A dynamic energy storage scheduling method for a modular photovoltaic system

Abstract

This invention discloses a dynamic energy storage scheduling method for a modular photovoltaic system, relating to the field of photovoltaic power generation system operation and control technology. The method involves dividing the photovoltaic power generation system into multiple autonomously operating modules. Each module includes a photovoltaic power generation unit, an energy storage unit, and a local control unit. The local control unit collects real-time data on the module's current irradiance, battery state of charge, temperature, grid power flow, and load power, and periodically uploads these parameters to the system's central coordination unit. The central coordination unit constructs a nonlinear load prediction model based on module-level load rate changes, external temperature sensitivity, and diurnal cycle characteristics to improve the response capability to temporary sudden load changes. This invention solves the problems of low operating efficiency, uneven energy storage utilization, inaccurate load prediction, and delayed scheduling response.

Description

Technical Field

[0001] This invention relates to the field of photovoltaic power generation system operation and control technology, specifically to a dynamic energy storage scheduling method for a modular photovoltaic system. Background Technology

[0002] With the global energy structure transformation and the vigorous development of renewable energy, photovoltaic power generation, as a clean and renewable energy form, is widely used in industrial parks, commercial buildings, residential communities, and other scenarios. Especially in industrial parks with large rooftop resources, the deployment of distributed photovoltaic systems enables on-site power generation and consumption, effectively reducing energy costs and improving energy efficiency.

[0003] However, photovoltaic (PV) power generation is intermittent and volatile, easily affected by weather and temporal changes. To improve system stability and self-consumption rate, energy storage systems are typically installed in conjunction with PV systems. Currently, modular PV and energy storage designs are widely used in industrial parks or microgrids, dividing the system into multiple functional modules or areas, each with independent PV modules and energy storage units. This modular design facilitates system expansion and maintenance, but it also brings new challenges to energy storage dispatch.

[0004] In existing technologies, most photovoltaic + energy storage systems still rely on static or semi-static scheduling strategies, such as timed charging and discharging rules or simple "peak shaving and valley filling" logic. These methods cannot be flexibly adjusted according to actual operating conditions, resulting in problems such as low system operating efficiency, uneven energy storage utilization, inaccurate load forecasting, and delayed scheduling response. Summary of the Invention

[0005] The purpose of this invention is to provide a dynamic energy storage scheduling method for modular photovoltaic systems to solve the problems mentioned in the background art.

[0006] To address the aforementioned technical problems, this invention provides the following technical solution: a dynamic energy storage scheduling method for a modular photovoltaic system, comprising the following steps:

[0007] Step 1: Divide the photovoltaic power generation system into multiple modules with autonomous operation capabilities. Each module includes a photovoltaic power generation unit, an energy storage unit, and a local control unit. The local control unit is used to collect the current light intensity, battery state of charge, temperature, grid power flow direction, and load power data of the module in real time, and periodically upload the above parameters to the system central coordination unit.

[0008] Step 2: The central coordination unit constructs a nonlinear load prediction model based on module-level load rate changes, external temperature sensitivity, and diurnal cycle characteristics to improve the response capability to temporary and sudden load changes;

[0009] Step 3: When it is detected that more than a threshold number of modules simultaneously have excess photovoltaic power generation capacity while the load is low within a certain time window, the central coordination unit determines whether the system has entered the photovoltaic power surge zone based on the analysis and calculation results.

[0010] Step 4: Once the photovoltaic power impact zone is determined, the energy storage weight function constructed by the suppression and excitation composite function is used to perform priority scheduling on each energy storage unit.

[0011] Step 5: The system sorts the modules according to their energy storage weight function values. Under the conditions of meeting SoC constraints, battery temperature rise constraints and local load requirements, the system prioritizes scheduling high-weight modules to receive excess power generation for charging.

[0012] Step Six: When the system detects an abnormal increase in load, a load forecast value that is significantly higher than the current power generation value, or when the power grid issues a demand control command, the central coordination unit initiates the emergency discharge scheduling process, prioritizing the scheduling of energy storage units with high SoC, suitable temperature, and large lifespan redundancy to participate in discharge support.

