Hybrid inverter control method, device, equipment and storage medium

Abstract

The present application relates to the field of hybrid inverter control technology, and discloses a hybrid inverter control method, device, equipment and storage medium, the method comprising: collecting solar panel output data, battery energy storage system status data and load demand data, and constructing a system dynamic model; predicting energy allocation schemes; calculating optimal power flow instructions; respectively calculating the maximum power output value of the solar power generation system and the optimal charge and discharge power of the battery energy storage system; based on the maximum power output value and the optimal charge and discharge power, adjusting the output current of the hybrid inverter to the power grid, and generating coordinated control instructions; generating a switch control sequence according to the coordinated control instruction, and calculating the reactive compensation value to obtain a real-time control signal, thereby reducing the harmonic content of the output voltage and improving the power quality. By calculating the reactive compensation value in real time and combining it with the switch control sequence, dynamic reactive power compensation of the hybrid inverter is realized.

Description

Technical Field

[0001] The present application relates to the technical field of hybrid inverter control, and in particular to a hybrid inverter control method, device, equipment and storage medium. Background Art

[0002] As a key device for integrating solar power generation and battery energy storage systems, hybrid inverters play an increasingly important role in distributed energy systems. Hybrid inverters can not only achieve efficient utilization of solar power generation, but also smooth power fluctuations through battery energy storage systems, thereby improving system stability and reliability. However, traditional hybrid inverter control methods face multiple challenges.

[0003] Due to the intermittent and volatile nature of solar power generation, as well as the uncertainty of load demand, accurately predicting short-term power generation and load demand has become a key issue for the effective operation of hybrid inverters. Secondly, in a complex system where multiple energy forms work together, how to achieve optimal power distribution and energy management to balance system efficiency, battery life and economy is a difficult problem that needs to be solved urgently. In addition, as an interface device between the power grid and distributed energy, the hybrid inverter also needs to meet the strict requirements of the power grid for power quality. How to achieve fast and accurate voltage and frequency regulation while ensuring system stability, while minimizing harmonic interference and reactive power loss, is another important challenge facing hybrid inverter control technology. Summary of the invention

[0004] The present application provides a hybrid inverter control method, device, equipment and storage medium, thereby reducing the harmonic content of the output voltage and improving the power quality. Dynamic reactive power compensation of the hybrid inverter is achieved by calculating the reactive compensation value in real time and combining it with the switch control sequence.

[0005] A first aspect of the present application provides a hybrid inverter control method, the hybrid inverter control method comprising:

[0006] Collect solar panel output data, battery energy storage system status data and load demand data, and build a system dynamic model;

[0007] Based on the system dynamic model, short-term load demand, solar power generation and grid electricity charges are predicted to obtain an energy allocation plan;

[0008] Calculate the power balance relationship between the solar power generation system, the battery energy storage system and the load according to the energy allocation scheme to obtain the optimal power flow instruction;

[0009] Based on the optimal power flow instruction, respectively calculating the maximum power output value of the solar power generation system and the optimal charging and discharging power of the battery energy storage system;

[0010] Based on the maximum power output value and the optimal charge and discharge power, adjusting the output current of the hybrid inverter to the power grid, and generating a coordinated control instruction;

[0011] A switch control sequence is generated according to the coordinated control instruction, and a reactive compensation value is calculated to obtain a real-time control signal of the hybrid inverter.

[0012] A second aspect of the present application provides a hybrid inverter control device, the hybrid inverter control device comprising:

[0013] The acquisition module is used to collect the solar panel output data, the battery energy storage system status data and the load demand data, and build a system dynamic model;

[0014] A prediction module, used for predicting short-term load demand, solar power generation and grid electricity charges based on the system dynamic model to obtain an energy allocation plan;

[0015] A calculation module, used to calculate the power balance relationship between the solar power generation system, the battery energy storage system and the load according to the energy allocation scheme, and obtain an optimal power flow instruction;

[0016] A processing module, used to calculate the maximum power output value of the solar power generation system and the optimal charging and discharging power of the battery energy storage system respectively based on the optimal power flow instruction;

[0017] A regulating module, configured to regulate the output current of the hybrid inverter to the power grid based on the maximum power output value and the optimal charge and discharge power, and generate a coordinated control instruction;

[0018] A generating module is used to generate a switch control sequence according to the coordinated control instruction, and calculate a reactive power compensation value to obtain a real-time control signal of the hybrid inverter.

[0019] A third aspect of the present application provides an electronic device, comprising: a memory and at least one processor, wherein the memory stores instructions; the at least one processor calls the instructions in the memory so that the electronic device executes the above-mentioned hybrid inverter control method.

[0020] A fourth aspect of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores instructions, which, when executed on a computer, enable the computer to execute the above-mentioned hybrid inverter control method.

[0021] Compared with the prior art, the present application has the following beneficial effects: by adopting the recursive least squares method to estimate the system parameters, combining the autoregressive moving average model and the support vector machine algorithm to predict the short-term load and power generation, the accuracy and prediction precision of the system dynamic model are improved, and the discrete time state space model is constructed by using the rolling time domain optimization algorithm to realize the dynamic power balance between the solar power generation system, the battery energy storage system and the load, and optimize the energy distribution efficiency of the system. The maximum power output value of the solar power generation system and the optimal charge and discharge power of the battery energy storage system are calculated respectively, which improves the energy utilization efficiency and battery life. Combined with the power buffer and the fuzzy logic controller, the precise adjustment of the output voltage and frequency of the hybrid inverter is realized, and the dynamic response capability and stability of the system are enhanced. The space vector pulse width modulation technology is used to generate the switch control sequence, and combined with the selective harmonic elimination algorithm and the dead time compensation, the harmonic content of the output voltage is significantly reduced, and the power quality is improved. By calculating the reactive compensation value in real time and combining it with the switch control sequence, the dynamic reactive power compensation of the hybrid inverter is realized, and the power factor and grid friendliness of the system are improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0023] The structures, proportions, sizes, etc. illustrated in the drawings of this specification are only used to match the contents disclosed in the specification so as to facilitate understanding and reading by persons familiar with this technology. They are not used to limit the conditions under which the present invention can be implemented, and therefore have no substantive technical significance. Any structural modification, change in proportion or adjustment of size, without affecting the effects and purposes that can be achieved by the present invention, should still fall within the scope of the technical contents disclosed by the present invention.

[0024] Figure 1 is a flow chart of a hybrid inverter control method provided by an embodiment of the present invention;

[0025] Figure 2 It is a schematic block diagram of the structure of a hybrid inverter control device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0026] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0027] The flowcharts shown in the accompanying drawings are only examples and do not necessarily include all the contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps may also be decomposed, combined or partially merged, so the actual execution order may change according to actual conditions.

[0028] It should also be understood that the terms used in this application specification are only for the purpose of describing specific embodiments and are not intended to limit the application. As used in this application specification and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include plural forms.

[0029] It should be further understood that the term "and / or" used in the specification and appended claims of this application refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations. Figure 1 , an embodiment of the hybrid inverter control method in the embodiment of the present application includes:

[0030] Step 100: Collect solar panel output data, battery energy storage system status data and load demand data, and build a system dynamic model;

[0031] It is understandable that the execution subject of the present application may be a hybrid inverter control device, or a terminal or a server, which is not specifically limited here. The present application embodiment is described by taking a server as the execution subject as an example.

[0032] Specifically, the output voltage and current of the solar panel are sampled to obtain a voltage and current data sequence reflecting the output characteristics of the solar panel, reflecting the output characteristics of the solar panel under different lighting conditions. At the same time, the terminal voltage, charge and discharge current and battery temperature of the battery energy storage system are measured to obtain a state data set of the battery energy storage system, describing the current operating state of the battery, including the battery state of charge (SOC) and temperature changes. The active power and reactive power at the load end are monitored in real time to obtain active and reactive power data sets reflecting the load demand, reflecting the actual load demand of the current system. The solar panel output data sequence, the battery energy storage system state data set and the load demand data set are denoised and standardized respectively to obtain a preprocessed data set. The preprocessed data set is subjected to time series analysis to capture the dynamic relationship between the system input and output. Based on the time series analysis, the objective function of the recursive least squares algorithm is constructed. The recursive least squares algorithm is an online parameter identification method that can continuously update the system parameters according to new data during the operation of the system to keep the system model accurately describing the current state. When constructing the objective function, the initial parameter value and the forgetting factor are set. The initial parameter value is used to give the initial state of the algorithm, while the forgetting factor is used to control the weight of the old data in the parameter estimation. It is usually between 0 and 1. By reasonably selecting the forgetting factor, while ensuring that the model has sufficient memory for historical data, it pays more attention to the latest data, thereby improving the real-time and adaptability of the system model. Based on these settings, the iterative formula for parameter estimation is obtained to guide the subsequent online parameter identification process. Based on the parameter estimation iterative formula, online parameter identification calculations are performed. New data is continuously collected during the operation of the system, and the system parameters are updated through the iterative formula to obtain the updated system parameter values. Using the updated system parameter values, the state equation and output equation of the hybrid inverter system are constructed to obtain the dynamic model of the system. The state equation describes the law of change of the internal state of the system, while the output equation describes the external performance of the system. These two equations comprehensively characterize the behavior of the hybrid inverter system under different working conditions.

