A global coordinated control method and system for photovoltaic energy storage system

Abstract

The present application relates to the field of information technology, and discloses a global coordinated control method and system for a photovoltaic energy storage system, wherein the method is applied to a global network comprising a number of intelligent agents, the intelligent agents including a photovoltaic system, an energy storage device and a converter; the method comprises: obtaining a current output sequence of the converter according to load demand forecast data; obtaining a dynamic overcurrent protection strategy according to the current output sequence and the actual current output of the converter; determining an optimal charge and discharge strategy for the energy storage device based on charge state data of the energy storage device, a photovoltaic power generation forecast value, a load demand forecast value and a global consistency constraint of the global network; performing a global optimization solution based on the current output sequence of the converter, the dynamic overcurrent protection strategy and the optimal charge and discharge strategy of the energy storage device through a distributed alternating direction multiplier method, and obtaining a global operation strategy of the global network at the current moment; achieving a technical effect of improving the output performance of the converter while ensuring safety.

Description

Technical Field

[0001] The present application relates to the field of information technology, and in particular to a global coordinated control method and system for a photovoltaic energy storage system. Background Art

[0002] In the coordinated control of photovoltaic (PV) and energy storage systems, the converter plays a crucial role. It not only converts and transmits energy but also strikes a delicate balance between protecting its own safety and achieving optimal power distribution. This balancing act presents a challenge: the converter's built-in overcurrent protection mechanism is designed to prevent damage from overload. However, coordinated control methods sometimes require the converter to temporarily exceed this protection limit, thereby outputting higher currents, for the benefit of the entire system. Therefore, an ideal control strategy is needed that balances converter safety and output performance while enabling coordinated control of the converter with the PV system and energy storage device. Summary of the Invention

[0003] The present application provides a global coordinated control method and system for a photovoltaic energy storage system, which solves the technical problem that the converter cannot take into account both safety and output performance, and achieves the technical effect of improving the output performance of the converter while ensuring safety.

[0004] In order to achieve the above objectives, the main technical solutions adopted in this application include:

[0005] In a first aspect, embodiments of the present application provide a global coordinated control method for a photovoltaic energy storage system, applied to a global network comprising a plurality of intelligent agents, wherein the intelligent agents include a photovoltaic system, an energy storage device, and a converter; the method comprising:

[0006] Obtaining a current output sequence of the converter according to load demand forecast data; wherein the load demand forecast data is predicted based on photovoltaic power generation forecast data of the photovoltaic system and historical operation data of the converter;

[0007] According to the current output sequence, an overcurrent protection strategy of the converter is obtained by grading, and the overcurrent protection strategy is adjusted according to the actual current output of the converter to obtain a dynamic overcurrent protection strategy; wherein the dynamic overcurrent protection strategy includes an overcurrent protection threshold and a trip time;

[0008] Determining an optimal charging and discharging strategy for the energy storage device based on the state of charge data of the energy storage device, the photovoltaic power generation forecast value and the load demand forecast value of the photovoltaic system, and the global consistency constraints of the global network; wherein the global consistency constraints include power balance constraints and equipment operation restrictions;

[0009] Through the distributed alternating direction multiplier method, a global optimization solution is performed based on the current output sequence of the converter, the dynamic overcurrent protection strategy and the optimal charging and discharging strategy of the energy storage device, and the global operation strategy of the global network at the current moment is obtained to achieve distributed collaborative optimization of the converter overcurrent protection, optimal power distribution and the charging and discharging strategy of the energy storage device.

[0010] The global coordinated control method of the photovoltaic energy storage system provided in the embodiment of the present application includes obtaining a dynamic overcurrent protection strategy of the converter, and by setting the overcurrent protection threshold and tripping time, the converter can output a current higher than the overcurrent protection threshold within the tripping time, thereby improving the output of the photovoltaic energy storage system and meeting the actual needs of the load; at the same time, if the output current is higher than the overcurrent protection threshold, it will automatically trip after the tripping time to cut off the current, ensuring the safety of each device in the photovoltaic energy storage system and reducing the probability of equipment failure. The dynamic overcurrent protection strategy is obtained through the current output sequence and actual current output of the converter, and the load demand forecast data used to obtain the current output sequence is obtained by predicting the photovoltaic power generation forecast data of the photovoltaic system and the historical operation data of the converter. Therefore, during the operation of the converter, the dynamic overcurrent protection strategy can dynamically change with changes in natural environmental factors and load demand, so that the operation of the photovoltaic system can adapt to changes in the external environment, thereby improving the energy efficiency and operation safety of the photovoltaic system.

[0011] In addition, the global coordinated control method for a photovoltaic energy storage system provided in an embodiment of the present application also includes performing a global optimization solution on a global network comprising a number of intelligent agents, thereby obtaining the global operating strategy of the global network at the current moment. In this method, the dynamic overcurrent protection strategy of the converter changes dynamically with changes in the external environment. Therefore, it is necessary to determine the optimal charging and discharging strategy of the connected energy storage device based on changes in the operating state of the photovoltaic system and the converter, thereby achieving distributed collaborative optimization of converter overcurrent protection, optimal power distribution, and energy storage device charging and discharging strategies, thereby improving the output performance and reliability of the photovoltaic energy storage system.

[0012] In a second aspect, embodiments of the present application provide a global coordinated control system for a photovoltaic energy storage system, which is applied to a global network comprising a plurality of intelligent agents, wherein the intelligent agents include a photovoltaic system, an energy storage device, and a converter; the system includes:

[0013] a current sequence determination module, configured to obtain a current output sequence of the converter according to load demand forecast data; wherein the load demand forecast data is obtained by forecasting photovoltaic power generation forecast data of the photovoltaic system and historical operation data of the converter;

[0014] a protection threshold adjustment module, configured to obtain an overcurrent protection strategy for the converter by grading according to the current output sequence, and adjust the overcurrent protection strategy according to real-time operating data of the converter to obtain a dynamic overcurrent protection strategy;

[0015] an energy storage strategy determination module, configured to determine an optimal charge and discharge strategy for the energy storage device based on the state of charge data of the energy storage device, the photovoltaic power generation forecast value and the load demand forecast value of the photovoltaic system, and the global consistency constraints of the global network; wherein the global consistency constraints include power balance constraints and equipment operation restrictions;

[0016] A global strategy determination module is used to perform a global optimization solution based on the current output sequence of the converter, the dynamic overcurrent protection strategy, and the optimal charge and discharge strategy of the energy storage device using a distributed alternating direction multiplier method to obtain the global operation strategy of the global network at the current moment, thereby achieving distributed collaborative optimization of the converter overcurrent protection, optimal power distribution, and energy storage device charge and discharge strategy.

[0017] In a third aspect, an embodiment of the present application provides a computer device comprising: a memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the global coordinated control method of the photovoltaic energy storage system described in any one of the above embodiments by executing the computer instructions.

[0018] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium having computer instructions stored thereon, wherein the computer instructions are used to enable a computer to execute the global coordinated control method for a photovoltaic energy storage system according to any one of the above embodiments.

[0019] In a fifth aspect, an embodiment of the present application provides a computer program product, comprising computer instructions, wherein the computer instructions are used to enable a computer to execute the global coordinated control method for a photovoltaic energy storage system according to any one of the above embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the specific implementation methods of the present application or the technical solutions in the prior art, the following is a brief introduction to the drawings required for use in the specific implementation methods or the description of the prior art. Obviously, the drawings described below are some implementation methods of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0021] Figure 1 A step diagram of a global coordinated control method for a photovoltaic energy storage system provided in an embodiment of the present application;

[0022] Figure 2a A diagram showing the steps for setting an overcurrent protection strategy in an embodiment of the present application;

[0023] Figure 2b A diagram showing the steps for adjusting the overcurrent protection strategy in an embodiment of the present application;

[0024] Figure 2c A diagram showing the steps of adaptively correcting the overcurrent protection strategy according to an embodiment of the present application;

[0025] Figure 3 A diagram showing the steps for obtaining a global operation strategy in an embodiment of the present application;

[0026] Figure 4 A module diagram of a global coordinated control system for a photovoltaic energy storage system provided in an embodiment of the present application;

[0027] Figure 5 A schematic diagram of the structure of a computer device provided in an embodiment of the present application.

[0028] The figure numbers of the drawings in the specification are as follows: 100. Current sequence determination module, 200. Protection threshold adjustment module, 300. Energy storage strategy determination module, 400. Global strategy determination module. DETAILED DESCRIPTION

[0029] To make the purpose, technical solutions, and advantages of the embodiments of the present application more clear, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without making creative efforts shall fall within the scope of protection of this application.

[0030] In the coordinated control of photovoltaic energy storage systems, converters play a crucial role. They are responsible not only for energy conversion and transmission but also for striking a delicate balance between protecting their own safety and achieving optimal power distribution. This balancing act presents a challenge: the converter's built-in overcurrent protection mechanism is designed to prevent damage from overload. However, sometimes, for the benefit of the entire system, the coordinated control algorithm requires the converter to temporarily exceed this protection limit and output higher current. This creates a technical dilemma: an overly conservative overcurrent protection strategy ensures device safety but may limit system efficiency; an overly aggressive overcurrent protection strategy increases current output but increases the probability of device failure.

[0031] Furthermore, the balance between converter safety and output performance changes dynamically with fluctuations in external environmental conditions. These conditions include both natural and equipment conditions. Natural conditions refer to the natural environment surrounding the PV energy storage system, including light intensity, temperature, and precipitation. Equipment conditions refer to the operating conditions of the equipment within and connected to the PV energy storage system, including load demand. This means that an ideal control strategy should be able to adjust in real time to adapt to the changing external environment, while ensuring that the converter maintains a balance between safety and output performance to meet the practical needs of the PV energy storage system.