[0013] According to the above technical solution, in step one, the local control unit periodically collects key operating status parameters of its module and constructs a standardized data structure for unified uploading, specifically including:

[0014] The local control unit uses a built-in light intensity sensor, current and voltage detection module, temperature sensor and power meter to perform high-frequency sampling of photovoltaic output power, real-time voltage and current of energy storage unit, energy storage battery temperature, load power and grid feed / extraction power parameters, and uploads them after averaging and filtering within a fixed time window.

[0015] The uploaded data structure includes a module unique identifier ID, sampling timestamp, parameter field set, and module status flag. The parameter field set covers: current photovoltaic power generation, energy storage unit state of charge, current load power, ambient temperature, battery temperature, grid interaction direction and power value. The status flag is used to indicate whether the module is in a special state such as communication interruption, energy storage failure, or battery temperature rise warning.

[0016] Data upload adopts a combination of event-driven mechanism and periodic synchronization mechanism. That is, while uploading basic operating parameters in each period, asynchronous fast upload is immediately triggered when a parameter change is detected, so that the central coordination unit can dynamically adjust the scheduling strategy according to abnormal changes.

[0017] According to the above technical solution, the local control unit autonomously identifies and marks the module's operating status, and reports this status to the central coordination unit in real time through a status flag field. This is used to assist in the implementation of subsequent energy storage scheduling and safety protection strategies. The status flag includes at least:

[0018] The communication status flag is used to indicate whether the communication between the module and the central coordination unit is normal. If no scheduling instruction is received for several consecutive cycles or the number of failed transmissions exceeds the limit, it will be automatically set to the "communication interruption" state.

[0019] The energy storage health status indicator is used to indicate whether the current energy storage unit has fault symptoms such as high internal resistance, voltage drift, abnormal charging and discharging, or the number of cycles reaching the warning value. When the health status is lower than the set threshold, the indicator is automatically set to "energy storage degradation" or "energy storage failure".

[0020] Thermal status indicators automatically generate "temperature rise warning" or "thermal runaway risk" indicators when the temperature is abnormal or there is a trend of excessively rapid temperature rise by jointly analyzing the battery temperature collection value and the temperature rise rate.

[0021] The load status flag is used to identify whether the local load is in a state of peak operation, sudden start-stop or irregular oscillation. If multiple power fluctuations exceed the limit in a short period of time, the system will mark it as a "load disturbance" state.

[0022] When any module reports an anomaly, the central coordination unit will implement scheduling degradation, scheduling ban, or local isolation strategies for that module in the energy storage scheduling sequence, and issue maintenance suggestions to ensure the stability and security of the overall system operation, and has good fault tolerance and online self-recovery capabilities.

[0023] According to the above technical solution, in step two, the calculation expression of the nonlinear load prediction model is:

[0024]

[0025] in, This indicates the actual load power in the current cycle; This indicates the current rate of load change. This indicates the temperature sensitivity factor, reflecting the tendency of the load to change with temperature; This indicates the current hour, used to simulate the day-night cycle. These are all system self-tuning coefficients, dynamically adjusted based on historical accuracy.

[0026] According to the above technical solution, the nonlinear prediction model in the central coordination unit used to predict the future load power of each module has an initialization self-learning, rolling update, and multi-period comparison mechanism in actual operation, specifically including:

[0027] When the system is first run or a new module is added, the load prediction model builds an initial prediction template based on hourly load power data over the past three days. At the same time, it automatically associates the corresponding time period's light level, ambient temperature, and actual electricity price to form an initial combination of feature factors, and then fits the module-level electricity consumption behavior characteristics accordingly.

[0028] Each time the model completes a prediction cycle, it triggers a parameter correction process, comparing the deviation between the predicted value and the actual load power of that cycle. If the deviation exceeds a set threshold, the weighting factor is automatically adjusted and some prediction factors are recalibrated.

[0029] The system is equipped with a multi-cycle comparison mechanism. Every 24-hour operation cycle, the predicted trajectory of the day is dynamically aligned with the actual load curve, and the daily average relative error value and trend deviation are calculated as the benchmark for updating the initial template for the next day, so as to form a prediction correction closed loop on a daily basis.