[0033] Step 200: predict short-term load demand, solar power generation and grid electricity charges based on the system dynamic model to obtain an energy allocation plan;

[0034] Specifically, according to the system dynamic model, the load demand data set and the output data sequence of the solar panel are decomposed into time series, and the original data is decomposed into multiple components to reveal their internal laws. Through time series decomposition, the load demand and the output data of the solar panel are decomposed into trend items, periodic items and random items. The trend item reflects the long-term trend of the data, the periodic item describes the periodic fluctuation of the data, and the random item reflects the unpredictable random fluctuation in the data. The trend item is differentially processed to transform the trend item into a stationary sequence. Through differential processing, the trend component in the sequence is eliminated to meet the stationary requirements. Based on the processed stationary sequence, an autoregressive moving average (ARMA) analysis is performed. The ARMA model is a time series prediction method that can effectively capture the linear relationship in the data and obtain the linear prediction results of the load demand and solar power generation. Through the autoregressive part, the model can use historical data to predict future trends, while the moving average part is used to process random errors in the data to improve the accuracy of the prediction. The nonlinear features of the periodic items and random items of the data are learned. The periodic items and random items are input into the support vector machine model for nonlinear feature learning. The support vector machine is a machine learning method that is good at dealing with complex nonlinear problems. By performing nonlinear learning on periodic terms and random terms, the complex fluctuation characteristics of load demand and solar power generation that are difficult to describe by linear models are captured, and more accurate nonlinear prediction results are obtained. The support vector machine maps data to high-dimensional space through kernel functions and finds the optimal hyperplane in high-dimensional space for prediction. The linear prediction results and nonlinear prediction results are weighted and fused to obtain the final short-term load demand prediction value and solar power generation prediction value. The historical power grid electricity fee data is obtained, and the electricity fee is predicted based on the short-term load demand prediction value and solar power generation prediction value to obtain the predicted power grid electricity fee data. Based on the short-term load demand prediction value, solar power generation prediction value and predicted power grid electricity fee data, the energy balance constraint equation is constructed and iterative calculation is performed to obtain the final energy allocation plan. The energy balance constraint equation describes the supply and demand balance relationship between different energy sources in the system, that is, the power balance between solar power generation, battery energy storage and load demand. By constructing the constraint equation, the energy flow between solar power generation and battery energy storage is reasonably allocated on the premise of meeting the system power demand to achieve optimal energy utilization. In the process of iterative calculation, the power distribution between various energy sources is continuously adjusted until the optimal solution for energy utilization of the entire system is found. Iterative optimization methods, such as gradient descent or Newton's method, are used to improve the efficiency and accuracy of the calculation.

[0035] Step 300: Calculate the power balance relationship between the solar power generation system, the battery energy storage system and the load according to the energy allocation plan to obtain the optimal power flow instruction;

[0036] It should be noted that based on the energy allocation scheme, the physical quantities such as voltage, current and power in the hybrid inverter system are discretized. The continuous signal is divided into several discrete points on the time axis to obtain the state variables and control variables of the system. The state equation and output equation of the hybrid inverter are constructed using the system state variables and control variables to obtain a discrete time state space model. The state space model is an effective tool to describe the dynamic characteristics of the system. The state equation describes the change law of the system state over time, while the output equation describes the external output characteristics of the system. For the hybrid inverter system, the state equation reflects the evolution process of voltage and current at different time steps, while the output equation describes how the output power of the inverter is affected by the input signal and state. According to the discrete time state space model, the prediction interval and control interval of the rolling time domain optimization are set. Rolling time domain optimization is a control strategy that can make the optimal control decision at the current moment by predicting and optimizing the system behavior within a certain period of time in the future. When setting the rolling time domain optimization, the prediction interval determines the length of time required for prediction in the future, and the control interval determines the time step for actual control. Based on the set prediction interval and control interval, a multi-objective optimization problem is established to achieve optimal control of the system operation state. In the multi-objective optimization problem, multiple objectives are included, such as minimizing the power balance deviation, maintaining the battery state of charge (SOC), and minimizing the inverter loss. Minimizing the power balance deviation is to ensure the supply and demand balance between power generation, energy storage and load, so as to achieve stable operation of the system; maintaining the battery state of charge is to prevent the battery from overcharging or over-discharging, so as to extend the battery life; and minimizing the inverter loss is to improve the energy utilization efficiency of the system and reduce energy loss. Weights are assigned to each objective to determine the importance of each objective in the entire optimization problem. The penalty function method is used to transform the multi-objective optimization problem into a single-objective optimization problem. The penalty function method introduces a penalty term to integrate multiple objectives into one objective, thereby simplifying the optimization process. After completing the conversion, a standardized optimization model is obtained. The standardized optimization model is substituted into the quadratic programming solver. Quadratic programming is suitable for dealing with optimization problems with quadratic objective functions and linear constraints. Through iterative calculation, the optimal control sequence at the current moment is obtained. The first control quantity is extracted from the optimal control sequence at the current moment as the control instruction at the current moment, and applied to the hybrid inverter system to achieve real-time power flow distribution. The result of real-time power flow distribution can effectively reflect the operating status of the current system and the energy flow between each subsystem. By applying the current control instruction, corresponding adjustments are made according to the predicted results to achieve optimal power management.The real-time power flow distribution result is compared with the preset threshold. If the power flow distribution result exceeds the set threshold range, it means that the current control scheme cannot meet the system's operating requirements, or some unpredictable disturbances occur. At this time, the re-optimization process is triggered to re-evaluate the system's operating status and calculate a new optimal control scheme. If the power flow distribution result does not exceed the threshold range, it means that the current control scheme is feasible, and it moves forward according to the prediction and control steps, rolling to the next moment, continuing the next round of prediction and control, and finally obtaining the optimal power flow instruction.

[0037] Step 400: Based on the optimal power flow instruction, the maximum power output value of the solar power generation system and the optimal charging and discharging power of the battery energy storage system are calculated respectively;

[0038] Specifically, the first target power value of the solar power generation system and the second target power value of the battery energy storage system are extracted from the optimal power flow instruction to obtain the power allocation target of the entire system. Based on the output data sequence of the solar panel, the IV characteristic curve model of the solar panel is constructed. The IV characteristic curve reflects the output characteristics of the solar panel under different voltage and current conditions. By constructing this model, the working state of the solar panel under different environmental conditions is simulated. The IV characteristic curve model is differentiated to calculate the slope of the power-voltage curve, that is, the power change rate, so as to analyze the trend of the power output of the solar panel with the voltage change, which helps to find the maximum power point. The maximum power point is the maximum power that the solar panel can output under specific conditions. Accurate tracking of it can significantly improve the efficiency of solar energy utilization. According to the power change rate, the perturbation observation method is used to iteratively adjust the working point of the solar panel to find the voltage and current values ​​of the maximum power point. The perturbation observation method is a maximum power point tracking algorithm that introduces small disturbances on the voltage or current to observe the change of power, thereby continuously approaching the maximum power point. If the power increases with the increase of voltage, the voltage continues to increase, otherwise the voltage is reduced until the position where the power no longer increases is found, that is, the maximum power point. The voltage and current values ​​at the maximum power point are multiplied to obtain the maximum power output value of the solar power generation system. At the same time, the battery state of charge (SOC) is estimated based on the battery energy storage system status data set to obtain the current battery SOC value. SOC is a key parameter in the battery management system, which reflects the remaining power of the battery and directly affects the battery charging and discharging strategy. According to the current battery SOC value and the second target power value of the battery energy storage system, the dynamic programming algorithm is used to calculate the optimal trajectory of the battery charging and discharging power. Dynamic programming is an effective method to solve multi-stage decision problems. By decomposing the problem into a series of sub-problems, the decision of each step is gradually solved and optimized to obtain the optimal solution of the entire process. Dynamic programming is used to calculate the optimal charging and discharging trajectory of the battery in different time periods to meet the power requirements of the system, while avoiding overcharging or over-discharging of the battery and extending the battery life. The optimal charging and discharging power of the battery energy storage system is obtained through the calculation of the dynamic programming algorithm. This power value ensures that the battery can maintain a reasonable SOC level while meeting the energy requirements of the system, avoiding battery performance degradation caused by excessive charging and discharging.

