A method and system for controlling the operating state of a bidirectional photovoltaic energy storage inverter

Abstract

The application provides a running state control method and system of a bidirectional photovoltaic energy storage inverter. It belongs to the technical field of new energy power electronic control. The method comprises the following steps: obtaining initial hardware topology data and real-time working condition data of the bidirectional photovoltaic energy storage inverter, analyzing the initial hardware topology data, determining the reconfigurable hardware modules, configuring the reconfigurable hardware modules according to the real-time working condition data through a preset reconfigurable rule, and obtaining a dynamic hardware topology structure; constructing a real-time hardware model of the bidirectional photovoltaic energy storage inverter according to the dynamic hardware topology structure; through deep integration of the bidirectional reconfigurable hardware topology and multi-agent dynamic energy game, the bidirectional photovoltaic energy storage inverter can dynamically adjust the hardware structure and energy interaction strategy according to the real-time working condition, realizing a paradigm change from static control to dynamic self-adaption, and greatly improving the adaptability of the inverter to complex and changeable working conditions.

Description

Technical Field

[0001] This invention proposes a method and system for controlling the operating status of a bidirectional photovoltaic energy storage inverter, which belongs to the field of new energy power electronic control technology. Background Technology

[0002] Traditional photovoltaic (PV) energy storage inverter operation status control mostly adopts static control methods, which control the inverter based on preset parameters and a fixed hardware topology. However, in actual operation, the operating conditions of PV systems are complex and variable, with factors such as sunlight intensity, temperature, and load demand constantly changing. Static control methods are difficult to flexibly adjust according to real-time operating conditions, resulting in low inverter operating efficiency, significant energy loss, and an inability to fully realize the performance advantages of PV energy storage systems. Therefore, a bidirectional PV energy storage inverter control method that can dynamically adjust its operating status according to real-time operating conditions is needed. Summary of the Invention

[0003] This invention provides a method and system for controlling the operating status of a bidirectional photovoltaic energy storage inverter, in order to solve the problems mentioned in the background art above:

[0004] This invention proposes a method for controlling the operating state of a bidirectional photovoltaic energy storage inverter, the method comprising:

[0005] S1: Obtain the initial hardware topology data and real-time operating condition data of the bidirectional photovoltaic energy storage inverter; analyze the initial hardware topology data to determine its reconfigurable hardware modules; configure the reconfigurable hardware modules according to the preset reconfigurable rules based on the real-time operating condition data to obtain the dynamic hardware topology structure; construct the real-time hardware model of the bidirectional photovoltaic energy storage inverter based on the dynamic hardware topology structure.

[0006] S2: Divide the bidirectional photovoltaic energy storage inverter system into multiple intelligent agents, extract energy features for each intelligent agent, construct an energy game model for each intelligent agent based on the extracted energy features, solve the energy game model for each intelligent agent using a multi-agent dynamic energy game algorithm, and obtain the energy game decision results for each intelligent agent.

[0007] S3: Based on the real-time hardware model obtained in S1 and the energy game decision results of each intelligent agent obtained in S2, perform energy flow simulation of the bidirectional photovoltaic energy storage inverter; obtain equipment energy transmission simulation data of the bidirectional photovoltaic energy storage inverter under different operating conditions;

[0008] S4: Obtain the operation demand data of the bidirectional photovoltaic energy storage inverter. Based on the energy game decision results of each intelligent agent in S2 and the equipment energy transmission simulation data in S3, evaluate the operation demand data and analyze whether the current operation status meets the various requirements. If not, calculate the energy adjustment amount required to meet the operation requirements and obtain the operation demand energy assessment data.

[0009] S5: Based on the equipment energy transmission simulation data in S3 and the operation demand energy assessment data in S4, determine the operation status adjustment strategy of the bidirectional photovoltaic energy storage inverter; combined with the reconfigurable characteristics of the dynamic hardware topology in S1, further optimize the hardware structure according to the operation status adjustment strategy, and fine-tune the energy interaction strategy of each agent according to the result of the multi-agent dynamic energy game, and perform collaborative optimization of hardware structure evolution and energy flow intelligent game decision-making; transmit the collaboratively optimized operation status control strategy to the control unit of the bidirectional photovoltaic energy storage inverter to execute the operation status control task and realize the dynamic adaptive adjustment of the inverter operation status.

[0010] The present invention proposes an operation status control system for a bidirectional photovoltaic energy storage inverter, comprising:

[0011] One or more processors;

[0012] Memory, used to store one or more programs.

[0013] Wherein, when the one or more programs are executed by the one or more processors, the one or more processors implement the method described in any one of the above.

[0014] The beneficial effects of this invention are as follows: By deeply integrating bidirectional reconfigurable hardware topology with multi-agent dynamic energy game, the bidirectional photovoltaic energy storage inverter can dynamically adjust its hardware structure and energy interaction strategy according to real-time operating conditions, realizing a paradigm shift from static control to dynamic adaptation, and greatly improving the inverter's adaptability to complex and variable operating conditions.

[0015] The synergistic effect of real-time evolution of hardware structure and intelligent game-theoretic decision-making of energy flow can optimize energy transmission paths and distribution methods, reduce energy loss, and improve the overall operating efficiency of photovoltaic energy storage inverters.

[0016] Multi-agent dynamic energy game takes into account the interests and constraints of each agent, and can balance the energy relationship between photovoltaic array, energy storage battery and load, avoid system instability caused by energy supply and demand imbalance, and enhance the stability of the entire photovoltaic energy storage system.