[0032] Furthermore, changes in control strategies can cause changes in the converter's operating state. Consequently, the operating states of the PV system and energy storage device connected to the converter must change synchronously with these changes. Therefore, global optimization of the PV energy storage system, which includes the PV system, energy storage device, and converter, is necessary. This allows for distributed, coordinated optimization of converter overcurrent protection, optimal power distribution, and energy storage device charging and discharging strategies, ultimately achieving optimal coordinated control of the entire system.

[0033] To address the aforementioned issues, this application provides a global coordinated control method and system for a photovoltaic energy storage system, applicable to a photovoltaic energy storage system comprising a photovoltaic system, an energy storage device, an inverter, and other equipment. Furthermore, this application is applicable to a global network comprising multiple intelligent agents, formed by treating each device in the photovoltaic energy storage system as an intelligent agent, wherein the intelligent agents are connected to each other through physical and communication connections.

[0034] The global coordinated control method for a photovoltaic energy storage system provided in this application includes: determining the dynamic overcurrent protection strategy of the converter based on the current output sequence and actual current output of the converter; wherein the current output sequence of the converter is obtained based on load demand forecast data, and the load demand forecast data is predicted based on the photovoltaic power generation forecast data of the photovoltaic system and the historical operation data of the converter. For the global network, a global optimization solution is performed based on the current output sequence of the converter and the dynamic overcurrent protection strategy and the optimal charge and discharge strategy of the energy storage device to obtain the global operation strategy of the global network at the current moment, thereby realizing distributed collaborative optimization of the converter overcurrent protection, optimal power distribution, and the charge and discharge strategy of the energy storage device; wherein the optimal charge and discharge strategy of the energy storage device is determined based on the current output sequence of the converter.

[0035] The present application provides a global coordinated control method and system for a photovoltaic energy storage system, which solves the technical problem in related technologies that the converter cannot take into account both safety and output performance, and achieves the technical effect of improving the output performance of the converter while ensuring safety.

[0036] According to an embodiment of the present application, an embodiment of a global coordinated control method for a photovoltaic energy storage system is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0037] In this embodiment, a global coordinated control method for a photovoltaic energy storage system is provided, which can be applied to the above-mentioned global network comprising a plurality of intelligent agents, wherein the intelligent agents include a photovoltaic system, an energy storage device, and a converter. Figure 1 The steps of the global coordinated control method of the photovoltaic energy storage system provided in the embodiment of the present application are shown in the figure, including the following steps:

[0038] S100. Obtaining a current output sequence of the converter according to load demand prediction data; wherein the load demand prediction data is obtained by prediction based on photovoltaic power generation prediction data of the photovoltaic system and historical operation data of the converter.

[0039] S200. According to the current output sequence, the overcurrent protection strategy of the converter is obtained by grading, and the overcurrent protection strategy is adjusted according to the actual current output of the converter to obtain a dynamic overcurrent protection strategy; wherein the dynamic overcurrent protection strategy includes an overcurrent protection threshold and a tripping time.

[0040] S300. Determine the optimal charging and discharging strategy for the energy storage device based on the state of charge data of the energy storage device, the photovoltaic power generation forecast value and load demand forecast value of the photovoltaic system, and the global consistency constraints of the global network; wherein the global consistency constraints include power balance constraints and equipment operation restrictions.

[0041] S400. Through the distributed alternating direction multiplier method, a global optimization solution is performed based on the current output sequence of the converter, the dynamic overcurrent protection strategy, and the optimal charging and discharging strategy of the energy storage device, and the global operation strategy of the global network at the current moment is obtained to achieve distributed collaborative optimization of the converter overcurrent protection, optimal power distribution, and charging and discharging strategy of the energy storage device.

[0042] It should be noted that the converter's current output sequence is a time series model, representing the relationship between the converter's output current and the time series over a given period of time. It should be noted that the current output sequence acquired in step S100 represents predicted data for the converter's required output current at the current moment. Therefore, during converter operation, current control instructions for the converter can be generated based on the current output sequence, thereby driving the converter's power devices and changing the converter's output current.

[0043] Specifically, load demand forecast data represents the power required by an external load connected to a photovoltaic energy storage system within a specified timeframe. This external load relies, or partially relies, on the output of the photovoltaic energy storage system to function properly. In some embodiments, methods for obtaining load demand forecast data include, but are not limited to, time series analysis, machine learning, physical model prediction, and hybrid methods incorporating these methods. It will be appreciated that a photovoltaic energy storage system includes a photovoltaic system, an energy storage device, and an inverter. In some embodiments, the inverter is connected between the photovoltaic system and the energy storage device.

[0044] Furthermore, the load demand forecast data is derived based on the photovoltaic system's photovoltaic power generation forecast data and the converter's historical operating data. It is understood that the photovoltaic system's photovoltaic power generation forecast data is related to light intensity, while the converter's historical operating data is related to the external load's historical data. Changes in load demand can be determined using this historical operating data. For example, when the power required by the external load increases during normal operation, the load demand forecast data increases, and the current values ​​in the converter's current output sequence also increase accordingly. When the power required by the external load decreases during normal operation, the load demand forecast data decreases, and the current values ​​in the converter's current output sequence also decrease accordingly. By predicting load demand based on the photovoltaic system's photovoltaic power generation forecast data and the converter's historical operating data, the converter's operating state, as determined by the current output sequence, can dynamically change based on changes in natural and equipment environmental conditions, thereby enhancing the converter's adaptability to external environmental conditions.

[0045] Specifically, the overcurrent protection strategy includes an overcurrent protection threshold and a trip time. The overcurrent protection threshold is the trigger point for overcurrent protection and corresponds to different current levels in the current output sequence. The upper limit of the overcurrent protection threshold is the maximum current that the converter hardware can carry. This limit is the maximum current that the converter can output above the overcurrent protection threshold for a short period of time without causing permanent damage to the converter. This limit is determined by the converter's rated current and short-term overload capacity. The trip time is the duration during which the converter outputs a current above the overcurrent protection threshold. After the trip time, the converter's current output is cut off, thereby protecting the converter's internal components. In some embodiments, the trip time can be determined by the converter's short-term overload capacity. Based on the converter's output current and short-term overload capacity, a trip time is determined that will not cause permanent damage to the converter. It will be appreciated that there is a one-to-one correspondence between the overcurrent protection threshold and the trip time. A higher overcurrent protection threshold results in a shorter trip time, thereby reducing damage to the converter.

[0046] Furthermore, the overcurrent protection strategy is adjusted according to the actual current output of the converter. It is understandable that when the actual current output of the converter changes, if the original overcurrent protection strategy is still applied, it may cause the converter to frequently perform overcurrent protection or fail to effectively perform overcurrent protection, resulting in an impact on the working efficiency of the converter. At this time, it is necessary to adjust the overcurrent protection strategy according to the real-time changes in the actual current output of the converter to obtain a dynamic overcurrent protection strategy to achieve dynamic changes in the overcurrent protection strategy. For example, if the actual current output of the converter increases, the overcurrent protection threshold is increased and the tripping time is shortened, thereby reducing the probability of the converter frequently performing overcurrent protection. It is understandable that the increased overcurrent protection threshold should be lower than the converter's limit value; if the actual current output of the converter decreases, the overcurrent protection threshold in the overcurrent protection strategy is reduced and the tripping time is prolonged, thereby effectively performing overcurrent protection on the converter.

[0047] Specifically, after determining the converter's current output sequence and dynamic overcurrent protection strategy, the converter's operating state is adjusted accordingly. At this time, the converter's operating state will change, thereby affecting the photovoltaic system and energy storage device connected to the converter. Therefore, it is necessary to determine the optimal charging and discharging strategy for the energy storage device based on the converter's operating state.

[0048] The optimal charging and discharging strategy for an energy storage device is related to the state-of-charge data of the energy storage device, the photovoltaic power generation of the photovoltaic system, the current output sequence of the converter, the load demand data, and the global consistency constraints of the global network. In some embodiments, the objectives of the optimal charging and discharging strategy for the energy storage device include, but are not limited to, maximizing the economic benefits of the photovoltaic energy storage system, extending the service life of the equipment, meeting load demands, and multiple objectives encompassing the aforementioned objectives. In different embodiments, the objectives can be determined based on actual application needs. It should be noted that the global consistency constraints of the global network are constraints between different intelligent agents in the global network, including, but not limited to, power balance constraints and equipment operation restrictions.

[0049] Specifically, the Distributed Alternating Direction Method of Multipliers (DADMM) is a method for solving distributed optimization problems. It combines the ideas of the Alternating Direction Method of Multipliers (ADMM) and distributed computing, and can effectively handle optimization problems of large-scale and distributed data. The main idea of ​​the Distributed Alternating Direction Multiplier Method is to construct a global network consisting of several nodes, split the global problem into multiple sub-problems in the process of optimizing the global problem, and distribute the sub-problems to each node, and each node solves its own sub-problem. In some embodiments, the sub-problems are solved by the primal-dual interior point method. An alternating update mechanism is introduced in the optimization calculation of the global problem, and the global parameters related to the global problem are updated according to the results of the sub-problems. Based on the information transmission and coordination between different nodes, the optimal solution of the global problem is obtained through iterative calculation.

[0050] Exemplarily, the global operating strategy of the global network at the current moment includes the operating strategies of all agents in the global network. The operating strategies of the agents include the current output sequence of the converter, the optimal charge and discharge strategy of the energy storage device, and the photovoltaic power generation data of the photovoltaic system. It is understood that the photovoltaic power generation data of the photovoltaic system is related to external environmental conditions and is not controlled by a specific method; the current output sequence of the converter is obtained based on the photovoltaic power generation data of the photovoltaic system and the historical operating data of the converter; and the optimal charge and discharge strategy of the energy storage device is obtained based on the photovoltaic power generation data of the photovoltaic system, the current output sequence of the converter, and the state of charge data of the energy storage device. Therefore, it can be seen that during the global optimization process, the operating strategies of each agent in the global network are interconnected. By controlling the operating strategy of the converter or energy storage device, a distributed collaborative optimization of converter overcurrent protection, optimal power distribution, and energy storage device charge and discharge strategy in the global network is achieved, thereby improving the output performance and reliability of the photovoltaic energy storage system.