[0039] Step 500: Based on the maximum power output value and the optimal charge and discharge power, adjust the output current of the hybrid inverter to the power grid and generate a coordinated control instruction;

[0040] Specifically, the maximum power output value and the optimal charge and discharge power are input into the power buffer to smooth the power fluctuation. The function of the power buffer is to eliminate short-term fluctuations in the power command. Through smoothing, a more stable power command is obtained, so that the control process of the hybrid inverter is smoother and the adverse effect of power fluctuations on the stability of system operation is reduced. Based on the smoothed power command, the target output voltage and frequency of the hybrid inverter are calculated to obtain the initial droop control parameters. Droop control is a technology commonly used in microgrid and inverter control. Through droop control, the output voltage and frequency of the inverter can be adjusted according to the power demand of the system, so that different energy sources can work together without communication. In this process, the output voltage and frequency of the hybrid inverter are measured in real time to obtain the current actual output value, reflecting the output state of the hybrid inverter at the current moment. The actual output value is compared with the target output voltage and frequency, and the deviation value between the two is calculated to obtain the control error. The control error is used to reflect the gap between the actual output and the target. By adjusting the error, the system gradually approaches the expected operating state. Based on the control error, the fuzzy logic controller is used to adjust the initial droop control parameters online to obtain the adaptive droop coefficient. The fuzzy logic controller is a controller based on fuzzy set theory, which is suitable for processing complex systems with uncertainty and nonlinear characteristics. In this process, the fuzzy logic controller can adaptively adjust the droop control parameters according to the real-time measured control error, so that the control process has higher robustness and adaptability. The droop control equation is updated according to the adaptive droop coefficient, and the corrected output voltage and frequency adjustment amount are calculated. The corrected output voltage and frequency adjustment amount are substituted into the power balance equation, and iterative calculation is performed to obtain the final output voltage and frequency setting value. Through this process, it is ensured that the power supply and demand of the system is balanced, and the output voltage and frequency are kept within a reasonable range, so as to achieve stable operation of the system. Based on the output voltage and frequency setting values, the output current of the hybrid inverter is adjusted to the grid, and the coordinated control instruction of the hybrid inverter is generated. The coordinated control instruction contains the voltage amplitude, phase angle and frequency adjustment amount, which is used to guide the actual operation of the hybrid inverter. Through these control instructions, the hybrid inverter can adjust its output according to the power demand and load characteristics of the system, and achieve the optimal configuration of energy utilization while meeting the load demand. The coordinated control method based on adaptive droop control and fuzzy logic control can effectively deal with various uncertainties and dynamic changes in system operation, so that the hybrid inverter can maintain stability and efficiency in a complex operating environment.

[0041] Step 600: Generate a switch control sequence according to the coordinated control instruction, calculate the reactive power compensation value, and obtain a real-time control signal of the hybrid inverter.

[0042] Specifically, the voltage amplitude and phase angle in the coordinated control instruction are Clarke transformed, and the three-phase stationary coordinate system is converted into a two-phase stationary coordinate system to obtain the voltage vector in the αβ coordinate system. Clarke transformation is a coordinate transformation method that simplifies the analysis and calculation of voltage and current by converting the three-phase voltage into a two-phase αβ coordinate system. Based on the voltage vector in the αβ coordinate system, the sector where the voltage space vector is located is determined, and the action time of two adjacent basic vectors is calculated to obtain the vector synthesis time. Voltage space vector modulation is a modulation method that obtains the target voltage vector by synthesizing different voltage vectors. Determining the sector where the voltage vector is located can clarify the state of the current system, so that the corresponding basic vectors are selected for synthesis according to different sectors, and the action time of adjacent basic vectors determines the actual performance of the target voltage vector in a switching cycle. According to the vector synthesis time, the conduction time of each switch tube is calculated using a symmetrical sinusoidal pulse width modulation algorithm to obtain an initial switch sequence. Symmetrical sinusoidal pulse width modulation can control the output voltage of the inverter by adjusting the conduction time of each switch. Since there is a delay in the actual operation of the switching device, in order to avoid the short-circuit problem caused by the simultaneous conduction of the switch tube, the initial switch sequence is compensated for the dead time to eliminate the shoot-through phenomenon during the conduction and shutdown of the switch tube, and the compensated switch sequence is obtained. Based on the compensated switch sequence, the harmonic content of the output voltage of the hybrid inverter is calculated, and the selective harmonic elimination algorithm is used to suppress the high-order harmonics to obtain the optimized switch sequence. High-order harmonics are the main harmonic components in the inverter output, which will have an adverse effect on the operating efficiency and stability of the system. The selective harmonic elimination algorithm reduces the high-order harmonics of a specific frequency by reasonably selecting the switch sequence, optimizes the harmonic characteristics of the system, and makes the output voltage waveform closer to the ideal sine wave. The optimized switch sequence is compared with the triangular carrier to generate a pulse width modulation waveform to obtain the final switch control sequence. In the process of generating the pulse width modulation waveform, the optimized switch sequence is used to modulate the triangular carrier to obtain the corresponding pulse width signal, which is used to control the switch tube of the inverter to achieve precise control of voltage and frequency. The voltage and current at the output of the hybrid inverter are sampled, and the actual active power and reactive power of the system are calculated by processing the sampled data to obtain the power calculation result. Based on the power calculation result and the preset power factor, the required compensation current is calculated using the instantaneous reactive power theory to obtain the reactive compensation value. The instantaneous reactive power theory is an effective tool for analyzing active and reactive power in three-phase power systems. By calculating the required compensation current, the reactive component in the system is eliminated and the power factor is improved without affecting the active power transmission. The reactive compensation value is combined with the switch control sequence to generate a real-time control signal for the hybrid inverter.The control signal contains all the necessary information required for the system operation, including voltage amplitude, phase angle and frequency adjustment, and through precise control of the switching tube, ensures that the system can operate according to the set power instructions and reactive power compensation requirements.

[0043] In the embodiment of the present application, the system parameters are estimated by using the recursive least squares method, and the short-term load and power generation forecasting is performed by combining the autoregressive moving average model and the support vector machine algorithm, so as to improve the accuracy and prediction precision of the system dynamic model, and the discrete time state space model is constructed by using the rolling time domain optimization algorithm, so as to realize the dynamic power balance between the solar power generation system, the battery energy storage system and the load, and optimize the energy distribution efficiency of the system. The maximum power output value of the solar power generation system and the optimal charge and discharge power of the battery energy storage system are calculated respectively, so as to improve the energy utilization efficiency and battery life. In combination with the power buffer and the fuzzy logic controller, the precise adjustment of the output voltage and frequency of the hybrid inverter is realized, and the dynamic response capability and stability of the system are enhanced. The switch control sequence is generated by using the space vector pulse width modulation technology, and the selective harmonic elimination algorithm and the dead time compensation are combined to significantly reduce the harmonic content of the output voltage and improve the power quality. By calculating the reactive compensation value in real time and combining it with the switch control sequence, the dynamic reactive power compensation of the hybrid inverter is realized, and the power factor and grid friendliness of the system are improved.

[0044] In a specific embodiment, the process of executing step 100 may specifically include the following steps:

[0045] The output voltage and current of the solar panel are sampled to obtain the solar panel output data sequence; the terminal voltage, charge and discharge current and battery temperature of the battery energy storage system are measured to obtain the battery energy storage system status data set; the active power and reactive power at the load end are monitored in real time to obtain the load demand data set;

[0046] The solar panel output data sequence, the battery energy storage system status data set and the load demand data set are denoised and standardized respectively to obtain the preprocessed data sets;

[0047] Perform time series analysis on the preprocessed data set to obtain the corresponding relationship between the system input and output. Based on the corresponding relationship between the system input and output, construct the objective function of the recursive least squares algorithm, set the initial parameter value and forgetting factor, and obtain the parameter estimation iteration formula.

[0048] Based on the parameter estimation iteration formula, online parameter identification calculation is performed to obtain updated system parameter values. The updated system parameter values ​​are used to construct the state equation and output equation of the hybrid inverter system to obtain the system dynamic model.