[0017] By optimizing energy management and hardware structure, the number and depth of charge and discharge cycles of energy storage batteries can be reduced, and the working stress of hardware modules can be lowered, thereby extending the service life of bidirectional photovoltaic energy storage inverters and related equipment. Attached Figure Description

[0018] Figure 1 This is a diagram of the method described in this invention. Detailed Implementation

[0019] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0020] One embodiment of the present invention, such as Figure 1 As shown, a method for controlling the operating state of a bidirectional photovoltaic energy storage inverter, the method comprising:

[0021] S1: Acquire the initial hardware topology data and real-time operating condition data of the bidirectional photovoltaic energy storage inverter. The real-time operating condition data includes light intensity, temperature, and load demand information. Analyze the initial hardware topology data to determine its reconfigurable hardware modules, which include power switches, inductors, and capacitors. Based on the real-time operating condition data, configure the reconfigurable hardware modules according to preset reconfigurable rules to perform real-time evolution of the hardware structure, including multi-level / multi-port / redundancy mode switching, to obtain a dynamic hardware topology. Construct a real-time hardware model of the bidirectional photovoltaic energy storage inverter based on the dynamic hardware topology.

[0022] S2: The bidirectional photovoltaic energy storage inverter system is divided into multiple intelligent agents, including photovoltaic array intelligent agents, energy storage battery intelligent agents, and load intelligent agents. Energy features are extracted for each intelligent agent: for the photovoltaic array intelligent agent, its output power, voltage, and current are extracted; for the energy storage battery intelligent agent, its state of charge and charging / discharging power are extracted; for the load intelligent agent, its power demand and power consumption mode are extracted. Based on the extracted energy features, an energy game model for each intelligent agent is constructed, considering the energy gains, costs, and constraints of each agent. The energy game model for each intelligent agent is solved using a multi-agent dynamic energy game algorithm to obtain the energy game decision results for each intelligent agent, i.e., the optimal strategy for each intelligent agent in energy interaction.

[0023] S3: Based on the real-time hardware model obtained in S1 and the energy game decision results of each intelligent agent obtained in S2, perform energy flow simulation of the bidirectional photovoltaic energy storage inverter; during the simulation, consider the impact of dynamic changes in the hardware structure on energy transmission and the role of energy game decisions of each intelligent agent on energy allocation; obtain equipment energy transmission simulation data of the bidirectional photovoltaic energy storage inverter under different operating conditions, the transmission simulation data includes parameters such as power, voltage and current of each port;

[0024] S4: Obtain the operation requirement data of the bidirectional photovoltaic energy storage inverter, including system efficiency requirements, power quality requirements, and energy storage battery life protection requirements; based on the energy game decision results of each agent in S2 and the equipment energy transmission simulation data in S3, evaluate the operation requirement data and analyze whether the current operating state meets the various requirements; if not, calculate the energy adjustment amount required to meet the operation requirements and obtain the operation requirement energy assessment data;

[0025] S5: Based on the equipment energy transmission simulation data in S3 and the operation demand energy assessment data in S4, determine the operation state adjustment strategy for the bidirectional photovoltaic energy storage inverter; combining the reconfigurable characteristics of the dynamic hardware topology in S1, further optimize the hardware structure according to the operation state adjustment strategy. The optimization configuration includes adjusting the number of multi-level signals and the connection method of multi-ports; simultaneously, based on the results of the multi-agent dynamic energy game, fine-tune the energy interaction strategy of each agent, and perform collaborative optimization of hardware structure evolution and energy flow intelligent game decision-making; transmit the collaboratively optimized operation state control strategy to the control unit of the bidirectional photovoltaic energy storage inverter to execute the operation state control task and realize the dynamic adaptive adjustment of the inverter's operation state.

[0026] The working principle and effects of the above technical solution are as follows: By dynamically adjusting the hardware topology, it can maintain stable operation under changing conditions such as light, temperature, and load, thus improving the adaptability of the bidirectional photovoltaic energy storage inverter to complex operating conditions; the multi-agent energy game model makes the energy interaction between photovoltaic, energy storage, and load more efficient, reduces energy waste, and improves the rationality of system energy allocation; the synergistic optimization of hardware structure and energy strategy reduces unnecessary power loss and improves system operating efficiency.

[0027] By precisely controlling the charging and discharging process and protection strategies, the battery life is extended; the rate of energy storage battery wear is reduced; the switching of redundancy modes and dynamic adjustment mechanisms reduce failures caused by hardware overload or parameter imbalance; and the probability of system failure is reduced.

[0028] The switching between multi-level and multi-port modes can meet the energy transmission needs of different scenarios; it enhances the stability of power quality, accurately controls parameters such as voltage and current, and reduces waveform distortion and fluctuations; it ensures that the load can obtain a stable power supply under various operating conditions, reduces the risk of power outages, and enhances the reliability of power supply for users.

[0029] In one embodiment of the present invention, S1 includes:

[0030] S11. Obtain the initial hardware topology data of the bidirectional photovoltaic energy storage inverter. The initial hardware topology data includes static attribute data, which includes the power conversion unit circuit structure, component parameter specifications, and module connection relationships. Simultaneously collect real-time operating condition data, which includes dynamic parameters, such as illuminance (resolution not less than 10W / m²). 2 The data includes ambient temperature (accuracy ±0.5℃), load requirements (including instantaneous power, average power, and fluctuation frequency), and grid voltage / frequency; the collected data are preprocessed to construct a standardized dataset.