[0030] According to the above technical solution, the analytical calculation expression used by the central coordination unit in step three is as follows:

[0031]

[0032] in, Indicates the total number of modules; This represents the photovoltaic power generation of the i-th module at time t; This represents the local load power of the i-th module at time t; This is a critical threshold jointly determined by the system capacity and the maximum access power of energy storage; it is reached when the sum of the total photovoltaic power generation of all modules in the system minus their total local load exceeds the threshold. This indicates that the system has entered the photovoltaic power surge zone.

[0033] According to the above technical solution, the mechanism for determining whether the system has entered the photovoltaic power impact zone in step three adopts a combination of hierarchical judgment and dynamic threshold, specifically including:

[0034] The preliminary judgment standard is that the total excess photovoltaic power of the modules exceeds 75% of the energy storage access capacity;

[0035] Dynamic threshold Adjustments will be made in real time based on the following three factors:

[0036] The current maximum charging power of all energy storage units;

[0037] The ratio of predicted peak load to average load;

[0038] Permissible power access value at grid connection point and electricity price incentive coefficient;

[0039] When the total excess power exceeds the dynamic threshold, it is considered to have entered the "photovoltaic power surge zone".

[0040] According to the above technical solution, in step four, the energy storage weight function constructed using the composite function of suppression and excitation has the following calculation expression:

[0041]

[0042] in, This indicates the current state of charge of the i-th energy storage unit; Indicates the current battery temperature; This indicates the optimal temperature setting; Indicates the remaining percentage of the cycle; This indicates the current real-time electricity price; This represents the scheduling incentive coefficient, used to adjust the commercialization weight.

[0043] According to the above technical solution, in step five, when a high-weight module fails to schedule due to a sudden anomaly or physical access bottleneck, the system immediately initiates an inter-module power redundancy migration mechanism, which includes:

[0044] A "real-time power reception capability query interface" is introduced, allowing each module to upload its maximum power reception value every cycle. When the system detects that the original target module is unavailable, it will adjust the current modules accordingly. The value is used to redistribute redundant power;

[0045] The original redundant power Divided into several sub-power units Prioritize distributing the power in batches according to the remaining power margin that the subordinate modules can safely receive;

[0046] If the total receiving capacity of all modules is still less than [a certain value] in the current scheduling period The system starts the "power abstraction scheduling cache pool" to temporarily record the remaining redundant power and prioritizes the allocation of newly added modules when starting the next scheduling cycle.

[0047] If multiple modules request to accept the same redundant power within the synchronization time window, the coordination unit will sort the conflicting modules according to their most recent scheduling participation frequency, energy storage health factor, and response speed.

[0048] Compared with existing technologies, the beneficial effects achieved by this invention are as follows: By introducing an architecture combining module-level control and global coordination, this invention enables the system to achieve rapid response and dynamic energy balance under various complex scenarios such as light fluctuations, load abrupt changes, and grid commands. The innovative energy storage weighting function and power redundancy migration mechanism effectively improve the scheduling efficiency and lifespan of energy storage resources, reducing curtailment rate and energy loss. Simultaneously, the constructed nonlinear load prediction model possesses strong temperature sensitivity analysis and periodic behavior capture capabilities, making system regulation more forward-looking and stable, significantly improving the intelligent operation level and economic benefits of the photovoltaic system. Attached Figure Description

[0049] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0050] Figure 1 This is a flowchart illustrating the dynamic energy storage scheduling method for the modular photovoltaic system of the present invention. Detailed Implementation

[0051] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0052] Please see Figure 1 The present invention provides a technical solution: a dynamic energy storage scheduling method for a modular photovoltaic system, comprising the following steps:

[0053] Step 1: Divide the photovoltaic power generation system into multiple modules with autonomous operation capabilities. Each module includes a photovoltaic power generation unit, an energy storage unit, and a local control unit. The local control unit is used to collect the current light intensity, battery state of charge, temperature, grid power flow direction, and load power data of the module in real time, and periodically upload the above parameters to the system's central coordination unit.