[0049] Specifically, data is collected from solar panels, battery energy storage systems, and load ends. For the sampling process of solar panels, the output voltage and current signals are obtained. By sampling these signals in real time, an output data sequence reflecting the performance of the solar panel is obtained, describing the output characteristics of the solar panel under different lighting conditions and load changes. For the battery energy storage system, its terminal voltage, charge and discharge current, and battery temperature are measured to obtain a state data set of the battery energy storage system. These parameters are important indicators for describing the working state of the battery. For example, the terminal voltage and charge and discharge current are used to calculate the state of charge (SOC) of the battery, while the temperature reflects the operating environment and health status of the battery. The active power and reactive power at the load end are monitored in real time to obtain a load demand data set to reflect the actual demand of the current system load. The solar panel output data sequence, the battery energy storage system state data set, and the load demand data set are denoised and standardized. The denoising process is to eliminate the measurement errors caused by environmental interference, equipment noise, and other factors to ensure the accuracy of the data. Denoising methods include moving average filtering, median filtering, and the like. For example, for high-frequency noise in the voltage signal, it is eliminated by moving average filtering, and the formula is:

[0050] ;

[0051] in, Represents the voltage signal after filtering, is the voltage value at the current moment, is the size of the filter window. Through this processing, the impact of random noise on the data analysis results can be effectively reduced. Standardization is performed to eliminate the dimensional differences of different data. The numerical range of each data set is adjusted to the same scale to improve the consistency and accuracy of different types of data in the subsequent modeling process. Time series analysis is performed on the preprocessed data set to obtain the corresponding relationship between the system input and output. In the time series analysis process, the recursive least squares algorithm is used for parameter identification. Construct the objective function of the recursive least squares algorithm. The objective function is usually in the form of minimizing the sum of squared errors. The formula is as follows:

[0052] ;

[0053] in, is the objective function, is the output of the system, is the input vector, which contains the input data and status information of the system. is the parameter vector to be estimated, is the forgetting factor, and its value range is , which is used to control the impact of new and old data on parameter estimation. Through the recursive least squares algorithm, the system parameters are estimated online, and the system model is continuously updated according to the new data. Based on the objective function, the initial parameter value and the forgetting factor are set to obtain the iterative formula for parameter estimation. The recursive least squares algorithm uses historical data and newly collected data to update parameters, and its iterative formula is:

[0054] ;

[0055] in, is the updated parameter vector, is the parameter estimate at the current moment, is the gain matrix, which is used to adjust the step size of parameter update. Through the iterative process, the system model parameters are constantly approaching the true values, and a more accurate system dynamic model is obtained. The state equation and output equation of the hybrid inverter system are constructed using the updated system parameter values. The state equation describes the change law of the internal state of the system over time, while the output equation describes the external performance of the system. For example, the state equation is expressed as:

[0056] ;

[0057] in, is the system state vector, is the state transfer matrix, is the control input matrix, is the control input vector. The output equation is expressed as:

[0058] ;

[0059] in, is the system output vector, and are the output matrix and direct transfer matrix respectively. These equations can describe the dynamic response characteristics of the hybrid inverter system under different input conditions.

[0060] In a specific embodiment, the process of executing step 200 may specifically include the following steps:

[0061] According to the system dynamic model, the load demand data set and the solar panel output data sequence are decomposed into time series to obtain trend terms, periodic terms and random terms;

[0062] The trend term is differentiated to obtain a stationary sequence, and an autoregressive moving average analysis is performed based on the stationary sequence to obtain the linear prediction results of load demand and solar power generation;

[0063] The periodic term and the random term are input into the support vector machine model for nonlinear feature learning to obtain the nonlinear prediction results of load demand and solar power generation;

[0064] The linear prediction results and the nonlinear prediction results are weighted and fused to obtain the short-term load demand prediction value and the solar power generation prediction value;

[0065] Obtain historical power grid electricity fee data, and make electricity fee forecasts based on short-term load demand forecast values ​​and solar power generation forecast values ​​to obtain forecast power grid electricity fee data;

[0066] Based on the short-term load demand forecast, solar power generation forecast and predicted grid electricity price data, an energy balance constraint equation is constructed and iterative calculation is performed to obtain the energy allocation plan.

[0067] Specifically, the dynamic model of the system is analyzed, and then the load demand data set and the solar panel output data sequence are decomposed into time series. The original data is decomposed into different components to facilitate a clearer understanding and processing of their inherent laws of change. The load demand data and the solar panel output data sequence are decomposed into trend terms, periodic terms, and random terms, respectively. The trend term reflects the long-term change direction of the data, the periodic term represents the periodic change pattern in the data, and the random term describes the short-term random fluctuations in the data. The trend term is differentially processed to convert it into a stationary sequence. The characteristic of a stationary sequence is that its statistical characteristics (such as mean and variance) do not change over time. Through differential operations, the non-stationary components in the trend term are eliminated to make the sequence stationary. For example, for a time series , the difference processing is realized by the following formula:

[0068] ;

[0069] in, represents the stationary series after difference, is the trend item value at the current moment, is the trend item value of the previous moment. Through differential processing, the long-term upward or downward trend is effectively eliminated, making the data more suitable for autoregressive analysis. Autoregressive moving average analysis is performed based on the stationary series. The autoregressive moving average analysis model combines the autoregressive part and the moving average part to capture the linear characteristics in the time series. The autoregressive part predicts the future value by regressing the past value of the sequence, while the moving average part uses the past errors to improve the accuracy of the prediction. The mathematical expression of the autoregressive moving average analysis model is:

[0070] ;

[0071] in, is the predicted time series value, is the autoregression coefficient, is a random error, is the moving average coefficient. In this way, the linear prediction results of load demand and solar power generation are obtained to capture the linear change law presented in the data. The periodic terms and random terms are analyzed for nonlinear feature learning. The support vector machine model is used to learn the nonlinear features of the periodic terms and random terms. Support vector machine is a machine learning method suitable for processing data with complex nonlinear relationships. By inputting the periodic terms and random terms into the support vector machine model, more accurate nonlinear prediction results of load demand and solar power generation are obtained. The support vector machine model maps the input data to a high-dimensional feature space through a kernel function, and finds the optimal hyperplane in the high-dimensional space to achieve data classification or regression, so that it can well cope with the complex characteristics of periodic changes and random fluctuations. The linear prediction results and the nonlinear prediction results are weighted and fused to obtain the final short-term load demand forecast value and solar power generation forecast value. The weighted fusion process is expressed by the following formula:

[0072] ;

[0073] in, represents the predicted value after fusion, is the linear prediction result, is the nonlinear prediction result, and are the weights of linear and nonlinear prediction results respectively, and satisfy . The selection of weights is determined based on the prediction accuracy of the model, verified by historical data, and the weight combination that minimizes the error is selected to ensure the accuracy of the fusion result. The historical power grid electricity price data is obtained, and the electricity price is predicted based on the short-term load demand forecast value and the solar power generation forecast value to obtain the predicted power grid electricity price data; based on the obtained short-term load demand forecast value, solar power generation forecast value and predicted power grid electricity price data, the energy balance constraint equation is constructed, and iterative calculation is performed to obtain the final energy allocation plan. The energy balance constraint equation describes the supply and demand balance relationship between different energy sources in the system, which is used to ensure that each energy source can be reasonably allocated under the premise of meeting the load demand. The energy balance constraint equation establishes a balance relationship between load demand and available energy. In order to find the best energy allocation plan, the iterative optimization method is used to adjust the output of each energy source so that the entire system can operate in the best way. In the iterative process, the predicted load demand and power generation data are used to gradually adjust the charging and discharging strategy of the battery energy storage system to achieve optimal management of system energy.

[0074] In a specific embodiment, the process of executing step 300 may specifically include the following steps:

[0075] Based on the energy allocation scheme, the voltage, current and power of the hybrid inverter system are discretized to obtain the system state variables and control variables;

[0076] Using the system state variables and control variables, the state equation and output equation of the hybrid inverter are constructed to obtain a discrete time state space model;

[0077] According to the discrete time state space model, the prediction interval and control interval of the rolling time domain optimization are set, and based on the prediction interval and control interval, a multi-objective optimization problem including minimizing power balance deviation, maintaining battery state of charge, and minimizing inverter loss is established;

[0078] The multi-objective optimization problem is weighted and converted into a single-objective optimization problem using the penalty function method to obtain a standardized optimization model. The standardized optimization model is then substituted into the quadratic programming solver for iterative calculation to obtain the optimal control sequence at the current moment.