[0051] The global coordinated control method of the photovoltaic energy storage system provided in this embodiment is applied to a global network containing several intelligent agents, so as to obtain the global operation strategy of the global network. Among them, the global operation strategy is optimized and solved by the distributed alternating direction multiplier method, and the global problem is decomposed into several sub-problems and distributed to each intelligent agent, so that each intelligent agent only needs to solve its own sub-problem on the basis of satisfying the global consistency condition. Among them, the sub-problems include calculating the current output sequence of the converter and the optimal charging and discharging strategy of the energy storage device. On this basis, this embodiment designs the overcurrent protection strategy of the converter, so that the converter can improve the current output capacity to a certain extent according to the load demand while ensuring the safety of the equipment to meet the load demand; at the same time, the overcurrent protection strategy is adjusted according to the external environmental conditions, and the resulting dynamic overcurrent protection strategy can be dynamically changed according to the natural environmental conditions and the equipment environmental conditions, so that the operation of the photovoltaic system can adapt to changes in the external environment. The beneficial effect of the global coordinated control method for the photovoltaic energy storage system provided in this embodiment is that by setting a dynamic overcurrent protection strategy for the converter, the energy efficiency and operational safety of the photovoltaic system are effectively improved, the stable operation of the system is ensured, and the utilization efficiency of the energy storage device is optimized, achieving dual optimization of economic and environmental benefits.

[0052] In addition, the global coordinated control method for a photovoltaic energy storage system provided in an embodiment of the present application also includes performing a global optimization solution on a global network comprising a number of intelligent agents, thereby obtaining the global operating strategy of the global network at the current moment. In this method, the dynamic overcurrent protection strategy of the converter changes dynamically with changes in the external environment. Therefore, it is necessary to determine the optimal charging and discharging strategy of the connected energy storage device based on changes in the operating state of the photovoltaic system and the converter, thereby achieving distributed collaborative optimization of converter overcurrent protection, optimal power distribution, and energy storage device charging and discharging strategies, thereby improving the output performance and reliability of the photovoltaic energy storage system.

[0053] As an embodiment of the present application, obtaining a current output sequence of a converter according to load demand prediction data includes:

[0054] S110. Establish a current output optimization model based on the load demand forecast data and the initial overcurrent protection strategy and optimal power allocation target of the converter; wherein, if there is a previous moment before the current moment, the initial overcurrent protection strategy is the dynamic overcurrent protection strategy of the previous moment.

[0055] S120. Determine an objective function based on the power loss and utilization rate of the converter, and optimize the current output optimization model based on the objective function to obtain a current output sequence.

[0056] Specifically, the current output optimization model is a weighted multi-objective optimization model, which includes an objective function and constraints. The objective function is obtained by setting weights based on the power loss and utilization of the converter, and its form is as follows:

[0057]

[0058] in, is the objective function; is the output current value of the converter at different time points, which can be obtained according to the current output sequence; is the power loss of the converter, which should be minimized in the optimization calculation; is the converter utilization, which should be maximized in the optimization calculation; and are the weight coefficients of the power loss and utilization rate of the converter respectively.

[0059] Furthermore, the power loss of the converter includes conduction loss, switching loss, dead zone loss, driving loss and passive component loss, wherein the conduction loss is the power loss caused by the current flowing through the switch device when the switch device is turned on due to the on-resistance of the switch device. The switching loss is the power loss caused by the simultaneous existence of voltage and current for a short period of time when the switch device switches from the on state to the off state or vice versa during the switching process. The dead zone loss is the power loss generated during the moment when the switch device switches from on to off, in order to avoid two complementary switch devices from being turned on at the same time, there will be an extremely short dead zone time. The driving loss is the power loss generated by the driving circuit of the switching device. The passive component loss is the power loss generated by passive components such as inductors and capacitors when current passes through. Therefore, the power loss of the converter satisfies the following formula:

[0060]

[0061] in, is the conduction loss; is the switching loss; is the dead zone loss; is the driving loss; is the passive component loss.

[0062] Furthermore, the utilization rate of the converter refers to the ratio of the actual power output of the converter to the maximum possible power output within a specific time, which can reflect the efficiency of the converter. The utilization rate of the converter satisfies the following formula:

[0063]

[0064] in, is the actual power output of the converter; is the maximum possible power output of the converter.

[0065] In some embodiments, the weight coefficients of the power loss and utilization of the converter can be determined according to the needs of the actual scenario. When the actual scenario places more emphasis on minimizing the power loss of the converter, the weight coefficient can be appropriately increased. And reduce the weight coefficient ; When the actual scenario places more emphasis on maximizing the utilization of the converter, the weight coefficient can be appropriately increased And reduce the weight coefficient ; It should be noted that the weight coefficient must satisfy For example, the weight coefficient in this embodiment is The weight coefficient can be 0.6. 0.4 can be taken.

[0066] Specifically, the constraints of the current output optimization model include converter capacity limitation, overcurrent protection strategy, and photovoltaic power generation limitation. The constraints are in the following form:

[0067]

[0068] in, for The power output of the moment converter; is the maximum capacity of the converter; for Current output of the moment converter; is the overcurrent protection threshold; for The photovoltaic power generation power of the photovoltaic system at this moment.

[0069] Specifically, the particle swarm optimization algorithm (PSO) is used to optimize the current output optimization model. Particle swarm optimization (PSO), a branch of evolutionary computation, is a random search algorithm that simulates biological activity in nature. PSO mimics the natural processes of bird and fish flocking and predation. It finds the global optimal solution through collaboration within the swarm. Proposed by American scholars Eberhart and Kennedy in 1995, PSO has been widely applied to optimization problems in various engineering fields.

[0070] Furthermore, the specific optimization calculation process includes: setting a particle swarm, a maximum number of iterations, and a termination condition, where the position of a particle represents the current output of the converter at different time points; updating the particle position and velocity through iteration; determining whether the termination condition is met, stopping iteration when the termination condition is met, and outputting the optimal solution that meets the constraints. For example, in this embodiment, the particle swarm size is 50, the maximum number of iterations is 100, and the termination condition is that the optimal solution changes by less than 0.1% over 10 consecutive iterations.

[0071] Furthermore, in each iteration of the particle swarm optimization algorithm, the current optimal solution is subjected to moving average filtering, thereby eliminating the mutation points in the current output sequence and ensuring the smoothness of the current output sequence. For example, the window size of the moving average filter can be selected as 3, then for The current output of the current transformer is 、 and The average value of the three moments is used as the smoothed current value to improve the smoothness of the current output sequence.

[0072] As an embodiment of the present application, load demand forecast data is obtained by:

[0073] S130. The photovoltaic power generation prediction data and the historical operation data are set in the same time series, and the time series is segmented using a sliding time window method.

[0074] S140. Use the ARIMA model to predict the load demand of different segments in the time series to obtain load demand forecast data.

[0075] Specifically, the sliding window technique is a common technique in data processing and algorithm design, primarily used to address problems requiring analysis or computation based on data within a fixed-size window. It has broad applications in fields such as time series analysis, stream data processing, and optimization. The sliding window technique enables processing and analysis of data streams or time series data using a fixed-size window, enabling efficient analysis and manipulation in dynamic environments. For example, in this embodiment, the window size is 24 hours, and the window sliding step is 1 hour.

[0076] Specifically, the ARIMA model is a statistical model commonly used for time series analysis and forecasting. Its full name is the Autoregressive Integrated Moving Average Model. The ARIMA model is commonly used to model and forecast time series data. It combines the three methods of autoregression (AR), difference (I) and moving average (MA). It can process non-stationary time series data and forecast it under certain conditions. Among them, the autoregressive order is Indicates the order of the autoregressive part of the model, that is, how many periods of lagged values ​​the model uses to predict the current value. , the model will use the values ​​of the previous two periods to predict the current period value. Indicates how many times the difference operation is needed to make the time series stationary. Usually, or Sufficient to meet the needs of most non-stationary series. Moving average order Indicates the order of the moving average part of the model, that is, how many periods of random error lag values ​​are used by the model to predict the current value. In actual scenarios, set appropriate parameters , which improves the complexity control of the ARIMA model when fitting historical data and reduces the probability of overfitting. In some embodiments, the parameters of the ARIMA model can be determined by various methods, including but not limited to: grid search and performance indicators (such as AIC, BIC, or RMSE) for comparing models. For example, the ARIMA model is represented as ARIMA(p, d, q). When the ARIMA model is ARIMA(2, 1, 2), it means that the model will use the autoregressive term of the previous two periods, the difference term of one period, and the moving average term of the previous two periods to predict future load power demand.

[0077] Reference Figure 2a , Figure 2a This is a diagram of the steps for setting an overcurrent protection strategy in an embodiment of the present application. As shown in the figure, as an embodiment of the present application, the overcurrent protection strategy of the converter is obtained by classification according to the current output sequence, including:

[0078] S210. Classify the current output sequence into current levels using a clustering algorithm; wherein the number of current levels is determined based on the size relationship between the silhouette coefficients corresponding to different current levels.

[0079] S220. Determine the current level intervals according to the number of current levels, and set overcurrent protection strategies for the current level intervals respectively.

[0080] Specifically, the current values ​​in the current output sequence are divided into several current levels, each of which contains multiple current value data. At this time, the optimal number of current levels is determined by calculating the contour coefficients of different numbers of current levels. The contour coefficient calculation method includes: for each sample point in the current level, calculating the average distance between the sample point and other sample points in the same current level; calculating the average distance between the sample point and all sample points in the nearest neighboring current level; and calculating the contour coefficient corresponding to the number of current levels based on the above distance calculation. The contour coefficient satisfies the following formula:

[0081]

[0082] in, is the silhouette coefficient; is the sample point; For sample points The average distance to other sample points of the same current level; For sample points The average distance to all sample points in the nearest neighbor current level. The range of the silhouette coefficient is between -1 and 1. If the silhouette coefficient is close to 1, it means that the sample point The matching degree with the current level range is high; if the silhouette coefficient is close to 0, it means that the sample point Located at the boundary of the current level range; if the silhouette coefficient is close to -1, it means that the sample point The matching degree with the current level interval is low, and the sample point may be classified into the wrong current level interval. For example, in this embodiment, the current level number with the largest corresponding silhouette coefficient is used as the final result, and the current values ​​in the current output sequence are graded according to the current level number.

