demand data of the grid connection point in the current period is obtained in real time, and it is compared with the target power allocation parameters of the current period item by item. The difference between the theoretical output power of each power source and the actual load demand is calculated to obtain the power difference that needs to be adjusted for each power source. The power differences of each power source are summed to obtain the total power difference. Based on the proportion of each power source in the target power allocation parameters, the total power difference is allocated to each power source proportionally to form their respective allocation adjustment amounts; Combining the historical output data and dynamic response capabilities of each power source in the previous cycle, the correction coefficient of the allocation adjustment amount for each power source is determined, and the corrected allocation adjustment amount is added to its target power allocation parameter to obtain the final target output value. The final target output value is checked for safety. If it exceeds the maximum or minimum power range allowed by the device, it is automatically corrected to the boundary value allowed by the device. The final target output value after safety verification is used as the target power allocation command for the current cycle.

8. The photovoltaic energy storage inverter power coordination control method based on adaptive algorithm according to claim 1, characterized in that, Simultaneously, it monitors the energy storage device's remaining power, charging / discharging power limits, and temperature safety range. Combining this with target power allocation parameters and load demand margin, it determines whether the energy storage device needs charging or discharging in the current cycle, and accordingly sets the charging / discharging power values, generating energy storage charging / discharging adjustment commands, including: Real-time monitoring of the remaining power, upper limit of charging and discharging power, and actual temperature of the energy storage device in the current cycle forms an energy storage status data input set; By comprehensively analyzing the energy storage status data input set with the target power allocation parameters and the load demand margin of the current period, it is determined whether the energy storage device should adopt charging mode or discharging mode in the current period, and the operation mode determination result is obtained. Based on the operation mode determination result, determine the direction of charging or discharging required by the energy storage device in the current cycle, and set the actual executable charging and discharging power value with the remaining power, the upper limit of the equipment's allowable power and the temperature safety range as constraints. The charging and discharging power values ​​are checked for safety and rationality against the target power allocation parameters and actual load requirements. If the values ​​exceed the equipment limits or there are safety hazards, they are automatically adjusted to the maximum or minimum values ​​within the allowable range of the equipment, generating corrected charging and discharging power values. Based on the revised charge and discharge power values, the energy storage charge and discharge regulation command for the current cycle is generated.

Images