[0079] Extracting the first control quantity from the optimal control sequence at the current moment as the control instruction at the current moment, and applying it to the hybrid inverter system to obtain the real-time power flow distribution result;

[0080] The real-time power flow allocation result is compared with the preset threshold. If it exceeds the threshold, re-optimization is triggered, otherwise it is rolled to the next moment to finally obtain the optimal power flow instruction.

[0081] Specifically, based on the energy allocation scheme, the voltage, current and power of the system are discretized to obtain the state variables and control variables of the system. The voltage, current and power signals are sampled and quantized into discrete data points on the time axis. These discrete data points are used to describe the operating state of the system at different time steps. Through discretization, the state variables of the system are obtained. and control variables ,in Represents a discrete time step. Based on the state variables and control variables of the system, the state equation and output equation of the hybrid inverter are constructed to obtain a discrete time state space model. The state space model can systematically describe the dynamic behavior of the hybrid inverter. The state equation describes the evolution of the system state over time, while the output equation reflects the relationship between the system output and the state and input. The state equation is expressed as:

[0082] ;

[0083] in, represents the state vector of the system, such as voltage and current, is the state transfer matrix, which describes the internal state changes of the system. To control the input vector, such as the switching control signal of the hybrid inverter, is the input matrix, describing the influence of control input on state change. The output equation is expressed as:

[0084] ;

[0085] in, is the output quantity of the system, such as output voltage and output power, is the output matrix, describing the influence of system state on output, is a direct transfer matrix that describes the direct impact of input on output. These state and output equations constitute the discrete time state space model of the hybrid inverter. According to the discrete time state space model, the prediction interval and control interval of the rolling horizon optimization are set. Rolling horizon optimization is a control strategy that predicts and optimizes the system behavior in the future at each time step, so as to make the best control decision at the current moment. When setting the rolling horizon optimization, the prediction interval determines the length of time required for prediction in the future, while the control interval determines the time step for actual implementation of control. Based on the prediction interval and the control interval, a multi-objective optimization problem is established to achieve the optimal control of the system. In this multi-objective optimization problem, multiple objectives are included, such as minimizing the power balance deviation, maintaining the battery state of charge (SOC), and minimizing the inverter loss. The goal of minimizing the power balance deviation is to ensure that the hybrid inverter system can maintain the power balance between power generation, energy storage and load under different load demands, so as to achieve stable operation of the system. The maintenance of the battery state of charge is to prevent the battery from overcharging or over-discharging, thereby extending the battery life; while the minimization of the inverter loss is aimed at improving the energy utilization efficiency of the system and reducing energy loss. In multi-objective optimization, it is expressed as the following objective function:

[0086] ;

[0087] in, is the overall optimization objective function, is the power balance deviation target, Maintaining goals for the SOC, is the inverter loss target, is the weight coefficient of each objective, which is used to reflect the relative importance of each objective in the overall optimization. Weights are assigned to each objective, and the multi-objective optimization problem is converted into a single-objective optimization problem through the penalty function method. The penalty function method introduces constraints into the optimization of the objective function by imposing penalties on solutions that do not meet the constraints. After the conversion is completed, a standardized optimization model is obtained, which is substituted into the quadratic programming (QP) solver for solution. Quadratic programming is an optimization method that is suitable for dealing with optimization problems with quadratic objective functions and linear constraints. Through iterative calculation, the optimal control sequence at the current moment is obtained. After obtaining the optimal control sequence, the first control quantity is extracted from it as the control instruction at the current moment, and it is applied to the hybrid inverter system to achieve real-time power flow distribution. For example, if the optimal control sequence represents the switching signal of the inverter or the control quantity for adjusting the voltage and frequency, then it is applied to the system at the current time step to adjust the output of the hybrid inverter to meet the current power demand. The real-time power flow distribution result reflects the operating status of the system at the current moment and the energy flow between the various subsystems. In order to ensure the safety and reliability of the system operation, the real-time power flow distribution results are compared with the preset thresholds. If the power flow distribution results exceed the set threshold range, it means that the current control scheme cannot meet the system operation requirements, or some unpredictable disturbances occur. At this time, the re-optimization process is triggered to re-evaluate the system operation status and calculate a new optimal control scheme. If the power flow distribution results do not exceed the threshold range, it means that the current control scheme is feasible, and the system continues to move forward according to the prediction and control steps, thereby rolling to the next moment, continuing the next round of prediction and control, and finally obtaining the optimal power flow instruction.

[0088] In a specific embodiment, the process of executing step 400 may specifically include the following steps:

[0089] Extracting a first target power value of the solar power generation system and a second target power value of the battery energy storage system from the optimal power flow instruction to obtain a system power allocation target;

[0090] Based on the solar panel output data sequence, an IV characteristic curve model of the solar panel is constructed, and the IV characteristic curve model is differentiated to calculate the slope of the power-voltage curve to obtain the power change rate;

[0091] According to the power change rate, the working point of the solar panel is iteratively adjusted using the perturbation observation method to obtain the voltage and current values ​​of the maximum power point, and the voltage and current values ​​of the maximum power point are multiplied to obtain the maximum power output value of the solar power generation system;

[0092] The battery state of charge is estimated based on the battery energy storage system status data set to obtain the current battery SOC value, and the optimal trajectory of the battery charging and discharging power is calculated using a dynamic programming algorithm based on the current battery SOC value and the second target power value of the battery energy storage system to obtain the optimal charging and discharging power of the battery energy storage system.

[0093] Specifically, the first target power value of the solar power generation system and the second target power value of the battery energy storage system are extracted from the optimal power flow instruction to obtain the power allocation target of the system. These power allocation targets guide the power output of the solar panels and the battery energy storage system to meet the load requirements. The IV characteristic curve model of the solar panel is constructed based on the output data sequence of the solar panel. The IV characteristic curve is a curve that describes the relationship between current and voltage of the solar panel under different light intensities and load conditions. The IV characteristic curve model is differentiated to obtain the slope of the power-voltage (PV) curve to calculate the power change rate. The power change rate is calculated by differentiating the power-voltage curve. Power With voltage The relationship between them is expressed as:

[0094] ;

[0095] in, is power, is the voltage, is the current. To obtain the power change rate, differentiate the above formula to obtain:

[0096] ;

[0097] in, It represents the rate of change of power to voltage. is the derivative of current with respect to voltage. When the power change rate is zero, that is, , at this time the solar panel operates at the maximum power point; when the power change rate is positive, it means that the power can continue to increase; when the power change rate is negative, it means that the power begins to decrease. By calculating the power change rate, the operating point is guided to move toward the maximum power point. Based on the power change rate, the perturbation observation method is used to iteratively adjust the operating point of the solar panel to find the voltage and current values ​​of the maximum power point. The perturbation observation method is a maximum power point tracking algorithm that applies a small disturbance to the working voltage and then observes the changes in power. If the power increases, it means that the disturbance direction is correct, and continue to adjust in this direction; if the power decreases, adjust in the opposite direction. Through continuous perturbation and observation, the maximum power point of the solar panel is found. After obtaining the voltage at the maximum power point and current The maximum power output value of the solar power generation system is calculated by the following formula:

[0098] ;

[0099] in, Indicates the maximum power output value, and They are the voltage and current at the maximum power point, respectively. The maximum power output value is used to guide the energy output of solar panels under different conditions, thereby maximizing the utilization efficiency of solar energy resources. At the same time, the battery energy storage system is controlled to achieve system power balance and effective utilization of the energy storage system. Based on the battery energy storage system status data set, the battery state of charge (SOC) is estimated. SOC is an important parameter that describes the current state of charge of the battery, usually expressed as a percentage, reflecting the remaining power of the battery. The estimation of SOC is achieved through the current integration method, and the formula is as follows:

[0100] ;

[0101] in, Indicates the current state of charge. is the SOC at the previous moment, is the rated capacity of the battery, is the charge and discharge current of the battery. Through this integration process, the current SOC value of the battery is obtained. According to the current SOC value of the battery and the second target power value of the battery energy storage system extracted from the optimal power flow instruction, the dynamic programming algorithm is used to calculate the optimal trajectory of the battery charge and discharge power. Dynamic programming is an effective method for solving multi-stage decision-making problems. It decomposes the problem into a series of sub-problems, and obtains the optimal solution to the entire problem by gradually solving each sub-problem. In battery charge and discharge control, the goal of the dynamic programming algorithm is to select the optimal charge and discharge strategy based on the current SOC state, so that the battery can extend the battery life and improve the energy utilization efficiency as much as possible while meeting the system power requirements. Assuming that the goal is to minimize energy loss and extend battery life during charging and discharging, the objective function is defined as:

[0102] ;

[0103] in, To optimize the objective function, For the The charge and discharge power of a time step, is the target state of charge, For the The SOC value of the time step, and are weight coefficients, which are used to balance the impact of energy loss and SOC deviation. Through dynamic programming, a set of charging and discharging power , so that the objective function Minimize and obtain the optimal charging and discharging power of the battery energy storage system.