[0083] For example, when the current values ​​in the current output sequence range from 0 to 300A, the number of current levels can be 2 to 10. Based on the different current levels, a clustering algorithm is used to obtain several current level intervals corresponding to different current level numbers. According to the above formula, the contour coefficients of different current level numbers are calculated. When the number of current levels is 5, the maximum contour coefficient is 0.75. Based on this, five current level intervals are obtained, with center values ​​of 50A, 100A, 150A, 200A, and 250A, respectively. The boundary values ​​of adjacent current level intervals are 75A, 125A, 175A, and 225A, respectively. Overcurrent protection thresholds are set for each current level interval. If the first current level interval is 0 to 75A, the overcurrent protection threshold for the first current level interval can be set to 90A. Similarly, if the second current level interval is 75A to 125A, the overcurrent protection threshold for the second current level interval can be set to 150A. Similarly, the overcurrent protection thresholds for the third, fourth, and fifth current level intervals can be obtained.

[0084] In some embodiments, the type of clustering algorithm can be a K-means clustering algorithm, which is a commonly used unsupervised learning algorithm used to divide samples in a data set into multiple groups or clusters. Its main goal is to divide the data into K clusters, where the data points within each cluster are similar to each other, while the similarity between data points in different clusters is low.

[0085] Furthermore, after determining the number of current levels, the current values ​​in the current output sequence are graded to obtain a plurality of current level intervals, each current level interval corresponding to a center value and a boundary value. According to the center value and boundary value of different current level intervals, an overcurrent protection threshold is set for each current level interval. The overcurrent protection threshold is a piecewise function, which increases with the increase of the current level. In the highest current level interval, the overcurrent protection threshold is the limit value of the converter. For example, if the rated current of the converter is 250A and the converter has a short-term overload capacity of 20%, the limit value of the converter is 300A. It should be noted that the overcurrent protection threshold corresponds to the tripping time one by one, and the overcurrent protection strategy is obtained according to the overcurrent protection threshold and the tripping time.

[0086] The global coordinated control method for a photovoltaic energy storage system provided in this embodiment sets an overcurrent protection threshold and a trip time, enabling the converter to output a current higher than the overcurrent protection threshold within the trip time, thereby increasing the output of the photovoltaic energy storage system and meeting the actual needs of the load. Furthermore, if the output current exceeds the overcurrent protection threshold, the converter automatically trips after the trip time, cutting off the current. This ensures the safety of each device in the photovoltaic energy storage system and reduces the probability of device failure.

[0087] Reference Figure 2b , Figure 2b This is a step diagram for adjusting the overcurrent protection strategy in an embodiment of the present application. As shown in the figure, as an embodiment of the present application, the overcurrent protection strategy is adjusted according to the actual current output of the converter to obtain a dynamic overcurrent protection strategy, including:

[0088] S230. Collect the actual current output of the converter, and determine the current level range in which the actual current output is located using a binary search algorithm.

[0089] S240. Obtain the output current change rate of the converter according to the actual current output using an exponentially weighted moving average method.

[0090] S250. Use the output current change rate as a coefficient to adjust the overcurrent protection strategy corresponding to the current level range.

[0091] Specifically, a binary search algorithm is used to quickly find a specific element in an ordered array. Its basic idea is to continuously divide the search range into two until the target element is found or determined to be non-existent. Based on the actual current output of the converter, the binary search algorithm is used to determine the current level range in which it falls, thereby determining the overcurrent protection strategy corresponding to the actual current output. In some embodiments, after the overcurrent protection strategy is determined using the binary search algorithm, it is written to a register of the converter controller and stored using a hash table structure, thereby improving data query efficiency.

[0092] Specifically, the Exponential Moving Average (EMA) method is a commonly used time series data smoothing technique used to eliminate random fluctuations in data and highlight long-term trends and cyclical changes. Compared with the simple moving average method, the exponentially weighted moving average method places more emphasis on the weight of recent data during the calculation process, and therefore can more quickly reflect the changing trend of the data. The actual current output of the converter is the actual current output by the converter within a certain period of time before the current moment. For example, the actual current output of the converter can be the current data of the converter within the past 24 hours. According to the changing trend of the actual current output of the converter, the overcurrent protection strategy is adjusted so that the overcurrent protection strategy can dynamically change according to the external environmental conditions, thereby obtaining a dynamic overcurrent protection strategy.

[0093] In some embodiments, after the dynamic overcurrent protection strategy is obtained, the overcurrent protection strategy stored in the register is updated according to the obtained dynamic overcurrent protection strategy.

[0094] Furthermore, the adjustment formula of the overcurrent protection threshold in the dynamic overcurrent protection strategy is as follows:

[0095]

[0096] in, is the adjusted overcurrent protection threshold; is the overcurrent protection threshold before adjustment; is the trend coefficient, and its value range is -0.1 to 0.1; is the adjustment coefficient, and its value range is 0.9 to 1.1. Accordingly, the adjustment formula of the trip time in the dynamic overcurrent protection strategy is as follows:

[0097]

[0098] in, is the adjusted trip time; = is the trip time before adjustment. Therefore, if the actual current output of the converter has an increasing trend, the trend coefficient is positive, and the overcurrent protection threshold increases, while the trip time decreases. If the actual current output of the converter has a decreasing trend, the trend coefficient is negative, and the overcurrent protection threshold decreases, while the trip time increases.

[0099] For example, when it is detected that the real-time current output of the converter is 180A, the binary search algorithm can be used to determine that the real-time current output is in the third current level range, wherein the third current level range is 150A to 200A. The overcurrent protection strategy corresponding to the third current level range obtained from the register includes: the overcurrent protection threshold is 220A, and the tripping time is 0.5 seconds. The output current change rate of the converter in the past 24 hours is calculated using the exponentially weighted moving average method, and the trend coefficient is 0.05. The adjustment coefficient is selected as 1.0. According to the adjustment formula of the overcurrent protection threshold, the adjusted overcurrent protection threshold is 231A; accordingly, the adjusted tripping time is 0.48 seconds. After the adjusted dynamic overcurrent protection strategy is stored in the register, the adjustment of the overcurrent protection strategy is completed. This embodiment provides a global coordinated control method for a photovoltaic energy storage system, in which a dynamic overcurrent protection strategy is obtained through the current output sequence and actual current output of the converter, and the load demand forecast data used to obtain the current output sequence is predicted by the photovoltaic power generation forecast data of the photovoltaic system and the historical operation data of the converter. Therefore, during the operation of the converter, the dynamic overcurrent protection strategy can dynamically change with changes in natural environmental factors and load demand, thereby enabling the operation of the photovoltaic system to adapt to changes in the external environment, thereby improving the energy efficiency and operational safety of the photovoltaic system.

[0100] Reference Figure 2c , Figure 2cThis is a step diagram for adaptively correcting the overcurrent protection strategy according to an embodiment of the present application. As shown in the figure, as an embodiment of the present application, after adjusting the overcurrent protection strategy according to the actual current output of the converter to obtain a dynamic overcurrent protection strategy, the following steps are also included:

[0101] S260. Collect the real-time current output of the converter, and correct the current output sequence according to the difference between the real-time current output and the current output sequence.

[0102] S270. Optimize the pulse width modulation signal of the converter according to the corrected current output sequence; wherein the optimization includes adjusting the duty cycle of the pulse width modulation signal.

[0103] Specifically, in step S260, a high-speed analog-to-digital converter is used to sample the real-time current output of the converter. The high-speed analog-to-digital converter performs sampling at a certain sampling frequency, which can be set based on the needs of the actual scenario. For example, if the real-time current output of the converter fluctuates significantly, a higher sampling frequency can be used; if the real-time current output of the converter fluctuates slightly, a lower sampling frequency can be used. For example, the high-speed analog-to-digital converter uses a 16-bit precision AD7606 chip, and its sampling frequency can be set to 5kHz, indicating that current data is collected every 200μs.

[0104] Furthermore, the real-time current output data obtained through sampling is stored in a ring buffer, which can store data within a certain period of time for subsequent reading during calculation. For example, the size of the ring buffer can be set to 1000 data points and can store data within 200ms. When calculating at each moment, a sliding window with a range of 20 data is set, and the 20 most recently stored data are read from the ring buffer. Using a fast averaging algorithm, the average of the read data is used as the real-time current output value at that moment. Based on the fast averaging algorithm, by maintaining the data in the sliding window and quickly calculating the average of the above data, the computational overhead is reduced and the computation speed is improved.

[0105] Furthermore, the real-time current output of the converter is the current output of the converter at the current moment. A Kalman filter is used to estimate the current deviation based on the real-time current output and the current output sequence. A feedforward compensation algorithm is then used to generate a correction based on the current deviation. The current deviation includes the current deviation value and the current deviation change rate. The current deviation is the difference between the real-time current output and the current output sequence, and the current deviation change rate is the change in the current deviation over time. Based on the current deviation value and the current deviation change rate, the current output sequence is corrected based on the correction value. The correction is as follows:

[0106]

[0107] in, is the correction value used to correct the current output sequence; is the current deviation value; is the current deviation change rate; and According to the dynamic adjustment of fuzzy logic controller, The adjustment range is [0.5, 2], The adjustment range is [0,0.5]. The fuzzy logic controller adjusts the current deviation value and the current deviation change rate in real time. and value.

[0108] In some embodiments, the Kalman filter uses the matrix operation library Eigen, in which the state transfer matrix is ​​set to [[1,0.001], [0,1]], the observation matrix is ​​[1,0], the initial state estimate is [0,0], and the initial error covariance matrix is ​​[[1,0], [0,1]].