[0054] Step 2: The central coordination unit constructs a nonlinear load prediction model based on module-level load rate changes, external temperature sensitivity, and diurnal cycle characteristics to improve the response capability to temporary and sudden load changes;

[0055] Step 3: When it is detected that more than a threshold number of modules simultaneously have excess photovoltaic power generation capacity while the load is low within a certain time window, the central coordination unit determines whether the system has entered the photovoltaic power surge zone based on the analysis and calculation results.

[0056] Step 4: Once the photovoltaic power impact zone is determined, the energy storage weight function constructed by the suppression and excitation composite function is used to perform priority scheduling on each energy storage unit.

[0057] Step 5: The system sorts the modules according to their energy storage weight function values. Under the conditions of meeting SoC constraints, battery temperature rise constraints and local load requirements, the system prioritizes scheduling high-weight modules to receive excess power generation for charging.

[0058] Step Six: When the system detects an abnormal increase in load, a load forecast value significantly higher than the current power generation value, or a demand control command issued by the grid, the central coordination unit initiates the emergency discharge scheduling process, prioritizing the scheduling of energy storage units with high SoC, suitable temperature, and large lifespan redundancy to participate in discharge support. Through periodic collection of module-level information and centralized processing of global information, on-demand scheduling and dynamic balancing of energy storage resources are achieved under different operating scenarios. It has cross-module linkage, adaptive adjustment capabilities, and scalability, which can improve photovoltaic utilization, extend energy storage life, reduce electricity costs, and significantly improve the operational economy and intelligent response level of the entire photovoltaic-storage system while ensuring system stability.

[0059] In step one, the local control unit periodically collects key operating status parameters of its module and constructs a standardized data structure for unified uploading, specifically including:

[0060] The local control unit uses a built-in light intensity sensor, current and voltage detection module, temperature sensor and power meter to perform high-frequency sampling of photovoltaic output power, real-time voltage and current of energy storage unit, energy storage battery temperature, load power and grid feed / pump power parameters, and uploads them after averaging and filtering within a fixed time window (preferably 5 to 30 seconds).

[0061] The uploaded data structure includes a module unique identifier ID, sampling timestamp, parameter field set, and module status flag. The parameter field set covers: current photovoltaic power generation, energy storage unit state of charge, current load power, ambient temperature, battery temperature, grid interaction direction and power value. The status flag is used to indicate whether the module is in a special state such as communication interruption, energy storage failure, or battery temperature rise warning.

[0062] Data upload adopts a combination of event-driven mechanism and periodic synchronization mechanism. That is, while uploading basic operating parameters in each period, when a sudden change in parameters is detected (such as load fluctuation exceeding the threshold or rapid rise in battery temperature), asynchronous fast upload is immediately triggered, so that the central coordination unit can dynamically adjust the scheduling strategy according to abnormal changes.

[0063] The local control unit autonomously identifies and marks the module's operating status, and reports this status to the central coordination unit in real time via a status flag field. This information is used to assist in the implementation of subsequent energy storage scheduling and safety protection strategies. The status flag includes at least the following:

[0064] The communication status flag is used to indicate whether the communication between the module and the central coordination unit is normal. If no scheduling instruction is received for several consecutive cycles or the number of failed transmissions exceeds the limit, it will be automatically set to the "communication interruption" state.

[0065] The energy storage health status indicator is used to indicate whether the current energy storage unit has fault symptoms such as high internal resistance, voltage drift, abnormal charging and discharging, or the number of cycles reaching the warning value. When the health status is lower than the set threshold, the indicator is automatically set to "energy storage degradation" or "energy storage failure".

[0066] Thermal status indicators automatically generate "temperature rise warning" or "thermal runaway risk" indicators when the temperature is abnormal or there is a trend of excessively rapid temperature rise by jointly analyzing the battery temperature collection value and the temperature rise rate.

[0067] The load status flag is used to identify whether the local load is in a state of peak operation, sudden start-stop or irregular oscillation. If multiple power fluctuations exceed the limit in a short period of time, the system will mark it as a "load disturbance" state.