[0104] In a specific embodiment, the process of executing step 500 may specifically include the following steps:

[0105] The maximum power output value and the optimal charge and discharge power are input into the power buffer, the power fluctuation is smoothed, and a smoothed power instruction is obtained;

[0106] Based on the smoothed power command, the target output voltage and frequency of the hybrid inverter are calculated to obtain the initial droop control parameters, and the output voltage and frequency of the hybrid inverter are measured in real time to obtain the actual output value;

[0107] The actual output value is compared with the target output voltage and frequency, the deviation value is calculated, and the control error is obtained. Based on the control error, the initial droop control parameters are adjusted online using a fuzzy logic controller to obtain an adaptive droop coefficient.

[0108] According to the adaptive droop coefficient, the droop control equation is updated, and the corrected output voltage and frequency adjustment amount are calculated. At the same time, the corrected output voltage and frequency adjustment amount are substituted into the power balance equation, and the output voltage and frequency setting values ​​are obtained by iterative calculation.

[0109] The output current of the hybrid inverter is adjusted to the power grid based on the output voltage and frequency setting values, and a coordinated control instruction of the hybrid inverter is generated, the coordinated control instruction including the voltage amplitude, phase angle and frequency adjustment amount.

[0110] Specifically, the maximum power output value and the optimal charge and discharge power of the battery energy storage system are input into the power buffer for processing. The main function of the power buffer is to smooth the power fluctuations to reduce the power disturbance caused by the rapid changes in load and power generation conditions in the system. This smoothing process is achieved through a low-pass filter, which eliminates the high-frequency power fluctuations and only retains the low-frequency part to obtain a more stable power command. Suppose the power command is The smooth power signal helps the subsequent control system to operate stably and avoids the adverse effects of rapid power fluctuations on the safety and control effect of the system. Calculate the target output voltage and frequency of the hybrid inverter. Based on droop control, automatic power distribution between different power sources is achieved by adjusting the droop characteristics of voltage and frequency. Calculating the target output voltage of the hybrid inverter and frequency , specifically expressed as:

[0111] ;

[0112] ;

[0113] in, and are the target output voltage and frequency respectively, and is the reference voltage and frequency when no load is applied, and is the droop control coefficient, is the smoothed active power command, is the smoothed reactive power command (usually obtained through similar smoothing). The initial droop control parameters are calculated to determine the target operating point of the hybrid inverter. The actual output voltage and frequency of the hybrid inverter are measured in real time to obtain the current output value. and . Compare the actual output value with the target output voltage and frequency Compare and calculate the deviation between the two to get the control error. The control error is expressed as:

[0114] ;

[0115] ;

[0116] in, and are the voltage and frequency errors respectively. Based on the control error, the fuzzy logic controller is used to adjust the initial droop control parameters online to obtain the adaptive droop coefficient. The fuzzy logic controller adjusts the droop coefficient online by reasoning the fuzzy rules of the error and the error change, so that the system can dynamically adjust the control strategy according to the actual operating state to adapt to different operating conditions and improve the robustness and response speed of the system. According to the adaptive droop coefficient, the droop control equation is updated to obtain the corrected output voltage and frequency adjustment. The corrected droop control equation is expressed as:

[0117] ;

[0118] ;

[0119] in, and are the corrected output voltage and frequency respectively, and The output voltage and frequency are adjusted to control the operation of the hybrid inverter, so that it can maintain stable operation under different load and power generation conditions. and frequency Substitute into the power balance equation and perform iterative calculation to obtain the final output voltage and frequency setting values. The power balance equation is used to ensure the balance of active and reactive power supply and demand in the system, and continuously adjust the output voltage and frequency through iteration until the power balance condition is met. For example, the power balance equation is expressed as:

[0120] ;

[0121] in, is the power required by the load, and are the power outputs of the solar power generation system and the battery energy storage system respectively. or , ensuring that the load demand is met and the system is in a balanced state. Based on the final output voltage and frequency setting values, the hybrid inverter output current is adjusted to the grid, and a coordinated control instruction for the hybrid inverter is generated. The coordinated control instruction contains voltage amplitude, phase angle, and frequency adjustment, which are used to actually guide the operation of the hybrid inverter. For example, the voltage amplitude , Phase Angle and frequency adjustment Set it up in the following form:

[0122]

[0123] in, It is a reference phase angle, usually used to ensure the synchronization of the hybrid inverter with other power sources. By coordinating control instructions, the hybrid inverter can be stably controlled in actual operation so that its output meets the system's requirements for voltage, frequency and phase, thereby ensuring the stability of the system and the quality of power.

[0124] In a specific embodiment, the process of executing step 600 may specifically include the following steps:

[0125] The voltage amplitude and phase angle in the coordinated control command are transformed by Clarke, and the three-phase stationary coordinate system is converted into a two-phase stationary coordinate system to obtain the voltage vector in the αβ coordinate system.

[0126] Based on the voltage vector in the αβ coordinate system, determine the sector where the voltage space vector is located, and calculate the action time of two adjacent basic vectors to obtain the vector synthesis time;

[0127] According to the vector synthesis time, the on-time of each switch tube is calculated by using the symmetrical sinusoidal pulse width modulation algorithm to obtain the initial switch sequence, and the dead time compensation is performed on the initial switch sequence to eliminate the shoot-through phenomenon during the on-off process of the switch tube, and obtain the compensated switch sequence;

[0128] Based on the compensated switching sequence, the harmonic content of the hybrid inverter output voltage is calculated, and the high-order harmonics are suppressed using the selective harmonic elimination algorithm to obtain the optimized switching sequence;

[0129] The optimized switching sequence is compared with the triangular carrier to generate a pulse width modulation waveform to obtain a switching control sequence, and the voltage and current at the output of the hybrid inverter are sampled to calculate the actual active power and reactive power of the system to obtain a power calculation result;

[0130] Based on the power calculation result and the preset power factor, the required compensation current is calculated using the instantaneous reactive power theory to obtain the reactive compensation value, which is then combined with the switch control sequence to generate a real-time control signal for the hybrid inverter.

[0131] Specifically, the voltage amplitude and phase angle in the coordinated control instruction are transformed by Clarke transformation, and the three-phase stationary coordinate system is transformed into a two-phase stationary coordinate system to obtain the voltage vector in the αβ coordinate system. Clarke transformation is a coordinate transformation method used to transform the quantities (such as voltage and current) of a three-phase system into two-phase quantities, thereby simplifying the complex three-phase analysis problem into a vector operation on a two-dimensional plane. In the three-phase stationary coordinate system, the three-phase voltages are , and , after Clarke transformation, the voltage vector in the αβ coordinate system is and Calculated by the following formula:

[0132] ;

[0133] ;

[0134] in, and are the voltage components on the α-axis and β-axis respectively, and these components together constitute the voltage vector in the αβ coordinate system. The result of this transformation maps the voltage components in the three-phase stationary coordinate system to a two-dimensional plane, which is convenient for calculation in voltage space vector modulation. Based on the voltage vector in the aβ coordinate system, the sector where the voltage space vector is located is determined. Voltage space vector modulation controls the output voltage and frequency of the inverter by synthesizing the target voltage vector on the space vector plane. In order to synthesize the target voltage vector, determine the sector where the vector is located. On the αβ plane, the voltage space is divided into six sectors, each sector representing a different basic vector combination. By calculating the voltage vector and The angle of the target voltage vector is determined to determine in which sector the target voltage vector is located. After determining the sector where the voltage vector is located, the action time of two adjacent basic vectors is calculated to synthesize the target voltage vector. Assume that the target voltage vector is located in the sector In this case, the two basic vectors of the sector are used and and zero vector The action time of the basic vector is calculated by the following formula:

[0135] ;

[0136] ;