[0109] In other embodiments, the fuzzy logic controller uses five input membership functions and five output membership functions, wherein the inputs include current deviation and current deviation change rate, and the outputs are and For example, if the current deviation is 2A and the current deviation change rate is 0.5A / s, the fuzzy controller outputs Adjustment amount 0.3, The final adjustment is 0.1. is 1.3, is 0.2.

[0110] Furthermore, after obtaining the correction value, it is superimposed on the original current output sequence to obtain a corrected current output sequence. The corrected current output sequence is converted into a PWM signal by a digital signal processor, which drives the power devices of the converter, thereby achieving precise current control. For example, the corrected current output sequence is converted into a PWM signal by a TI TMS320F28335 DSP processor, where the PWM frequency is set to 10 kHz and the dead time is 1 μs. This signal drives the H-bridge circuit composed of IGBTs within the converter, achieving precise control of the converter's real-time current output.

[0111] Specifically, in step S270, a space vector pulse width modulation (SVM) algorithm is used to determine a basic pulse width modulation (PWM) signal for the corrected current output sequence. The calculus of variations is used to dynamically adjust the duty cycle of the basic PWM signal, and an optimal duty cycle sequence is obtained through iterative calculation. An adaptive control algorithm is introduced, combining random pulse width modulation with dead time adjustment. The random frequency offset range and dead time adjustment coefficient are dynamically adjusted based on the current magnitude and current rate of change in the corrected current output sequence. If the current rate of change exceeds a preset threshold, the random frequency offset range is reduced while the dead time adjustment coefficient is increased, thereby achieving smooth transition and precise control of the converter's current output. Space vector pulse width modulation (SVM) is a modulation technique used to control three-phase voltage source inverters. It can effectively achieve voltage control in motor drive systems. The calculus of variations is a mathematical method for studying extreme value problems of functions. It is closely related to ordinary differential equations and partial differential equations, as well as optimization problems and variational principles in physics. The main idea of ​​the calculus of variations is to find a function that makes a certain functional reach its extreme value given a set of functions.

[0112] In some embodiments, the switching timing of each bridge arm of the converter is calculated using a space vector pulse width modulation algorithm. The corresponding sector is determined based on the voltage vector projection of the modified current output sequence in the α-β coordinate system. The sector area can be calculated using the triangle area method to calculate the action time of the modified current output sequence. Based on the obtained action time, a seven-segment switching sequence is generated to obtain a basic pulse width modulation signal.

[0113] In other embodiments, when dynamically adjusting the duty cycle of a basic pulse-width modulation signal using a variational method, the rate of change of the duty cycle is limited to or less than 2% / ms. The objective function for dynamic adjustment is the current error, i.e., the mean square error between the converter's real-time current output and a corrected current output sequence. Constraints include switching frequency and dead-time limits. Based on the objective function and constraints, the duty cycle of the basic pulse-width modulation signal is iterated to obtain an optimal duty cycle sequence.

[0114] Furthermore, the dead time calculation formula is as follows:

[0115]

[0116] in, is the adjusted dead time; As the basic dead time, take 1μs; is the adjustment coefficient, take 0.002μs / (A / s); is the absolute value of the current change rate.

[0117] Specifically, an adaptive control algorithm is introduced. When the current change rate is greater than 100A / s, the random frequency offset range is reduced to ±100Hz, and the adjustment coefficient is increased to 0.004μs / (A / s), achieving smooth transition and precise control of the output current.

[0118] For example, when the corrected current output sequence, i.e., the three-phase voltage command is [220V, -110V, -110V], the α-β coordinates obtained by the space vector pulse width modulation algorithm are [220V, 127V]. By comparing the signs and magnitudes of α and β, it can be determined that the voltage vector falls in the first sector. At this time, the triangle area method is used to calculate the action time of the basic vector, and the result is 0.8ms, Based on the obtained action time, a seven-segment switching sequence [000, 100, 110, 111, 110, 100, 000] is generated to form a basic pulse width modulation signal.

[0119] For the basic pulse width modulation signal obtained, the objective function is set as ,in, is the corrected current output sequence; is the real-time current output of the converter. The constraints are set to 0.1ms≤Ton≤0.9ms, where Ton is the on-time of the switching device. The optimal duty cycle sequence is obtained by iterating 50 times using the gradient descent method.

[0120] In random pulse width modulation, the linear congruential method is used to generate pseudo-random numbers, where the seed value uses the system clock and the random frequency offset value is updated every 100μs. In the dynamic adjustment of dead time, when the current change rate is detected to be 150A / s, the calculated dead time 0.7μs.

[0121] The adaptive control algorithm uses a fuzzy logic controller. Its inputs are the current error and current rate of change, and its outputs are corrections to the random frequency offset range adjustment coefficient and the dead time adjustment coefficient. For example, if the current error is 5A and the current rate of change is 200A / s, the fuzzy controller outputs a random frequency offset range adjustment coefficient of -0.5 and a dead time adjustment coefficient of +0.001μs / (A / s). Ultimately, the random frequency offset range becomes ±100Hz, and the dead time becomes 0.55μs and 0.85μs.

[0122] As an embodiment of the present application, determining an optimal charging and discharging strategy for the energy storage device based on state of charge data of the energy storage device, a photovoltaic power generation forecast value and a load demand forecast value of the photovoltaic system, and a global consistency constraint of the global network includes:

[0123] S310. Based on the state of charge data of the energy storage device, the photovoltaic power generation forecast value and load demand forecast value of the photovoltaic system, and the global consistency constraints of the global network, a nonlinear programming model is established with minimizing the operating cost of the energy storage device as the objective function; wherein the operating cost of the energy storage device includes the electricity purchase cost, energy storage loss and battery life.

[0124] S320. Solve the nonlinear programming model to obtain the optimal charging and discharging strategy.

[0125] Specifically, a nonlinear programming model is established based on the state of charge data of the energy storage device, the photovoltaic power generation prediction data, and the current output sequence to obtain the optimal charging and discharging strategy of the energy storage system. The objective function of the nonlinear programming model is to minimize the system operating cost of the energy storage device, including the electricity purchase cost, energy storage loss, and battery life; the constraints of the nonlinear programming model include power balance and the energy storage capacity limit, charging and discharging power limit, charging and discharging rate limit, and daily maximum charging and discharging times limit of the energy storage device. In some embodiments, the battery life as a constraint condition can be quantified by setting a penalty term for the number of charging and discharging times of the energy storage device. Exemplarily, the number of charging and discharging times of the energy storage device can be set to no more than 3 times per day to meet the battery life protection requirements.

[0126] In some embodiments, the photovoltaic power generation forecast value and load demand forecast value of the photovoltaic system are predicted by the Prophet model. When using the Prophet model, weather data can also be used as an auxiliary input to improve the prediction accuracy. The Prophet model is an open source model for time series prediction, which is particularly suitable for predicting time series data with obvious seasonal cycles, such as temperature, commodity sales, traffic flow, etc. In other embodiments, the photovoltaic power generation forecast data and load demand forecast data obtained in the above steps can also be used. Exemplarily, the Prophet model is trained using historical data from the past 30 days, taking into account periodicity and holiday effects, and combined with the temperature and cloud cover forecast data for the next 24 hours, it is predicted that the peak power of photovoltaic power generation is 50kW and the peak power demand is 80kW.

[0127] Furthermore, the nonlinear programming model is solved using a genetic algorithm combined with a branch-and-bound method. A genetic algorithm (GA) is an optimization algorithm that simulates Darwinian biological evolution and is used to solve complex search and optimization problems. Its basic concept is to search for optimal or near-optimal solutions by simulating natural selection, genetic mechanisms, and mutation. For example, in this embodiment, the genetic algorithm population size can be set to 100, the crossover probability to 0.8, and the mutation probability to 0.1. The initial solution of the genetic algorithm is obtained after 500 generations of evolution.

[0128] Furthermore, the Branch and Bound method is an algorithm for solving discrete optimization problems, mainly used in combinatorial optimization problems and mixed integer programming problems. It mainly searches the solution space step by step through the strategies of branching, bounding, and pruning to find the global optimal solution or an approximation of the optimal solution. Exemplarily, the branching strategy in this embodiment adopts the maximum score principle, and the bounds are calculated by the Lagrangian relaxation method, where the pruning rules include feasibility pruning and optimality pruning. By solving the nonlinear programming model using the above method, the optimal charge and discharge power of the energy storage device at each time point at the current moment can be obtained. The result is converted into a time-related sequence, and a scheduling instruction sequence containing the optimal charge and discharge power and time point can be obtained.

[0129] The global coordinated control method for the photovoltaic energy storage system provided in this embodiment enables the photovoltaic energy storage system to operate efficiently and economically at the current moment, reduces energy waste, lowers operating costs, and extends battery life, while ensuring grid stability and power supply and demand balance.

[0130] As an embodiment of the present application, after solving the nonlinear programming model to obtain the optimal charging and discharging strategy, the following steps are further included:

[0131] S330. If the optimal charge and discharge strategy does not meet the feasibility constraints, the optimal charge and discharge strategy is locally adjusted through a simulated annealing algorithm; wherein the feasibility constraints include the charge and discharge rate limit, depth limit, and life limit of the energy storage device.

[0132] Specifically, simulated annealing is a stochastic optimization algorithm used to find the global optimal solution or a near-optimal solution in a large search space. It is inspired by the principle of solid annealing process, in which the solid is gradually cooled at high temperature to reach a stable state.

[0133] Furthermore, the local adjustment considers the neighborhood structure, cooling strategy and termination condition, and through iteration, the adjusted optimal charging and discharging strategy satisfies all feasibility constraints.

[0134] For example, in a simulated annealing algorithm, the initial temperature is set to 100, the cooling coefficient is set to 0.95, and a neighborhood search is performed 50 times for each temperature. The neighborhood structure is a random adjustment of the charging and discharging power by ±5kW. The search is stopped when the temperature drops to 0.01 or when there is no improvement after 100 consecutive times. The final scheduling instruction sequence that satisfies all constraints is obtained.