[0031] S12. Based on the initial hardware topology data, a modular decomposition algorithm is used, combined with graph theory and circuit topology analysis, to identify reconfigurable hardware modules. The reconfigurable hardware modules can be specifically divided into: power switching modules (including IGBT / MOSFET and drive circuits), inductor modules (including core materials and winding parameters), capacitor modules (including capacitance and withstand voltage levels), and interface conversion modules. A reconfigurability evaluation matrix for each module is established, and the reconfiguration modes it supports are marked. The reconfiguration modes include series / parallel switching and adjustable parameter ranges.

[0032] S13. Based on the preprocessed real-time operating data, a preset reconfigurable rule base is invoked. The reconfigurable rule base includes light-load matching rules and temperature-power limiting rules. Dynamic configuration is performed on the reconfigurable hardware modules. The dynamic configuration includes activating multi-level mode (increasing the output voltage level) under high light and high load, switching multi-port mode (expanding the energy interaction channel) when multiple sources are connected, and activating redundancy mode (switching to backup modules) when component fault warnings are issued. Real-time switching between modules is performed through the hardware configuration instruction set to generate a dynamic hardware topology.

[0033] S14. Based on the dynamic hardware topology and combined with the real-time parameters of each module, including the on-resistance of the switching transistor and the equivalent impedance of the inductor, a real-time hardware model of a bidirectional photovoltaic energy storage inverter based on state-space equations is constructed; the least squares method is used to calibrate the model parameters online to form a high-precision model that can be directly used for simulation analysis.

[0034] The working principle and effects of the above technical solution are as follows: By accurately identifying and dynamically configuring reconfigurable modules, the same set of hardware can adapt to different working conditions, reducing equipment idleness and improving the utilization efficiency of hardware topology; high-precision acquisition and standardized processing of dynamic parameters such as illumination and temperature provide a reliable basis for subsequent hardware configuration and improve the accuracy of data acquisition; after online calibration, the real-time hardware model built based on state-space equations can more realistically reflect the operating status of the equipment, providing strong support for system optimization and improving the accuracy of model simulation.

[0035] By flexibly reconfiguring modules to replace some of the new hardware requirements, the expenses for equipment procurement and replacement are reduced, thus lowering hardware investment costs. Dynamic configuration ensures that the hardware always works in a state that matches the current needs, reducing ineffective energy loss and energy consumption caused by improper adaptation to operating conditions. Preset rule bases automatically drive module switching, reducing the frequency of manual adjustments and the probability of errors, and reducing the workload of manual configuration.

[0036] The flexible switching between multi-level and multi-port modes can cope with various situations such as sudden changes in light intensity and load fluctuations, enhancing the system's adaptability to complex environments. The same module can achieve multiple functions through reconfiguration modes such as series / parallel connection, improving the comprehensive utilization value of the equipment and enhancing the reusability of hardware modules. The redundancy mode can switch to the backup module in time when a fault warning is issued, reducing the risk of system interruption due to single point of failure and enhancing the stability of system operation.

[0037] In one embodiment of the present invention, S2 includes:

[0038] S21. Using a distributed agent modeling method, the bidirectional photovoltaic energy storage inverter system is divided into three types of core agents: photovoltaic array agent, energy storage battery agent, and load agent. The photovoltaic array agent is used for light energy capture and conversion; the energy storage battery agent is used for energy storage and release; and the load agent is used for energy consumption and demand response. The communication boundaries (using the IEEE 1519 standard communication protocol) and energy interaction ranges (defining the power thresholds of the input / output ports) of each agent are determined.

[0039] S22. For the photovoltaic array intelligent agent, the first characteristic parameter is collected and extracted by the synchronous phasor measurement unit. The first characteristic parameter includes, but is not limited to, power fluctuation rate, voltage distortion rate, and maximum tracking point (MPPT) offset. For the energy storage battery intelligent agent, the second characteristic parameter is obtained by processing with the Kalman filter algorithm. The second characteristic parameter includes, but is not limited to, state of charge (SOC), state of health (SOH), charge and discharge efficiency curve, and cycle life decay rate. For the load intelligent agent, the third characteristic parameter is extracted by load characteristic analysis. The third characteristic parameter includes, but is not limited to, peak-valley difference in power demand, electricity consumption pattern label (such as resistive / inductive), and demand response sensitivity. Construct a time series database of characteristic parameters.

[0040] S23. Based on non-cooperative game theory, construct an energy game model for each agent; the energy game model includes: maximizing the power generation revenue of the photovoltaic array agent as the objective function, minimizing the charging and discharging cost of the energy storage battery agent as the objective function, and maximizing the electricity consumption satisfaction of the load agent as the objective function; introduce constraints, including the upper limit of photovoltaic power output, the safe range of battery SOC (usually 20%-80%), and the load power supply reliability requirements (power outage time ≤50ms); establish a multi-objective optimization game equilibrium equation;

[0041] S24. An improved particle swarm optimization algorithm is used to solve the energy game model. The strategy space of each agent is initialized. The strategy space includes the photovoltaic output adjustment step size, the battery charging and discharging power level, and the load demand adjustment coefficient. The Nash equilibrium point is found through iterative calculation. The optimal strategy set of each agent is output. The optimal strategy set includes the real-time output plan of the photovoltaic array, the charging and discharging sequence of the energy storage battery, and the load demand response scheme, forming the energy game decision result.