[0137] in, and are the action times of the two basic vectors, is the DC bus voltage, and are the a-axis and β-axis components of the target voltage vector. The target voltage vector is synthesized by the action time of these two basic vectors and the action time of the zero vector, and the vector synthesis time is obtained. According to the calculated vector synthesis time, the conduction time of each switch tube is calculated by the symmetrical sinusoidal pulse width modulation algorithm to obtain the initial switch sequence. The symmetrical sinusoidal pulse width modulation algorithm controls the output voltage of the inverter by adjusting the conduction time of each switch. Since there is a delay in the actual operation of the inverter switch device, in order to avoid the short circuit problem caused by the simultaneous conduction of the switch tube, the initial switch sequence is compensated for the dead time. The purpose of dead time compensation is to set a time interval between two relative switch tubes to ensure that one switch is completely turned off before the other switch is turned on, so as to eliminate the direct-through phenomenon during the conduction and shutdown process. Through this compensation process, the compensated switch sequence is obtained, thereby ensuring the safe operation of the inverter. Based on the compensated switch sequence, the harmonic content of the output voltage of the hybrid inverter is calculated, and the selective harmonic elimination algorithm is used to suppress the high-order harmonics. Selective harmonic elimination is a modulation technology that eliminates specific harmonics by precisely selecting a switching sequence. It optimizes the switching angle so that the harmonic component of a specific frequency is zero. By suppressing high-order harmonics, the output power quality of the hybrid inverter is significantly improved, making the output voltage closer to an ideal sine wave, thereby reducing harmonic interference to the load. After harmonic suppression, an optimized switching sequence is obtained. The optimized switching sequence is compared with the triangular carrier to generate a pulse width modulation (PWM) waveform to obtain the final switching control sequence. The PWM waveform is used to control the on and off states of each switch tube of the inverter, thereby achieving precise regulation of the output voltage and frequency. The voltage and current at the output of the hybrid inverter are sampled in real time, and the sampled data is used to calculate the actual active power and reactive power of the system. Active power and reactive power The calculation is performed by the following formula:

[0138] ;

[0139] ;

[0140] in, and are the direct-axis component and quadrature-axis component of the voltage, respectively. and They are the direct-axis component and quadrature-axis component of the current, respectively. By calculating the dq components of the voltage and current, the active and reactive power of the system are obtained. Based on the calculation results of the active power and reactive power, and the preset power factor, the instantaneous reactive power theory is used to calculate the required compensation current to obtain the reactive compensation value. The instantaneous reactive power theory is used to analyze the active and reactive components in the three-phase system in real time. It determines the required compensation amount by calculating the instantaneous active and reactive power to achieve the purpose of power factor correction. The output of the inverter is adjusted by the calculated reactive compensation value, so that the reactive power of the system is close to zero, improving the power factor and the power utilization rate of the system. The reactive compensation value is combined with the optimized switch control sequence to generate a real-time control signal for the hybrid inverter. The control signal contains all the necessary information to control the switch tube of the inverter to achieve precise regulation of voltage, frequency and power factor.

[0141] The hybrid inverter control method in the embodiment of the present application is described above. The hybrid inverter control device in the embodiment of the present application is described below. Figure 2 In the embodiment of the present application, an embodiment of the hybrid inverter control device includes:

[0142] The acquisition module 11 is used to collect the solar panel output data, the battery energy storage system status data and the load demand data, and to build a system dynamic model;

[0143] A prediction module 12 is used to predict short-term load demand, solar power generation and grid electricity charges based on a system dynamic model to obtain an energy allocation plan;

[0144] The calculation module 13 is used to calculate the power balance relationship between the solar power generation system, the battery energy storage system and the load according to the energy allocation plan to obtain the optimal power flow instruction;

[0145] The processing module 14 is used to calculate the maximum power output value of the solar power generation system and the optimal charging and discharging power of the battery energy storage system based on the optimal power flow instruction;

[0146] The regulating module 15 is used to regulate the output current of the hybrid inverter to the power grid based on the maximum power output value and the optimal charging and discharging power, and generate a coordinated control instruction;

[0147] The generating module 16 is used to generate a switch control sequence according to the coordinated control instruction, and calculate the reactive power compensation value to obtain a real-time control signal of the hybrid inverter.

[0148] Through the synergy of the above components, the accuracy and prediction precision of the system dynamic model are improved by using the recursive least squares method to estimate the system parameters, combining the autoregressive moving average model and the support vector machine algorithm to predict the short-term load and power generation. The discrete time state space model is constructed by using the rolling time domain optimization algorithm to achieve the dynamic power balance between the solar power generation system, the battery energy storage system and the load, and optimize the energy distribution efficiency of the system. The maximum power output value of the solar power generation system and the optimal charge and discharge power of the battery energy storage system are calculated respectively, which improves the energy utilization efficiency and battery life. Combined with the power buffer and the fuzzy logic controller, the precise adjustment of the output voltage and frequency of the hybrid inverter is achieved, and the dynamic response capability and stability of the system are enhanced. The space vector pulse width modulation technology is used to generate the switch control sequence, and combined with the selective harmonic elimination algorithm and dead time compensation, the harmonic content of the output voltage is significantly reduced, and the power quality is improved. By calculating the reactive compensation value in real time and combining it with the switch control sequence, the dynamic reactive power compensation of the hybrid inverter is achieved, and the power factor and grid friendliness of the system are improved.

[0149] The present application also provides an electronic device, which includes a memory and a processor, wherein the memory stores computer-readable instructions, and when the computer-readable instructions are executed by the processor, the processor executes the steps of the hybrid inverter control method in the above-mentioned embodiments.

[0150] The present application also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium. Instructions are stored in the computer-readable storage medium. When the instructions are executed on a computer, the computer executes the steps of the hybrid inverter control method.

[0151] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0152] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions to enable an electronic device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, etc., various media that can store program codes.

[0153] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A hybrid inverter control method, characterized in that: The method comprises: Collect solar panel output data, battery energy storage system status data and load demand data, and build a system dynamic model; Based on the system dynamic model, short-term load demand, solar power generation and grid electricity charges are predicted to obtain an energy allocation plan; Calculate the power balance relationship between the solar power generation system, the battery energy storage system and the load according to the energy allocation scheme to obtain the optimal power flow instruction; Based on the optimal power flow instruction, respectively calculating the maximum power output value of the solar power generation system and the optimal charging and discharging power of the battery energy storage system; Based on the maximum power output value and the optimal charge and discharge power, the output current of the hybrid inverter is adjusted to the power grid, and a coordinated control instruction is generated; specifically, the maximum power output value and the optimal charge and discharge power are input into a power buffer, the power fluctuation is smoothed, and a smoothed power instruction is obtained; based on the smoothed power instruction, the target output voltage and frequency of the hybrid inverter are calculated to obtain an initial droop control parameter, and the output voltage and frequency of the hybrid inverter are measured in real time to obtain an actual output value; the actual output value is compared with the target output voltage and frequency, a deviation value is calculated, a control error is obtained, and Based on the control error, the initial droop control parameter is adjusted online by using a fuzzy logic controller to obtain an adaptive droop coefficient; according to the adaptive droop coefficient, the droop control equation is updated, and the corrected output voltage and frequency adjustment amount are calculated, and at the same time, the corrected output voltage and frequency adjustment amount are substituted into the power balance equation, and iterative calculation is performed to obtain the output voltage and frequency setting value; based on the output voltage and the frequency setting value, the output current of the hybrid inverter is adjusted to the power grid, and a coordinated control instruction of the hybrid inverter is generated, and the coordinated control instruction includes a voltage amplitude, a phase angle and a frequency adjustment amount; Generate a switch control sequence according to the coordinated control instruction, calculate the reactive compensation value, and obtain a real-time control signal of the hybrid inverter; specifically include: perform Clarke transformation on the voltage amplitude and phase angle in the coordinated control instruction, convert the three-phase stationary coordinate system into a two-phase stationary coordinate system, and obtain the voltage vector in the αβ coordinate system; based on the voltage vector in the αβ coordinate system, determine the sector where the voltage space vector is located, and calculate the action time of two adjacent basic vectors to obtain the vector synthesis time; according to the vector synthesis time, use a symmetrical sinusoidal pulse width modulation algorithm to calculate the conduction time of each switch tube to obtain an initial switch sequence, and perform dead time compensation on the initial switch sequence to eliminate the through phenomenon during the conduction and shutdown of the switch tube The invention relates to a method for controlling the output voltage of the hybrid inverter and obtaining a compensated switching sequence; based on the compensated switching sequence, calculating the harmonic content of the output voltage of the hybrid inverter, and suppressing high-order harmonics by using a selective harmonic elimination algorithm to obtain an optimized switching sequence; comparing the optimized switching sequence with a triangular carrier to generate a pulse width modulation waveform to obtain a switching control sequence, and sampling the voltage and current at the output end of the hybrid inverter to calculate the actual active power and reactive power of the system to obtain a power calculation result; based on the power calculation result and a preset power factor, calculating the required compensation current by using the instantaneous reactive power theory to obtain a reactive compensation value, and combining the reactive compensation value with the switching control sequence to generate a real-time control signal for the hybrid inverter.