[0135] Reference Figure 3 , Figure 3This is a diagram of the steps for obtaining a global operating strategy in an embodiment of the present application. As shown in the figure, as an embodiment of the present application, a distributed alternating direction multiplier method is used to perform a global optimization solution based on the current output sequence of the converter, the dynamic overcurrent protection strategy, and the optimal charge and discharge strategy of the energy storage device to obtain the global operating strategy of the global network at the current moment, so as to achieve distributed coordinated optimization of the converter overcurrent protection, optimal power distribution, and the charge and discharge strategy of the energy storage device, including:

[0136] S410. After obtaining the global operation strategy of the global network at the current moment, update the Lagrange multiplier and the slack variable according to the global operation strategy at the current moment and the coupling constraints between the intelligent agents.

[0137] S420. At the next moment after the current moment, the global operation strategy of the global network at the next moment is obtained by the distributed alternating direction multiplier method based on the updated Lagrange multipliers and slack variables.

[0138] S430. Iteratively calculate the global operation strategy of the global network until the global operation strategy meets the convergence criterion; wherein the convergence criterion includes the number of iterations, the change before and after the Lagrange multiplier is updated, or the residual of the operation strategy of any intelligent agent in the global operation strategy.

[0139] Specifically, the structure between agents in the global network is a graph, and the communication topology between agents is defined through an adjacency matrix. For example, the elements in the adjacency matrix can be set to 0 or 1 based on the physical connection relationship and communication capabilities between the agents, where 0 indicates no connection between the agents and 1 indicates a connection between the agents. The Gossip protocol is used for communication between agents to achieve information exchange and state synchronization. The Gossip protocol (also known as the rumor protocol) is a decentralized communication protocol used for information dissemination and state synchronization in distributed systems. Its basic concept is to achieve an eventual consensus state for all nodes in the entire system by exchanging information between randomly selected nodes. This protocol is commonly used to address issues such as data replication, event propagation, and state updates in distributed systems.

[0140] Furthermore, each agent in the global network is equipped with a local decision-making module based on a deep Q network, which includes a state space and an action space. The state space includes the operating parameters of the agent itself and the operating parameters of the agents in the adjacent nodes. The operating parameters include the agent's current operating voltage, operating current, and operating power. The action space includes the agent's power regulation and charge and discharge instructions. For example, the power regulation range can be between -10kW and 10kW, and the step size in this range can be 0.1kW, which is used to represent the power transmitted outward through the converter to achieve power distribution; the charge and discharge instructions can range from -1 to 1, where 1 indicates charging into the agent, -1 indicates discharging outside the agent, and 0 indicates that there is no power exchange.

[0141] Furthermore, the local decision module includes an adaptive adjustment mechanism, which includes reward and penalty terms. The reward terms include factors such as overcurrent protection, power allocation, and charge / discharge efficiency, and corresponding penalty terms are applied to these factors. For example, a soft penalty term is provided for overcurrent protection. If the converter's current output exceeds the overcurrent protection threshold, a first penalty term is applied. If the converter's current output exceeds a certain multiple of the overcurrent protection threshold, a second penalty term is applied. The second penalty term increases quadratically with respect to the first penalty term.

[0142] In some embodiments, the learning rate of the global optimization solution is dynamically adjusted using the Adam optimizer. For example, the learning rate is initially set to 0.001, and the penalty term is adjusted using an exponential decay method with an initial value of 1 and a decay of 0.995 times with each iteration. The global operation strategy is updated every 5 minutes using a rolling optimization method.

[0143] Furthermore, when determining the global network's current global operational strategy, if the previous moment does not exist, the Lagrange multipliers and slack variables are initialized; if the previous moment exists, the Lagrange multipliers and slack variables are updated several times from the initial values. After determining the global network's current global operational strategy, the Lagrange multipliers and slack variables are updated based on the global strategy and the coupling constraints between the agents. When the global operational strategy is determined at the next moment, the Lagrange multipliers and slack variables used are the updated values ​​at the current moment.

[0144] In some embodiments, the subproblems in the global optimization are steps S100 to S300 provided in the embodiments of the present application. Each iteration of the global optimization includes steps S100 to S300 and updates to the Lagrange multipliers and slack variables. After multiple iterations, the global operation strategy is terminated when the global operation strategy meets the convergence criterion, resulting in a final global operation strategy.

[0145] For example, the global network in this embodiment includes five photovoltaic system nodes, three energy storage device nodes, and eight converter nodes. The resulting adjacency matrix is ​​16×16, and the communication cycle between any two nodes is 100ms. The deep Q network in each node's agent adopts a fully connected network structure, with 20 input layer neurons, including 10 parameters and 10 neighbor information; 50 hidden layer neurons, and 21 output layer neurons, corresponding to 21 discrete actions.

[0146] In the local decision module, weights are set for the reward items, where the weight of overcurrent protection is 0.5, the weight of power distribution is 0.3, and the weight of charge and discharge efficiency is 0.2. The rated current of the converter in the converter node is 100A, the limit value is 120A, and the soft penalty item starts to take effect at 110A. The exponential decay rate β1 of the first-order moment estimate in the Adam optimizer is set to 0.9, the exponential decay rate β2 of the second-order moment estimate is set to 0.999, and ε is set to 10 -8 .

[0147] For the global network, the distributed alternating direction multiplier method is used to iterate 50 times. In each iteration, the subproblems of each agent are solved by the primal-dual interior point method. The maximum number of iterations is 100 and the convergence threshold is 10. -6 During one optimization run, the output power of the first PV system was 50 kW, the state of charge of the first energy storage device was 75%, and the output current of the first converter was 95 A. The Deep Q network's output action was to increase power by 2.3 kW. After distributed optimization, the final decision was to increase power by 2.1 kW, discharge the first energy storage device by 0.5 kW, and increase the output current of the first converter to 98 A. At this point, the converter's output met the overcurrent protection requirements.

[0148] The present application also provides a global coordinated control method for a photovoltaic energy storage system, which is applied to a global network comprising a plurality of intelligent agents, wherein the intelligent agents include a photovoltaic system, an energy storage device, and a converter. The method comprises the following steps:

[0149] S502. The photovoltaic power generation prediction data and the historical operation data are set in the same time series, and the time series is segmented using a sliding time window method.

[0150] S504. Use the ARIMA model to predict the load demand of different segments in the time series to obtain load demand forecast data.

[0151] S506. Establish a current output optimization model based on the load demand forecast data, the initial overcurrent protection strategy of the converter, and the optimal power allocation target; wherein, if there is a previous moment before the current moment, the initial overcurrent protection strategy is the dynamic overcurrent protection strategy of the previous moment.

[0152] S508. Determine an objective function based on the power loss and utilization rate of the converter, and perform optimization calculation on the current output optimization model based on the objective function to obtain a current output sequence.

[0153] Specifically, steps S502 through S508 combine the converter's historical operating data with the system's photovoltaic power generation forecast data to generate load demand forecast data for the PV energy storage system. This process requires advanced data analysis and machine learning algorithms to accurately predict future power demand. The accuracy of these forecasts directly impacts the effectiveness of all subsequent steps.

[0154] S510. Classify the current output sequence into current levels using a clustering algorithm; wherein the number of current levels is determined based on the size relationship between the silhouette coefficients corresponding to different current levels.

[0155] S512. Determine the current level intervals according to the number of current levels, and set overcurrent protection strategies for the current level intervals respectively.

[0156] S514. Collect the actual current output of the converter and determine the current level range in which the actual current output is located using a binary search algorithm.

[0157] S516. Obtain the output current change rate of the converter according to the actual current output using an exponentially weighted moving average method.

[0158] S518. Use the output current change rate as a coefficient to adjust the overcurrent protection strategy corresponding to the current level range.

[0159] Specifically, steps S510 to S518 include dynamically adjusting the overcurrent protection strategy according to different current levels, optimizing the output performance of the photovoltaic energy storage system while ensuring equipment safety. This not only improves the flexibility of the photovoltaic energy storage system, but also enhances its response to transient events.

[0160] S520. Collect the real-time current output of the converter, and correct the current output sequence according to the difference between the real-time current output and the current output sequence.

[0161] Specifically, step S520 introduces the concepts of online estimation and feedforward control. Online estimation is used to determine the difference between the real-time current output and the current output sequence. Feedforward control employs a feedforward compensation algorithm to determine the correction used to modify the current output sequence. This combination of online estimation and feedforward control enables the system to rapidly respond to current deviations and generate corrected control commands, thereby improving system stability and efficiency and meeting the power system's requirements for adapting to dynamic environments.

[0162] S522. Optimize the pulse width modulation signal of the converter according to the corrected current output sequence; wherein the optimization includes adjusting the duty cycle of the pulse width modulation signal.

[0163] Specifically, step S522 optimizes the pulse width modulation signal of the converter according to the corrected current control instruction, and achieves smooth transition and precise control of the converter output current by dynamically adjusting the duty cycle.

[0164] S524. Based on the state of charge data of the energy storage device, the photovoltaic power generation forecast value and load demand forecast value of the photovoltaic system, and the global consistency constraints of the global network, a nonlinear programming model is established with minimizing the operating cost of the energy storage device as the objective function; wherein the operating cost of the energy storage device includes the electricity purchase cost, energy storage loss and battery life.

[0165] S526. Solve the nonlinear programming model to obtain the optimal charging and discharging strategy.

[0166] S528. If the optimal charge and discharge strategy does not meet the feasibility constraints, the optimal charge and discharge strategy is locally adjusted through a simulated annealing algorithm; wherein the feasibility constraints include the charge and discharge rate limit, depth limit, and life limit of the energy storage device.

[0167] Specifically, steps S524 to S528 are directly related to solving the problem of how to optimize energy storage and release in photovoltaic energy storage systems. The above steps use intelligent algorithms to predict and determine when to charge or discharge, as well as the power levels of charging and discharging, to ensure that the system can operate efficiently and economically in the future time period, helping to avoid energy waste, reduce operating costs, and extend battery life, while ensuring grid stability and balance between electricity supply and demand.