[0068] When any module reports an anomaly, the central coordination unit will implement scheduling degradation, scheduling ban, or local isolation strategies for that module in the energy storage scheduling sequence, and issue maintenance suggestions to ensure the stability and security of the overall system operation, and has good fault tolerance and online self-recovery capabilities.

[0069] In step two, the calculation expression for the nonlinear load prediction model is:

[0070]

[0071] in, This indicates the actual load power in the current cycle; This indicates the current rate of load change. This indicates the temperature sensitivity factor, reflecting the tendency of the load to change with temperature; This indicates the current hour (e.g., 3 PM is 15), used to simulate the diurnal cycle. These are all system self-tuning coefficients, dynamically adjusted based on historical accuracy.

[0072] In this formula, by directly introducing a nonlinear growth term and a periodic correction term, prediction enhancement can be achieved for sudden load spikes (such as equipment startup) and specific time periods (such as production peaks), and the first-order load difference is also incorporated. It can identify "slope abrupt changes", enhance the response to "critical bursts", and clearly demonstrate the physical mechanism of load changes caused by temperature changes, avoiding the problem of large errors in simple historical averages.

[0073] The nonlinear prediction model used in the central coordination unit to predict the future load power of each module has a self-learning initialization mechanism, a rolling update mechanism, and a multi-period comparison mechanism in actual operation, specifically including:

[0074] When the system is first run or a new module is added, the load prediction model builds an initial prediction template based on hourly load power data over the past three days. At the same time, it automatically associates the corresponding time period's light level, ambient temperature, and actual electricity price to form an initial combination of feature factors, and then fits the module-level electricity consumption behavior characteristics accordingly.

[0075] Each time the model completes a prediction cycle, it triggers a parameter correction process, comparing the deviation between the predicted value and the actual load power of that cycle. If the deviation exceeds a set threshold, the weighting factor is automatically adjusted and some prediction factors are recalibrated to ensure that the model maintains its ability to quickly adapt to real-world fluctuations.

[0076] The system is equipped with a multi-cycle comparison mechanism. Every 24-hour operation cycle, the predicted trajectory of the day is dynamically aligned with the actual load curve, and the daily average relative error value and trend deviation are calculated as the benchmark for updating the initial template for the next day, so as to form a prediction correction closed loop on a daily basis.

[0077] In step three, the analytical calculation expression used by the central coordination unit is as follows:

[0078]

[0079] in, Indicates the total number of modules; This represents the photovoltaic power generation of the i-th module at time t; This represents the local load power of the i-th module at time t; This is a critical threshold jointly determined by the system capacity and the maximum access power of energy storage; it is reached when the sum of the total photovoltaic power generation of all modules in the system minus their total local load (i.e., the total excess photovoltaic power of the system) exceeds the threshold. This indicates that the system has entered the photovoltaic power surge zone.

[0080] Step three, which determines whether the system has entered the photovoltaic power surge zone, employs a combination of hierarchical judgment and dynamic thresholds, specifically including:

[0081] The preliminary judgment standard is that the total excess photovoltaic power of the modules exceeds 75% of the energy storage access capacity;

[0082] Dynamic threshold Adjustments will be made in real time based on the following three factors:

[0083] The current maximum charging power of all energy storage units;

[0084] The ratio of predicted peak load to average load;

[0085] Permissible power access value at grid connection point and electricity price incentive coefficient;

[0086] When the total excess power exceeds the dynamic threshold, it is considered to have entered the "photovoltaic power surge zone".

[0087] In step four, the energy storage weight function constructed using the composite function of suppression and excitation is expressed as follows:

[0088]

[0089] in, This indicates the current state of charge of the i-th energy storage unit; Indicates the current battery temperature; This indicates the optimal temperature setting; Indicates the remaining percentage of the cycle; This indicates the current real-time electricity price; This represents the scheduling incentive coefficient, used to adjust the commercialization weight;

[0090] In the formula, an exponential function is used to "suppress" the extreme state of the SoC and "stimulate" the low power state to prevent overcharging or over-discharging; temperature regulation adopts a sigmoid-like denominator function to give dynamic penalties to batteries that deviate from the optimal temperature, introduces a logarithmic term of the battery life factor to reduce priority when approaching the end of the life and prevent overuse in critical stages, and finally uses the electricity price as the final multiplicative incentive term to clearly highlight the role of economic factors during the dominant period.