[0042] The working principle and effects of the above technical solution are as follows: The system is divided into different intelligent agents and a game theory model is constructed, making the energy coordination between photovoltaics, energy storage, and load more closely match actual needs, reducing waste caused by energy mismatch, and improving the rationality of energy allocation; Specialized acquisition and processing methods are used for different intelligent agents, such as the Kalman filter algorithm for analyzing battery state, making parameters more reflective of the true state, providing a reliable basis for decision-making, and improving the accuracy of feature parameter extraction; The improved particle swarm optimization algorithm can quickly find the equilibrium point, and the output optimal strategy set makes the operation of each part more coordinated, resulting in higher overall energy efficiency and improving the efficiency of strategy optimization.

[0043] By rationally planning the charging and discharging sequence through a game theory model, overcharging and discharging are avoided, the aging rate of batteries is slowed down, and the loss of energy storage batteries is reduced; a clear demand response scheme can adjust power consumption according to actual conditions, reducing the situation of fluctuating power supply and reducing the fluctuation of load power supply; autonomous interaction of intelligent agents and automatic solution of algorithms replace some manual adjustment, reducing human input and operational errors, and reducing the cost of human intervention.

[0044] Each intelligent agent cooperates within clearly defined communication boundaries and interaction ranges, meshing together like gears, improving the overall smoothness of operation and enhancing system synergy; the multi-objective optimization game model can find the optimal solution while satisfying multiple constraints, allowing the system to operate stably even under changes in lighting and load fluctuations, enhancing its ability to cope with complex working conditions; the demand response scheme of the load intelligent agent can better adapt to power consumption patterns, reducing the troubles caused by power outages or instability, and enhancing user satisfaction with power usage.

[0045] In one embodiment of the present invention, S24 includes:

[0046] Based on the energy game model constructed by S23, the policy space boundaries of each agent are determined, specifically including: the output adjustment step size range of the photovoltaic array agent (0-5% of rated power / step), the charging and discharging power levels of the energy storage battery agent (5 levels, each corresponding to 20% of rated power), and the demand adjustment coefficient range of the load agent (0.8-1.2). The parameters of the improved particle swarm optimization algorithm are initialized, including the particle population size (50-100), the maximum number of iterations (500 times), the initial inertia weight (0.9), and the learning factor (c1=c2=2.0), generating an initial particle swarm (each particle corresponds to a set of agent policy combinations).

[0047] For each particle (strategy combination) in the initial particle swarm, the objective function and constraints of the energy game model are substituted to calculate the quantitative values ​​of photovoltaic array power generation revenue, energy storage charging and discharging costs, and load power consumption satisfaction; it is checked whether the constraints such as photovoltaic power limit, battery SOC range, and load power supply reliability are met (particles that do not meet the constraints are given a penalty fitness); a weighted summation method (weights are set according to the importance of the objectives) is used to transform the multi-objective function into a single-objective fitness value, forming a particle fitness evaluation set;

[0048] Based on the particle fitness evaluation set, an iterative process of an improved particle swarm optimization algorithm is executed, introducing an adaptive inertia weight (which decreases linearly from 0.9 to 0.4 with each iteration) to balance global exploration and local exploitation capabilities. Based on the historical best position and the global best position of each particle, the particle velocity and position (strategy combination) are updated to ensure the new position falls within the policy space boundary. A mutation operation is performed every 50 iterations (randomly perturbing 10% of the particle positions) to avoid getting trapped in local optima, generating an updated particle swarm and corresponding fitness set.

[0049] During the iteration process, the change of the optimal fitness value of the population is monitored in real time. When the fluctuation range of the optimal fitness value is ≤0.1% for 30 consecutive iterations, the algorithm is considered to have converged, and the optimal position of the population at this time corresponds to the Nash equilibrium point. If it has not converged and the maximum number of iterations has been reached, a local search strategy (performing a fine search around the current optimal position) is adopted to further optimize until the convergence condition is met, and the strategy combination corresponding to the final Nash equilibrium point is determined.

[0050] From the strategy combinations corresponding to the Nash equilibrium point, the specific strategies of each agent are extracted: the real-time output plan of the photovoltaic array agent (including the power setpoint and MPPT adjustment command every 15 minutes), the charging and discharging sequence of the energy storage battery agent (including the start / end time of charging and discharging, and the power level switching node), and the demand response scheme of the load agent (including the power reduction ratio and time window of the adjustable load). These strategies are integrated into a structured optimal strategy set. After verifying the effectiveness of the strategies (substituting them into the model to verify whether all constraints and objectives are met), the final energy game decision result is formed.

[0051] The working principle and effects of the above technical solution are as follows: Clearly defining the policy space boundaries of each agent provides a clear scope for the optimization process; coupled with the iterative optimization of the improved particle swarm optimization algorithm, the found Nash equilibrium point better matches actual operational needs, improving the accuracy of policy solution; the combination of adaptive inertial weights and periodic mutation operations avoids blind search, allowing the optimal solution to emerge faster, reducing unnecessary computation time, and improving the algorithm's convergence speed; finally, the extracted specific policies undergo validity verification to ensure they meet all constraints, making them more reliable for implementation and improving policy feasibility.

[0052] The strategies of each intelligent agent are coordinated to avoid problems such as wasted photovoltaic power output and ineffective battery charging and discharging, thereby reducing system energy loss and the irrationality of energy allocation. The full-process control from strategy space setting to final verification makes the output optimal strategy set more reliable, reduces operational failures caused by decision bias, and lowers the probability of decision errors. The algorithm automatically generates detailed power output plans, charging and discharging sequences, etc., replacing some manual operations, saving labor costs, and reducing the frequency of manual adjustments.