[0168] S530. Through the distributed alternating direction multiplier method, a global optimization solution is performed based on the current output sequence of the converter, the dynamic overcurrent protection strategy, and the optimal charging and discharging strategy of the energy storage device, and the global operation strategy of the global network at the current moment is obtained to achieve distributed collaborative optimization of the converter overcurrent protection, optimal power distribution, and charging and discharging strategy of the energy storage device.

[0169] S532. After obtaining the global operation strategy of the global network at the current moment, update the Lagrange multiplier and slack variable according to the global operation strategy at the current moment and the coupling constraints between the intelligent agents.

[0170] S534. At the next moment after the current moment, the global operation strategy of the global network at the next moment is obtained through the distributed alternating direction multiplier method based on the updated Lagrange multipliers and slack variables.

[0171] S536. Iteratively calculate the global operation strategy of the global network until the global operation strategy meets the convergence criterion; wherein the convergence criterion includes the number of iterations, the change before and after the Lagrange multiplier is updated, or the residual of the operation strategy of any intelligent agent in the global operation strategy.

[0172] Specifically, in steps S530 to S536, the photovoltaic system, energy storage device, and converter are collaboratively optimized as intelligent agents in the network. Through communication and collaboration between intelligent agents, the global optimization goal can be achieved while maintaining the local autonomy of each component. Steps S530 to S536 solve the problem of how to coordinate the optimal charging and discharging strategy of the energy storage device with the dynamic overcurrent protection and optimal power distribution of the converter. As an embodiment of the present application, the specific iterative calculation includes repeating steps S502 to S528, and updating the Lagrange multiplier and slack variable after each iterative calculation.

[0173] Furthermore, through the global network constructed by intelligent agents, coordinated control of the photovoltaic system and energy storage device is achieved, ensuring a dynamic balance between converter overcurrent protection threshold adjustment, optimal power distribution, and energy storage charging and discharging strategies. This not only improves the overall energy efficiency and safety of the photovoltaic energy storage system, but also optimizes the utilization efficiency of the energy storage device, ultimately achieving dual optimization of economic and environmental benefits.

[0174] Accordingly, please refer to Figure 4 , Figure 4 This is a module diagram of a global coordinated control system for a photovoltaic energy storage system provided in an embodiment of the present application. As shown in the figure, an embodiment of the present application provides a global coordinated control system for a photovoltaic energy storage system, which is applied to a global network including several intelligent agents, including photovoltaic systems, energy storage devices, and converters. The system includes:

[0175] The current sequence determination module 100 is used to obtain the current output sequence of the converter according to the load demand prediction data; wherein the load demand prediction data is predicted based on the photovoltaic power generation prediction data of the photovoltaic system and the historical operation data of the converter.

[0176] The protection threshold adjustment module 200 is used to obtain the overcurrent protection strategy of the converter by classification according to the current output sequence, and adjust the overcurrent protection strategy according to the real-time operation data of the converter to obtain a dynamic overcurrent protection strategy.

[0177] The energy storage strategy determination module 300 is used to determine the optimal charging and discharging strategy of the energy storage device based on the state of charge data of the energy storage device, the photovoltaic power generation forecast value and load demand forecast value of the photovoltaic system, and the global consistency constraints of the global network; wherein the global consistency constraints include power balance constraints and equipment operation restrictions.

[0178] The global strategy determination module 400 is used to perform a global optimization solution based on the current output sequence of the converter, the dynamic overcurrent protection strategy, and the optimal charge and discharge strategy of the energy storage device through a distributed alternating direction multiplier method, thereby obtaining the global operation strategy of the global network at the current moment, so as to achieve distributed collaborative optimization of the converter overcurrent protection, optimal power distribution, and the charge and discharge strategy of the energy storage device.

[0179] As an embodiment of the present application, the current sequence determination module 100 includes a model building unit and an optimization calculation unit, wherein:

[0180] The model building unit is used to establish a current output optimization model based on load demand prediction data and the initial overcurrent protection strategy and optimal power allocation target of the converter; wherein, if there is a previous moment before the current moment, the initial overcurrent protection strategy is the dynamic overcurrent protection strategy of the previous moment.

[0181] The optimization calculation unit is used to determine the objective function according to the power loss and utilization rate of the converter, and optimize the current output optimization model according to the objective function to obtain the current output sequence.

[0182] As an embodiment of the present application, the current sequence determination module 100 further includes a data segmentation unit and a load prediction unit, wherein:

[0183] The data segmentation unit is used to set the photovoltaic power generation prediction data and historical operation data in the same time series, and use the sliding time window method to segment the time series.

[0184] The load forecasting unit is used to use the ARIMA model to forecast the load demand of different segments in the time series to obtain load demand forecast data.

[0185] As an embodiment of the present application, the protection threshold adjustment module 200 includes a current grading unit and a strategy setting unit, wherein:

[0186] The current grading unit is used to grade the current output sequence into current grades by using a clustering algorithm; wherein the number of current grades is determined according to the size relationship between the silhouette coefficients corresponding to different current grades.

[0187] The strategy setting unit is used to determine the current level interval according to the number of current levels, and set the overcurrent protection strategy for each current level interval.

[0188] As an embodiment of the present application, the protection threshold adjustment module 200 further includes an interval determination unit, a change calculation unit, and a strategy adjustment unit, wherein:

[0189] The interval determination unit is used to collect the actual current output of the converter and determine the current level interval in which the actual current output is located through a binary search algorithm.

[0190] The change calculation unit is used to obtain the output current change rate of the converter according to the actual current output by using an exponentially weighted moving average method.

[0191] The strategy adjustment unit is used to adjust the overcurrent protection strategy corresponding to the current level range by taking the output current change rate as a coefficient.

[0192] As an embodiment of the present application, the protection threshold adjustment module 200 further includes a sequence correction unit and a signal optimization unit, wherein:

[0193] The sequence correction unit is used to collect the real-time current output of the converter and correct the current output sequence according to the difference between the real-time current output and the current output sequence.

[0194] The signal optimization unit is used to optimize the pulse width modulation signal of the converter according to the corrected current output sequence; wherein the optimization includes adjusting the duty cycle of the pulse width modulation signal.

[0195] As an embodiment of the present application, the energy storage strategy determination module 300 includes a planning model establishment unit and a planning model solving unit, wherein:

[0196] The planning model establishment unit is used to establish a nonlinear programming model with the objective function of minimizing the operating cost of the energy storage device based on the charge state data of the energy storage device, the photovoltaic power generation forecast data, the load demand forecast value of the photovoltaic system, and the global consistency constraints of the global network. The operating cost of the energy storage device includes the electricity purchase cost, energy storage loss and battery life.

[0197] The planning model solving unit is used to solve the nonlinear programming model to obtain the optimal charging and discharging strategy.

[0198] As an embodiment of the present application, the energy storage strategy determination module 300 also includes a local adjustment unit for adjusting the optimal charge and discharge strategy. If the optimal charge and discharge strategy does not meet the feasibility constraints, the optimal charge and discharge strategy is locally adjusted through a simulated annealing algorithm; wherein the feasibility constraints include the charge and discharge rate limit, depth limit, and life limit of the energy storage device.

[0199] As an embodiment of the present application, the global policy determination module 400 includes a global update unit, a policy update unit, and an iteration unit, wherein:

[0200] The global update unit is used to update the Lagrange multiplier and the slack variable according to the global operation strategy at the current moment and the coupling constraints between the intelligent agents after obtaining the global operation strategy of the global network at the current moment.

[0201] The strategy updating unit is used to obtain the global operation strategy of the global network at the next moment through the distributed alternating direction multiplier method based on the updated Lagrange multipliers and slack variables at the next moment after the current moment.

[0202] The iterative unit is used to iteratively calculate the global operation strategy of the global network until the global operation strategy meets the convergence criterion; wherein the convergence criterion includes the number of iterations, the change in the Lagrange multiplier before and after the update, or the residual of the operation strategy of any intelligent agent in the global operation strategy.

[0203] See also Figure 5 , Figure 5 This is a schematic diagram of the structure of a computer device provided in an embodiment of the present application. Figure 5 As shown, the computer device includes: one or more processors 10, memory 20, and interfaces for connecting various components, including high-speed interfaces and low-speed interfaces. Various components utilize different buses to communicate with each other and can be installed on a common mainboard or installed in other ways as needed. The processor can process the instructions executed in the computer device, including instructions stored in the memory or on the memory to display the graphical information of the GUI on an external input / output device (such as, a display device coupled to the interface). In some optional embodiments, if necessary, multiple processors and / or multiple buses can be used together with multiple memories and multiple memories. Equally, multiple computer devices can be connected, and each device provides part of the necessary operations (for example, as a server array, a group of blade servers, or a multi-processor system). Figure 5 A processor 10 is taken as an example.

[0204] The processor 10 may be a central processing unit (CPU), a network processor (NPU), or a combination thereof. The processor 10 may further include a hardware chip. The hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof. The programmable logic device (PLD) may be a complex programmable logic device (CPLD), a field programmable gate array (FPGA), a general purpose array logic (GAL), or any combination thereof. The memory 20 stores instructions executable by at least one processor 10, causing the at least one processor 10 to implement the methods described in the above embodiments.

[0205] The memory 20 may include a program storage area and a data storage area, wherein the program storage area may store an operating system and application programs required for at least one function; the data storage area may store data created based on the use of the computer device, etc. In addition, the memory 20 may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some optional embodiments, the memory 20 may optionally include a memory remotely located relative to the processor 10, and these remote memories may be connected to the computer device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0206] The memory 20 may include volatile memory, such as random access memory; the memory may also include non-volatile memory, such as flash memory, a hard disk, or a solid-state drive; the memory 20 may also include a combination of the aforementioned types of memory. The computer device also includes a communication interface 30 for communicating with other devices or a communication network.