[0091] In step five, if a high-weight module fails to schedule due to a sudden anomaly or physical access bottleneck, the system immediately initiates an inter-module power redundancy migration mechanism, which includes:

[0092] A "real-time power reception capability query interface" is introduced, allowing each module to upload its maximum power reception value every cycle. This value takes into account the current SoC, temperature, battery aging status, and DC-DC operating current limitations. When the system detects that the original target module is unavailable, it adjusts the current module's current configuration accordingly. The value is used to redistribute redundant power;

[0093] The original redundant power Divided into several sub-power units The power is distributed in batches according to the amount of power margin that the subordinate modules can still safely receive. In order to avoid overshoot or internal overheating of the target modules caused by instantaneous power injection, the central coordination unit adopts a batch transfer strategy to achieve seamless transitional power injection.

[0094] If the total receiving capacity of all modules is still less than [a certain value] in the current scheduling period The system starts the "power abstract scheduling cache pool" to temporarily record the remaining redundant power, and prioritizes the allocation of newly added modules when starting the next scheduling cycle, so as to achieve energy delay utilization rather than direct curtailment of light as much as possible;

[0095] If multiple modules request to accept the same redundant power within the synchronization time window, the coordination unit sorts the conflicting modules according to their most recent scheduling participation frequency, energy storage health factor, and response speed, ensuring that the optimal target gets the injection right, thereby improving the overall redundancy processing efficiency.

[0096] Through the inter-module power redundancy migration mechanism, the energy storage scheduling is transformed from single-module injection to multi-module collaborative evolution, which significantly enhances the scheduling fault tolerance and energy utilization of the photovoltaic system under uncertain loads and variable environments, and ensures that the system maintains high efficiency and responsiveness under abnormal scenarios.

[0097] This application proposes a dynamic energy storage scheduling method for modular photovoltaic (PV) systems, enabling self-sensing of state among multiple modules, nonlinear load prediction, and intelligent collaborative scheduling of energy storage during system operation. By introducing an architecture combining module-level control and global coordination, the system achieves rapid response and dynamic energy balance under various complex scenarios such as solar radiation fluctuations, load abrupt changes, and grid commands. Innovative energy storage weighting functions and power redundancy migration mechanisms effectively improve the scheduling efficiency and lifespan of energy storage resources, reducing curtailment rates and energy losses. Simultaneously, the constructed nonlinear load prediction model possesses strong temperature sensitivity analysis and periodic behavior capture capabilities, making system regulation more forward-looking and stable, significantly improving the intelligent operation level and economic benefits of the PV system.

[0098] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1A device that provides the functions specified in one or more boxes.

[0099] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0100] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0101] The embodiments of the present invention have been described above with reference to the accompanying drawings. However, the present invention is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of the present invention without departing from the spirit and scope of the claims. All of these forms are within the protection scope of the present invention.