[0053] The strategies of photovoltaics, energy storage, and load are matched with each other, operating like a precise process, which improves the smoothness of the overall system and enhances the coordination of system operation. The strategy space covers different adjustment ranges, and with the dynamic optimization of the algorithm, it can adapt to changes in conditions such as light and load, maintain system stability, and enhance the flexibility in dealing with variables. The structured optimal strategy set includes specific time nodes, power values, etc., which makes it easy to understand the operating logic of each agent, facilitates subsequent analysis and adjustment, and enhances the interpretability of decision results.

[0054] In one embodiment of the present invention, S3 includes:

[0055] S31. Integrate the real-time hardware model parameters (e.g., module switching response time, power loss coefficient) from S14 with the game decision results (e.g., energy interaction instructions of each agent) from S24 to construct a simulation input system containing 58 parameters; set the simulation time step and operating condition scenario labels, wherein the simulation time step is no greater than 100ms; the operating condition scenario labels include sunny / cloudy days and peak / valley loads.

[0056] S32. A coupled simulation platform for hardware structure and energy flow is built based on MATLAB / Simulink. During the simulation, a dynamic hardware topology model is called in real time to calculate the impact of hardware module switching on the energy transmission path (including power loss caused by impedance changes). The energy game decisions of each agent are substituted to simulate the dynamic flow process of energy between photovoltaic-energy storage-load-grid. Random disturbance factors are introduced, including sudden changes in illumination and load impacts, to simulate the system response under extreme conditions.

[0057] S33. Collect key parameters during the simulation process, including real-time power (accuracy ±1%), effective voltage (±0.5%), current waveform distortion rate (THD), energy conversion efficiency, etc. of each port; divide the collected data into time series segments (divided according to the points of change in operating conditions), calculate the statistical characteristics of each time period, including the mean, variance, and peak value; construct a device energy transmission simulation database and associate it with the corresponding hardware topology state and game decision parameters.

[0058] The working principle and effects of the above technical solution are as follows: It integrates hardware model parameters and game decision results. The input system of 58 parameters makes the simulation scenario closer to the actual operation, avoids the omission of key factors, and improves the comprehensiveness of energy flow simulation; the high-precision acquisition of parameters such as power and voltage at each port, coupled with statistical analysis after time series segmentation, makes the simulation results more valuable and improves the accuracy of data acquisition; the introduction of random disturbance factors such as sudden changes in illumination and load impact can detect potential problems of the system under special conditions in advance, provide direction for subsequent optimization, and improve the ability to cope with extreme working conditions.

[0059] The simulation platform built using MATLAB / Simulink can simulate various working conditions in a laboratory environment, reducing the number of on-site debugging sessions and resource investment, and lowering the cost of field testing. During the simulation, issues such as power loss and energy transmission bottlenecks caused by hardware switching are exposed in advance, avoiding potential failures in actual operation and reducing the risk of system operation. The equipment energy transmission simulation database links hardware status, game decisions, and transmission parameters, making subsequent queries and analysis more convenient, reducing data processing time, and lowering the difficulty of data association.

[0060] The coupled simulation platform calls the dynamic hardware topology model in real time, which can intuitively reflect the impact of hardware changes on energy transmission, making the cooperation between the two more coordinated and enhancing the synergy between hardware and energy flow. The simulation covers multiple scenario labels such as sunny / cloudy days and peak / off-peak loads, allowing the system to find a suitable operating mode in different environments, improving the system's adaptability. The strategy based on detailed simulation data is more reliable than relying solely on experience, improving the stability and efficiency of system operation.

[0061] In one embodiment of the present invention, S33 includes:

[0062] By using the real-time monitoring module of the coupled simulation platform, key parameters in the simulation process are collected. These key parameters include the real-time power (accuracy ±1%), effective voltage value (±0.5%), current waveform distortion rate (THD), energy conversion efficiency, and status parameters of hardware modules such as temperature and ripple, forming the original simulation parameter set.

[0063] Based on the operating condition change characteristics in the original simulation parameter set, including power abrupt changes and temperature surges, the collected data is segmented into time series segments according to the operating condition change points, dividing different stable operating condition intervals and transition intervals to obtain segmented time series datasets.

[0064] For each interval in the segmented time series dataset, calculate the statistical characteristics of each time period and extract a set of feature indicators that can reflect the characteristics of the data in that interval;

[0065] The feature index set is associated and matched with the corresponding hardware topology status data and game decision parameters. An indexing mechanism is established to build a device energy transmission simulation database for data tracing and querying.

[0066] The working principle and effects of the above technical solution are as follows: It comprehensively collects key parameters such as power, voltage, and temperature, even down to details like hardware ripple, ensuring the raw data fully reflects the system's operating status and improving the integrity of the simulation data; it segments the data according to operating condition characteristics such as sudden power changes and rapid temperature rises, separating the stable and transitional intervals for more focused data analysis and improving data relevance; and it uses an indexing mechanism to link characteristic indicators with hardware status and game theory decisions, allowing for quick location of any desired data without sifting through massive amounts of information, thus improving data query efficiency.

[0067] After segmentation, irrelevant and interfering information is removed, allowing subsequent analysis to focus on valuable data and reducing interference from invalid data. The database integrates scattered parameters, states, and decisions, saving the trouble of manual comparison, reducing the possibility of errors, and reducing the complexity of data association. Based on the feature index set, the problems under different working conditions can be clearly identified, the optimization direction is clearer, and the blind attempts to optimize the system are avoided.

[0068] From raw parameters to final decisions, every step can be traced step by step, allowing for quick identification of the root cause of any problem and enhancing data traceability. It specifically captures key nodes such as power surges and temperature spikes, enabling the system to respond more promptly to abnormal situations and enhancing its sensitivity to changes in operating conditions. The characteristic indicators of each interval can uncover specific operating patterns, providing a solid basis for subsequent system upgrades.