2. The hybrid inverter control method according to claim 1, characterized in that: The collecting of solar panel output data, battery energy storage system status data and load demand data, and building of a system dynamic model, includes: The output voltage and current of the solar panel are sampled to obtain the solar panel output data sequence; the terminal voltage, charge and discharge current and battery temperature of the battery energy storage system are measured to obtain the battery energy storage system status data set; the active power and reactive power at the load end are monitored in real time to obtain the load demand data set; Respectively performing denoising and standardization processing on the solar panel output data sequence, the battery energy storage system state data set and the load demand data set to obtain preprocessed data sets; Performing time series analysis on the preprocessed data set to obtain a corresponding relationship between system input and output, and based on the corresponding relationship between system input and output, constructing an objective function of a recursive least squares algorithm, setting initial parameter values ​​and a forgetting factor, and obtaining a parameter estimation iteration formula; Based on the parameter estimation iteration formula, an online parameter identification calculation is performed to obtain updated system parameter values, and the updated system parameter values ​​are used to construct the state equation and output equation of the hybrid inverter system to obtain a system dynamic model.

3. The hybrid inverter control method according to claim 2, characterized in that: The method of predicting short-term load demand, solar power generation and grid electricity charges based on the system dynamic model to obtain an energy allocation plan includes: According to the system dynamic model, time series decomposition is performed on the load demand data set and the solar panel output data sequence to obtain trend items, period items and random items; Performing differential processing on the trend item to obtain a stationary sequence, and performing autoregressive moving average analysis based on the stationary sequence to obtain a linear prediction result of load demand and solar power generation; Inputting the periodic term and the random term into a support vector machine model for nonlinear feature learning to obtain nonlinear prediction results of load demand and solar power generation; Performing weighted fusion on the linear prediction result and the nonlinear prediction result to obtain a short-term load demand prediction value and a solar power generation prediction value; Acquire historical power grid electricity fee data, and perform electricity fee forecasting based on the short-term load demand forecast value and the solar power generation forecast value to obtain forecast power grid electricity fee data; Based on the short-term load demand forecast value, the solar power generation forecast value and the forecast power grid electricity fee data, an energy balance constraint equation is constructed and iterative calculation is performed to obtain an energy allocation plan.

4. The hybrid inverter control method according to claim 1, characterized in that: The step of calculating the power balance relationship between the solar power generation system, the battery energy storage system and the load according to the energy allocation scheme to obtain the optimal power flow instruction includes: Based on the energy allocation scheme, the voltage, current and power of the hybrid inverter system are discretized to obtain system state variables and control variables; Using the system state variables and the control variables, constructing a state equation and an output equation of the hybrid inverter to obtain a discrete time state space model; According to the discrete-time state-space model, a prediction interval and a control interval of a rolling time domain optimization are set, and based on the prediction interval and the control interval, a multi-objective optimization problem including minimization of power balance deviation, maintenance of battery state of charge, and minimization of inverter loss is established; The multi-objective optimization problem is weighted and converted into a single-objective optimization problem using a penalty function method to obtain a standardized optimization model, and the standardized optimization model is substituted into a quadratic programming solver for iterative calculation to obtain an optimal control sequence at the current moment; Extracting a first control quantity from the optimal control sequence at the current moment as a control instruction at the current moment, and applying the first control quantity to the hybrid inverter system to obtain a real-time power flow distribution result; The real-time power flow allocation result is compared with a preset threshold value. If it exceeds the threshold value, re-optimization is triggered, otherwise it is rolled to the next moment to finally obtain the optimal power flow instruction.

5. The hybrid inverter control method according to claim 1, characterized in that: The calculating, based on the optimal power flow instruction, the maximum power output value of the solar power generation system and the optimal charging and discharging power of the battery energy storage system respectively include: Extracting a first target power value of the solar power generation system and a second target power value of the battery energy storage system from the optimal power flow instruction to obtain a system power allocation target; Based on the solar panel output data sequence, an IV characteristic curve model of the solar panel is constructed, and a differential operation is performed on the IV characteristic curve model to calculate the slope of the power-voltage curve to obtain a power change rate; According to the power change rate, the working point of the solar panel is iteratively adjusted by using a disturbance observation method to obtain the voltage and current values ​​of the maximum power point, and the voltage and current values ​​of the maximum power point are multiplied to obtain the maximum power output value of the solar power generation system; The battery state of charge is estimated based on the battery energy storage system status data set to obtain the current battery SOC value, and the optimal trajectory of the battery charging and discharging power is calculated using a dynamic programming algorithm according to the current battery SOC value and the second target power value of the battery energy storage system to obtain the optimal charging and discharging power of the battery energy storage system.

6. A hybrid inverter control device, characterized in that: The hybrid inverter control device comprises: The acquisition module is used to collect the solar panel output data, the battery energy storage system status data and the load demand data, and build a system dynamic model; A prediction module, used for predicting short-term load demand, solar power generation and grid electricity charges based on the system dynamic model to obtain an energy allocation plan; A calculation module, used to calculate the power balance relationship between the solar power generation system, the battery energy storage system and the load according to the energy allocation scheme, and obtain an optimal power flow instruction; A processing module, used to calculate the maximum power output value of the solar power generation system and the optimal charging and discharging power of the battery energy storage system respectively based on the optimal power flow instruction; The regulating module is used to regulate the output current of the hybrid inverter to the power grid based on the maximum power output value and the optimal charge and discharge power, and generate a coordinated control instruction; specifically comprising: inputting the maximum power output value and the optimal charge and discharge power into a power buffer, smoothing the power fluctuation, and obtaining a smoothed power instruction; based on the smoothed power instruction, calculating the target output voltage and frequency of the hybrid inverter, obtaining an initial droop control parameter, and measuring the output voltage and frequency of the hybrid inverter in real time to obtain an actual output value; comparing the actual output value with the target output voltage and frequency, calculating a deviation value, and obtaining a control error. difference, and based on the control error, use a fuzzy logic controller to adjust the initial droop control parameter online to obtain an adaptive droop coefficient; according to the adaptive droop coefficient, update the droop control equation, and calculate the corrected output voltage and frequency adjustment amount, and at the same time, substitute the corrected output voltage and frequency adjustment amount into the power balance equation, perform iterative calculation, and obtain the output voltage and frequency setting value; based on the output voltage and the frequency setting value, adjust the output current of the hybrid inverter to the power grid, and generate a coordinated control instruction of the hybrid inverter, the coordinated control instruction includes a voltage amplitude, a phase angle and a frequency adjustment amount; A generation module is used to generate a switch control sequence according to the coordinated control instruction, calculate the reactive power compensation value, and obtain the real-time control signal of the hybrid inverter; specifically comprising: performing Clarke transformation on the voltage amplitude and phase angle in the coordinated control instruction, converting the three-phase stationary coordinate system into a two-phase stationary coordinate system, and obtaining the voltage vector in the αβ coordinate system; based on the voltage vector in the αβ coordinate system, determining the sector where the voltage space vector is located, and calculating the action time of two adjacent basic vectors to obtain the vector synthesis time; according to the vector synthesis time, using a symmetrical sinusoidal pulse width modulation algorithm to calculate the conduction time of each switch tube to obtain an initial switch sequence, and performing dead time compensation on the initial switch sequence to eliminate the dead time during the conduction and shutdown of the switch tube The shoot-through phenomenon is eliminated to obtain a compensated switching sequence; based on the compensated switching sequence, the harmonic content of the output voltage of the hybrid inverter is calculated, and the high-order harmonics are suppressed by using a selective harmonic elimination algorithm to obtain an optimized switching sequence; the optimized switching sequence is compared with a triangular carrier to generate a pulse width modulation waveform to obtain a switching control sequence, and the voltage and current at the output end of the hybrid inverter are sampled to calculate the actual active power and reactive power of the system to obtain a power calculation result; based on the power calculation result and a preset power factor, the required compensation current is calculated using the instantaneous reactive power theory to obtain a reactive compensation value, and the reactive compensation value is combined with the switching control sequence to generate a real-time control signal for the hybrid inverter.

7. An electronic device, characterized in that: The electronic device comprises: a memory and at least one processor, wherein instructions are stored in the memory; The at least one processor calls the instructions in the memory to enable the electronic device to execute the hybrid inverter control method according to any one of claims 1 to 5.

8. A computer-readable storage medium having instructions stored thereon, characterized in that: When the instructions are executed by a processor, the hybrid inverter control method according to any one of claims 1 to 5 is implemented.

Images