[0207] The embodiments of the present application also provide a computer-readable storage medium. The above-mentioned method according to the embodiment of the present application can be implemented in hardware, firmware, or implemented as a computer code that can be recorded in a storage medium, or implemented as a computer code that is originally stored in a remote storage medium or a non-temporary machine-readable storage medium and downloaded through a network and will be stored in a local storage medium, so that the method described herein can be stored in such software processing on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only storage memory, a random access memory, a flash memory, a hard disk or a solid-state drive, etc.; further, the storage medium can also include a combination of the above-mentioned types of memory. It can be understood that a computer, a processor, a microprocessor controller or programmable hardware includes a storage component that can store or receive software or computer code. When the software or computer code is accessed and executed by a computer, a processor or hardware, the method shown in the above embodiment is implemented.

[0208] An embodiment of the present application provides a computer program product, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform a method according to any embodiment of the present application.

[0209] The systems, devices, modules, or units described in the above embodiments may be implemented by computer chips or entities, or by products having certain functions. A typical implementation device is a computer. Specifically, the computer may be, for example, a personal computer, a laptop computer, a cellular phone, a camera phone, a smartphone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.

[0210] For the convenience of description, the above devices are described as being divided into various units according to their functions. Of course, when implementing this application, the functions of each unit can be implemented in the same or multiple software and / or hardware.

[0211] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware. Furthermore, the present application may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0212] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0213] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0214] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0215] It should also be noted that the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, commodity, or apparatus that includes a series of elements includes not only those elements but also other elements not explicitly listed, or includes elements inherent to such process, method, commodity, or apparatus. In the absence of further limitations, an element defined by the phrase "comprises a ..." does not exclude the presence of other identical elements in the process, method, commodity, or apparatus that includes the element.

[0216] The various embodiments in this specification are described in a progressive manner. Similar parts between the various embodiments can be referred to in conjunction with each other. Each embodiment focuses on the differences between the other embodiments. In particular, the system embodiments are generally similar to the method embodiments, so the description is relatively simple. For relevant parts, refer to the description of the method embodiments.

[0217] The above are merely embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application should all be included within the scope of the claims of the present application.

[0218] Although the embodiments of the present application have been described with reference to the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present application, and such modifications and variations shall fall within the scope defined by the appended claims.

Claims

1. A global coordinated control method for a photovoltaic energy storage system, characterized in that: Applied to a global network comprising several intelligent agents, the intelligent agents including a photovoltaic system, an energy storage device, and a converter; the method comprises: Obtaining a current output sequence of the converter according to load demand forecast data; wherein the load demand forecast data is predicted based on photovoltaic power generation forecast data of the photovoltaic system and historical operation data of the converter; According to the current output sequence, an overcurrent protection strategy of the converter is obtained by grading, and the overcurrent protection strategy is adjusted according to the actual current output of the converter to obtain a dynamic overcurrent protection strategy, including: grading the current output sequence into current levels; determining current level intervals according to the number of current levels, and setting overcurrent protection strategies for the current level intervals respectively; wherein the dynamic overcurrent protection strategy includes an overcurrent protection threshold and a tripping time, the overcurrent protection threshold being a trigger point for overcurrent protection, corresponding to different current sizes in the current output sequence, and the number of current levels being determined according to the size relationship between contour coefficients corresponding to different current levels; Determining an optimal charging and discharging strategy for the energy storage device based on the state of charge data of the energy storage device, the photovoltaic power generation forecast value and the load demand forecast value of the photovoltaic system, and the global consistency constraints of the global network; wherein the global consistency constraints include power balance constraints and equipment operation restrictions, and the optimal charging and discharging strategy is determined based on the operating state of the converter in the event that the operating state of the converter changes after the current output sequence and dynamic overcurrent protection strategy of the converter are determined; Through the distributed alternating direction multiplier method, a global optimization solution is performed based on the current output sequence of the converter, the dynamic overcurrent protection strategy and the optimal charging and discharging strategy of the energy storage device, and the global operation strategy of the global network at the current moment is obtained to achieve distributed collaborative optimization of the converter overcurrent protection, optimal power distribution and the charging and discharging strategy of the energy storage device.

2. The method according to claim 1, characterized in that The step of obtaining the current output sequence of the converter according to load demand prediction data includes: Establishing a current output optimization model based on the load demand forecast data, the initial overcurrent protection strategy of the converter, and the optimal power allocation target; wherein, if there is a previous moment before the current moment, the initial overcurrent protection strategy is the dynamic overcurrent protection strategy of the previous moment; An objective function is determined according to the power loss and utilization rate of the converter, and the current output optimization model is optimized and calculated according to the objective function to obtain the current output sequence.

3. The method according to claim 2, characterized in that The load demand forecast data is obtained by: The photovoltaic power generation prediction data and the historical operation data are set in the same time series, and the time series is segmented using a sliding time window method; The ARIMA model is used to predict the load demands of different segments in the time series to obtain the load demand prediction data.

4. The method according to claim 1, wherein The step of obtaining the overcurrent protection strategy of the converter by grading according to the current output sequence includes: The current output sequence is classified into current levels using a clustering algorithm.

5. The method according to claim 1, wherein The step of adjusting the overcurrent protection strategy according to the actual current output of the converter to obtain a dynamic overcurrent protection strategy includes: Collecting the actual current output of the converter, and determining the current level interval in which the actual current output is located by using a binary search algorithm; Obtaining an output current change rate of the converter according to the actual current output by an exponentially weighted moving average method; The output current change rate is used as a coefficient to adjust the overcurrent protection strategy corresponding to the current level range.

6. The method according to claim 5, characterized in that After the overcurrent protection strategy is adjusted according to the actual current output of the converter to obtain a dynamic overcurrent protection strategy, the method further includes: collecting a real-time current output of the converter, and correcting the current output sequence according to a difference between the real-time current output and the current output sequence; The pulse width modulation signal of the converter is optimized according to the corrected current output sequence; wherein the optimization includes adjusting the duty cycle of the pulse width modulation signal.

7. The method according to claim 1, characterized in that The determining of the optimal charging and discharging strategy of the energy storage device based on the state of charge data of the energy storage device, the photovoltaic power generation forecast value and the load demand forecast value of the photovoltaic system, and the global consistency constraint of the global network includes: Establishing a nonlinear programming model with minimizing the operating cost of the energy storage device as an objective function based on the state of charge data of the energy storage device, the photovoltaic power generation forecast value and the load demand forecast value of the photovoltaic system, and the global consistency constraint of the global network; wherein the operating cost of the energy storage device includes the electricity purchase cost, energy storage loss, and battery life; The nonlinear programming model is solved to obtain the optimal charging and discharging strategy.

8. The method according to claim 7, characterized in that After solving the nonlinear programming model to obtain the optimal charge and discharge strategy, the method further includes: If the optimal charge and discharge strategy does not meet the feasibility constraints, the optimal charge and discharge strategy is locally adjusted through a simulated annealing algorithm; wherein the feasibility constraints include the charge and discharge rate limit, depth limit and life limit of the energy storage device.

9. The method according to claim 1, characterized in that The distributed alternating direction multiplier method is used to perform global optimization based on the current output sequence of the converter, the dynamic overcurrent protection strategy, and the optimal charge and discharge strategy of the energy storage device to obtain the global operation strategy of the global network at the current moment, so as to achieve distributed collaborative optimization of converter overcurrent protection, optimal power distribution, and energy storage device charge and discharge strategy, including: After obtaining the global operation strategy of the global network at the current moment, updating the Lagrange multiplier and the slack variable according to the global operation strategy at the current moment and the coupling constraints between the intelligent agents; At a moment next to the current moment, a global operation strategy of the global network at the next moment is obtained by a distributed alternating direction multiplier method based on the updated Lagrange multipliers and slack variables; The global operation strategy of the global network is iteratively calculated until the global operation strategy satisfies a convergence criterion; wherein the convergence criterion includes the number of iterations, the change before and after the update of the Lagrange multiplier, or the residual of the operation strategy of any of the intelligent agents in the global operation strategy.

10. A global coordinated control system for a photovoltaic energy storage system, characterized in that: Applied to a global network comprising several intelligent agents, including a photovoltaic system, an energy storage device, and a converter; the system comprises: a current sequence determination module, configured to obtain a current output sequence of the converter according to load demand forecast data; wherein the load demand forecast data is obtained by forecasting photovoltaic power generation forecast data of the photovoltaic system and historical operation data of the converter; A protection threshold adjustment module is used to obtain an overcurrent protection strategy for the converter by grading the current output sequence, and adjust the overcurrent protection strategy according to the real-time operating data of the converter to obtain a dynamic overcurrent protection strategy, including: grading the current output sequence into current levels; determining current level intervals according to the number of current levels, and setting overcurrent protection strategies for the current level intervals respectively; wherein the dynamic overcurrent protection strategy includes an overcurrent protection threshold and a tripping time, the overcurrent protection threshold being a trigger point for overcurrent protection, corresponding to different current sizes in the current output sequence, and the number of current levels being determined according to the size relationship between the contour coefficients corresponding to different current levels; an energy storage strategy determination module, configured to determine an optimal charge and discharge strategy for the energy storage device based on the state of charge data of the energy storage device, the photovoltaic power generation forecast value and load demand forecast value of the photovoltaic system, and the global consistency constraints of the global network; wherein the global consistency constraints include power balance constraints and equipment operation restrictions, and the optimal charge and discharge strategy is determined based on the operating state of the converter in the event that the operating state of the converter changes after the current output sequence and dynamic overcurrent protection strategy of the converter are determined; A global strategy determination module is used to perform a global optimization solution based on the current output sequence of the converter, the dynamic overcurrent protection strategy, and the optimal charge and discharge strategy of the energy storage device using a distributed alternating direction multiplier method to obtain the global operation strategy of the global network at the current moment, thereby achieving distributed collaborative optimization of the converter overcurrent protection, optimal power distribution, and energy storage device charge and discharge strategy.

Images