[0069] In one embodiment of the present invention, step S4 includes:

[0070] S41. Obtain the operational requirements data of the bidirectional photovoltaic energy storage inverter through the system control center, specifically including: system efficiency requirements (weighted average efficiency ≥ 95%), power quality requirements (voltage deviation ≤ ±5%, frequency deviation ≤ ±0.2Hz, THD ≤ 5%), energy storage battery life protection requirements (single charge / discharge depth ≤ 80%, cycle attenuation coefficient ≤ 0.01 / cycle), and grid connection requirements (meeting the fault ride-through capability of IEEE 1547 standard); quantify the requirements data into assessable indicator thresholds;

[0071] S42. Based on the energy game decision results of S24 and the transmission simulation data of S33, construct an operation status evaluation model: use the analytic hierarchy process (AHP) to determine the weights of each demand indicator, specifically the weights of efficiency (30%), power quality (25%), battery life (25%), and grid connection requirements (20%); establish a deviation function between the actual value and the threshold of the indicator, and calculate the compliance rate of individual indicators and the comprehensive compliance index.

[0072] S43. Substitute the simulated data into the evaluation model and calculate the comprehensive compliance index under the current operating state: if the comprehensive compliance index is ≥0.9 (set threshold), it is determined that the operating requirements are met; if it is <0.9, locate the non-compliant indicators (such as insufficient efficiency, excessive THD, etc.) and analyze their correlation with hardware topology and energy game strategy.

[0073] S44. For operational requirements that do not meet the standards, the required energy adjustment is calculated based on the deviation function. For efficiency problems, the power allocation correction value required to improve efficiency is calculated. For power quality problems, the compensation current amplitude and phase required to suppress harmonics are calculated. For battery life problems, the power limit value for optimizing the depth of charge and discharge is calculated. All adjustment values ​​are integrated into operational demand energy assessment data.

[0074] The working principle and effects of the above technical solution are as follows: Efficiency, power quality, and other requirements are quantified into specific thresholds, and the weight of each indicator is clarified using the analytic hierarchy process (AHP), making the evaluation results more closely aligned with the core needs of actual operation and improving the accuracy of operational requirement assessment; The comprehensive compliance index quickly determines whether the system meets the requirements, and if it fails to meet the requirements, the specific problem can be directly identified, saving time spent on blind troubleshooting and improving the efficiency of problem localization; Specific adjustment amounts are calculated for different non-compliant items, such as correcting power distribution for insufficient efficiency and compensating for excessive harmonics, making optimization measures more precise and improving the targeting of energy adjustments;

[0075] By assessing mandatory requirements such as battery life protection and grid connection in advance, we can avoid equipment damage or safety issues caused by overcharging and discharging or non-compliant grid connection, thus reducing the risk of system operation. Through adjustments such as power allocation correction and charge / discharge depth limits, we can reduce ineffective energy consumption and unnecessary energy loss. By replacing experience-based judgments with data models and quantitative indicators, we can make the assessment results more objective and reduce human error.

[0076] Strict adherence to standards such as IEEE 1547 ensures grid connection operations comply with industry norms, avoiding penalties for non-compliance; precise control of battery charge / discharge depth and cycle count slows battery aging and extends lifespan; stable voltage, qualified frequency, and low harmonic content ensure more stable operation of electrical equipment and reduce malfunctions caused by power quality issues.

[0077] In one embodiment of the present invention, step S5 includes:

[0078] S51. Based on the equipment energy transmission simulation data of S33 and the operation demand energy assessment data of S44, an initial adjustment strategy is generated using a fuzzy control algorithm. The initial adjustment strategy includes clarifying the direction of hardware structure adjustment (such as increasing the number of levels, switching port connections) and the correction amount of energy interaction strategy (such as photovoltaic output fine-tuning value, battery charging and discharging power correction coefficient); and formulating multiple alternative strategy schemes (no less than 3 sets).

[0079] S52. Combining the dynamic hardware topology reconfigurability feature of S13, perform collaborative optimization on the candidate strategies. Use a genetic algorithm to jointly optimize the hardware structure parameters and energy interaction strategies. The hardware structure parameters include the number of power levels and the port connection method. The energy interaction strategy includes the power allocation ratio of each agent. The objective function is set to maximize the comprehensive achievement index. Constraints are introduced to ensure the feasibility of hardware switching (e.g., switching time ≤ 10ms) and the safety of energy adjustment (e.g., not exceeding the rated power of the device). Output the optimal collaborative optimization strategy.

[0080] S53. Simulate and verify the optimal collaborative optimization strategy on the real-time hardware model, simulate the hardware state changes and energy flow response after the strategy is executed, and verify whether it meets all operating requirements (comprehensive compliance index ≥ 0.95); perform early warning analysis on potential risks (such as hardware overcurrent, voltage surge), and if there are risks, return to S51 to regenerate the strategy.

[0081] S54. The verified collaborative optimization strategy is converted into standardized control commands (compliant with the Modbus-RTU protocol) and transmitted to the control unit of the bidirectional photovoltaic energy storage inverter via industrial Ethernet. After parsing the commands, the control unit drives the hardware module to perform a reconfiguration operation, which includes adjusting the turn-on / turn-off timing of the power switching transistors and issuing energy interaction correction commands to each intelligent agent. The strategy execution effect is monitored in real time to form a closed-loop control and realize the dynamic adaptive adjustment of the inverter's operating state.

[0082] The working principle and effects of the above technical solution are as follows: an initial strategy is generated by combining simulation data and evaluation results, and then the hardware and energy strategies are jointly optimized through a genetic algorithm, so that the adjustment scheme can not only fit the actual working conditions but also maximize the comprehensive benefits, thus improving the scientific nature of the operation status adjustment; the simulation verification process is strictly controlled to ensure that the optimized strategy can meet all operational requirements and can also provide early warning of potential risks, reducing the possibility of problems during actual execution; the hardware switching time is controlled within 10ms, and with closed-loop monitoring and real-time adjustment, the inverter can quickly adapt to changes in operating conditions, maintain stable operation, and improve the timeliness of system response;

[0083] By using fuzzy control and genetic algorithms to find the optimal hardware configuration, the wear and tear on components caused by blind adjustments is avoided, the lifespan of the hardware is extended, and the cost of hardware reconfiguration is reduced. The precise power allocation ratio and charge / discharge correction coefficient reduce unnecessary energy conversion losses, improve overall energy efficiency, and reduce energy regulation losses. The entire process from strategy generation to execution is automated, and the closed-loop control automatically corrects deviations, reducing the workload and error rate of manual operation and the frequency of human intervention.

[0084] The dynamic hardware topology combined with the coordinated optimization of energy strategies can flexibly respond to various changes in lighting, load, etc., and can find a suitable operating mode no matter how the operating conditions change, thus enhancing the system's adaptability. Strictly adhering to constraints such as power limits and overcurrent warnings avoids equipment damage or safety accidents caused by hardware overload or abnormal voltage, thus enhancing the safety of equipment operation. Hardware reconfiguration and energy interaction adjustment are carried out simultaneously, and the cooperation between various intelligent agents and hardware modules is more tacit, allowing the entire system to operate as an organic whole with high efficiency.

[0085] One embodiment of the present invention provides an operation status control system for a bidirectional photovoltaic energy storage inverter, comprising:

[0086] One or more processors;

[0087] Memory, used to store one or more programs.

[0088] Wherein, when the one or more programs are executed by the one or more processors, the one or more processors implement the method described in any one of the above.

[0089] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for controlling the operating state of a bidirectional photovoltaic energy storage inverter, characterized in that, The method includes: S1: Obtain the initial hardware topology data and real-time operating condition data of the bidirectional photovoltaic energy storage inverter; analyze the initial hardware topology data to determine its reconfigurable hardware modules; configure the reconfigurable hardware modules according to the preset reconfigurable rules based on the real-time operating condition data to obtain the dynamic hardware topology structure; construct the real-time hardware model of the bidirectional photovoltaic energy storage inverter based on the dynamic hardware topology structure. S2: Divide the bidirectional photovoltaic energy storage inverter system into multiple intelligent agents, extract energy features for each intelligent agent, construct an energy game model for each intelligent agent based on the extracted energy features, solve the energy game model for each intelligent agent using a multi-agent dynamic energy game algorithm, and obtain the energy game decision results for each intelligent agent. S3: Based on the real-time hardware model obtained in S1 and the energy game decision results of each intelligent agent obtained in S2, perform energy flow simulation of the bidirectional photovoltaic energy storage inverter; obtain equipment energy transmission simulation data of the bidirectional photovoltaic energy storage inverter under different operating conditions; S4: Obtain the operation demand data of the bidirectional photovoltaic energy storage inverter. Based on the energy game decision results of each intelligent agent in S2 and the equipment energy transmission simulation data in S3, evaluate the operation demand data and analyze whether the current operation status meets the various requirements. If not, calculate the energy adjustment amount required to meet the operation requirements and obtain the operation demand energy assessment data. S5: Based on the equipment energy transmission simulation data in S3 and the operation demand energy assessment data in S4, determine the operation status adjustment strategy of the bidirectional photovoltaic energy storage inverter; combined with the reconfigurable characteristics of the dynamic hardware topology in S1, further optimize the hardware structure according to the operation status adjustment strategy, and fine-tune the energy interaction strategy of each agent according to the result of the multi-agent dynamic energy game, and perform collaborative optimization of hardware structure evolution and energy flow intelligent game decision-making; transmit the collaboratively optimized operation status control strategy to the control unit of the bidirectional photovoltaic energy storage inverter to execute the operation status control task and realize the dynamic adaptive adjustment of the inverter operation status.

2. The method for controlling the operating state of a bidirectional photovoltaic energy storage inverter according to claim 1, characterized in that, S1 includes: S11. Obtain the initial hardware topology data of the bidirectional photovoltaic energy storage inverter, synchronously collect real-time operating condition data, preprocess the collected data, and construct a standardized dataset. S12. Based on the initial hardware topology data, the modular decomposition algorithm is used, combined with graph theory and circuit topology analysis, to identify reconfigurable hardware modules, establish a reconfigurability evaluation matrix for each module, and label the reconfiguration modes it supports. S13. Based on the preprocessed real-time operating data, call the preset reconfigurable rule library to perform dynamic configuration on the reconfigurable hardware modules, and perform real-time switching between modules through the hardware configuration instruction set to generate a dynamic hardware topology. S14. Based on the dynamic hardware topology and combined with the real-time parameters of each module, a real-time hardware model of a bidirectional photovoltaic energy storage inverter based on state-space equations is constructed; the least squares method is used to calibrate the model parameters online, forming a high-precision model that can be directly used for simulation analysis.

3. The method for controlling the operating state of a bidirectional photovoltaic energy storage inverter according to claim 2, characterized in that, The initial hardware topology data includes static attribute data, which includes the power conversion unit circuit structure, component parameter specifications, and module connection relationships. The real-time operating data includes dynamic parameters, such as light intensity, ambient temperature, load demand, and grid voltage / frequency.

4. The method for controlling the operating state of a bidirectional photovoltaic energy storage inverter according to claim 1, characterized in that, The S2 includes: S21. Using a distributed intelligent agent modeling method, the bidirectional photovoltaic energy storage inverter system is divided into three types of core intelligent agents, and the communication boundaries and energy interaction range of each intelligent agent are determined. S22. For the photovoltaic array intelligent agent, the first characteristic parameter is collected and extracted by the synchronous phasor measurement unit. For the energy storage battery intelligent agent, the second characteristic parameter is obtained by processing with the Kalman filter algorithm. For the load intelligent agent, the third characteristic parameter is extracted by load characteristic analysis. A time series database of characteristic parameters is constructed. S23. Based on non-cooperative game theory, construct an energy game model for each agent; introduce constraints and establish a multi-objective optimization game equilibrium equation. S24. An improved particle swarm optimization algorithm is used to solve the energy game model: the policy space of each agent is initialized, and the Nash equilibrium point is found through iterative calculation; the optimal policy set of each agent is output to form the energy game decision result.

5. The method for controlling the operating state of a bidirectional photovoltaic energy storage inverter according to claim 4, characterized in that, The three types of core intelligent agents include photovoltaic array intelligent agents, energy storage battery intelligent agents, and load intelligent agents; the photovoltaic array intelligent agents are used for light energy capture and conversion; the energy storage battery intelligent agents are used for energy storage and release; and the load intelligent agents are used for energy consumption and demand response.

6. The method for controlling the operating state of a bidirectional photovoltaic energy storage inverter according to claim 4, characterized in that, The first characteristic parameters include, but are not limited to, power fluctuation rate, voltage distortion rate, and maximum tracking point offset; The second characteristic parameter includes, but is not limited to, state of charge, state of health, charge / discharge efficiency curve, and cycle life decay rate; The second characteristic parameter includes, but is not limited to, state of charge, state of health, charge / discharge efficiency curve, and cycle life decay rate.

7. The method for controlling the operating state of a bidirectional photovoltaic energy storage inverter according to claim 1, characterized in that, The S3 includes: S31. Integrate the real-time hardware model parameters of S14 with the game decision results of S24 to construct a simulation input system containing 58 parameters; set the simulation time step and working condition scenario labels; S32. Based on MATLAB / Simulink, a coupled simulation platform for hardware structure and energy flow is built. During the simulation, the dynamic hardware topology model is called in real time to calculate the impact of hardware module switching on energy transmission path. The energy game decision of each intelligent agent is substituted to simulate the dynamic flow process of energy between photovoltaic-energy storage-load-grid. Random disturbance factors are introduced to simulate the system response under extreme conditions. S33. Collect key parameters during the simulation process, segment the collected data into time series segments, calculate the statistical characteristics of each time period, construct a device energy transmission simulation database, and associate the corresponding hardware topology status with game decision parameters.

8. The method for controlling the operating state of a bidirectional photovoltaic energy storage inverter according to claim 1, characterized in that, The S4 includes: S41. Obtain the operation demand data of the bidirectional photovoltaic energy storage inverter through the system control center, and quantify the demand data into assessable indicator thresholds. S42. Based on the energy game decision results of S24 and the transmission simulation data of S33, construct an operation status evaluation model: use the analytic hierarchy process to determine the weight of each demand indicator, establish the deviation function between the actual value of the indicator and the threshold, and calculate the compliance rate of individual indicators and the comprehensive compliance index. S43. Substitute the simulated data into the evaluation model and calculate the comprehensive compliance index under the current operating state: if the comprehensive compliance index is ≥0.9, it is determined that the operating requirements are met; if it is <0.9, locate the non-compliant indicators and analyze their correlation with hardware topology and energy game strategy. S44. For operational requirements that do not meet the standards, the required energy adjustment amount is deduced based on the deviation function, and the various adjustment amounts are integrated into operational requirement energy assessment data.

9. The method for controlling the operating state of a bidirectional photovoltaic energy storage inverter according to claim 1, characterized in that, The S5 includes: S51. Based on the equipment energy transmission simulation data of S33 and the operation demand energy assessment data of S44, an initial adjustment strategy is generated using a fuzzy control algorithm. S52. Combining the dynamic hardware topology reconfigurability of S13, perform collaborative optimization on the alternative strategies. Use a genetic algorithm to jointly optimize the hardware structure parameters and energy interaction strategies, and introduce constraints to ensure the feasibility of hardware switching and the safety of energy adjustment; output the optimal collaborative optimization strategy. S53. Simulate and verify the optimal collaborative optimization strategy on the real-time hardware model, simulate the hardware state changes and energy flow response after the strategy is executed, and verify whether all operational requirements are met; perform early warning analysis on potential risks, and if risks exist, return to S51 to regenerate the strategy. S54. The verified collaborative optimization strategy is converted into standardized control commands and transmitted to the control unit of the bidirectional photovoltaic energy storage inverter via industrial Ethernet. After parsing the commands, the control unit drives the hardware module to perform the reconfiguration operation and sends energy interaction correction commands to each intelligent agent. The strategy execution effect is monitored in real time to form a closed-loop control.

10. An operation status control system for a bidirectional photovoltaic energy storage inverter, comprising: One or more processors; Memory, used to store one or more programs. Wherein, when the one or more programs are executed by the one or more processors, the one or more processors implement the method of any one of claims 1 to 9.

Images