Claims

1. A dynamic energy storage scheduling method for a modular photovoltaic system, characterized in that, Includes the following steps: Step 1: Divide the photovoltaic power generation system into multiple modules with autonomous operation capabilities. Each module includes a photovoltaic power generation unit, an energy storage unit, and a local control unit. The local control unit is used to collect the current light intensity, battery state of charge, temperature, grid power flow direction, and load power data of the module in real time, and periodically upload the above parameters to the system central coordination unit. Step 2: The central coordination unit constructs a nonlinear load prediction model based on module-level load rate changes, external temperature sensitivity, and diurnal cycle characteristics to improve the response capability to temporary and sudden load changes; Step 3: When it is detected that more than a threshold number of modules simultaneously have excess photovoltaic power generation capacity while the load is low within a certain time window, the central coordination unit determines whether the system has entered the photovoltaic power surge zone based on the analysis and calculation results. Step 4: Once the photovoltaic power impact zone is determined, the energy storage weight function constructed by the suppression and excitation composite function is used to perform priority scheduling on each energy storage unit. Step 5: The system sorts the modules according to their energy storage weight function values. Under the conditions of meeting SoC constraints, battery temperature rise constraints and local load requirements, the system prioritizes scheduling high-weight modules to receive excess power generation for charging. Step 6: When the system detects an abnormal increase in load, a load forecast value that is significantly higher than the current power generation value, or when the power grid issues a demand control command, the central coordination unit initiates the emergency discharge dispatch process, prioritizing the dispatch of energy storage units with high SoC, suitable temperature, and large lifespan redundancy to participate in discharge support. In step four, the energy storage weight function constructed using the composite function of suppression and excitation is calculated as follows: in, This indicates the current state of charge of the i-th energy storage unit; Indicates the current battery temperature; This indicates the optimal temperature setting; Indicates the remaining percentage of the cycle; This indicates the current real-time electricity price; This represents the scheduling incentive coefficient, used to adjust the commercialization weight.

2. The dynamic energy storage scheduling method for a modular photovoltaic system according to claim 1, characterized in that: In step one, the local control unit periodically collects key operating status parameters of its module and constructs a standardized data structure for unified uploading, specifically including: The local control unit uses a built-in light intensity sensor, current and voltage detection module, temperature sensor and power meter to perform high-frequency sampling of photovoltaic output power, real-time voltage and current of energy storage unit, energy storage battery temperature, load power and grid feed / extraction power parameters, and uploads them after averaging and filtering within a fixed time window. The uploaded data structure includes a module unique identifier ID, sampling timestamp, parameter field set, and module status flag. The parameter field set covers: current photovoltaic power generation, energy storage unit state of charge, current load power, ambient temperature, battery temperature, grid interaction direction and power value. The status flag is used to indicate whether the module is in a communication interruption, energy storage fault, or battery temperature rise warning state. Data upload adopts a combination of event-driven mechanism and periodic synchronization mechanism. That is, while uploading basic operating parameters in each period, asynchronous fast upload is immediately triggered when a parameter change is detected, so that the central coordination unit can dynamically adjust the scheduling strategy according to abnormal changes.

3. The dynamic energy storage scheduling method for a modular photovoltaic system according to claim 2, characterized in that: The local control unit autonomously identifies and marks the module's operating status, and reports this status to the central coordination unit in real time via a status flag field. This assists in the implementation of subsequent energy storage scheduling and safety protection strategies. The status flag includes at least the following: The communication status flag is used to indicate whether the communication between the module and the central coordination unit is normal. If no scheduling instruction is received for several consecutive cycles or the number of failed transmissions exceeds the limit, it will be automatically set to the "communication interruption" state. The energy storage health status indicator is used to indicate whether the current energy storage unit has fault symptoms such as high internal resistance, voltage drift, abnormal charging and discharging, or reaching the warning value of the number of cycles. When the health status is lower than the set threshold, the indicator is automatically set to "energy storage degradation" or "energy storage failure". Thermal status indicators, through joint analysis of battery temperature acquisition values ​​and temperature rise rate, automatically generate "temperature rise warning" or "thermal runaway risk" indicators when the temperature is abnormal or there is an excessively rapid temperature rise trend; The load status flag is used to identify whether the local load is operating at peak, experiencing sudden start-stop, or exhibiting irregular fluctuations. If multiple power fluctuations exceed the limit within a short period of time, the system will mark it as a "load disturbance" state. When any module reports an anomaly, the central coordination unit will implement scheduling degradation, scheduling ban, or local isolation strategies for that module in the energy storage scheduling sequence, and issue maintenance suggestions to ensure the stability and security of the overall system operation, and has good fault tolerance and online self-recovery capabilities.

4. The dynamic energy storage scheduling method for a modular photovoltaic system according to claim 1, characterized in that: In step two, the calculation expression for the nonlinear load prediction model is: in, This indicates the actual load power in the current cycle; This indicates the current rate of load change. This indicates the temperature sensitivity factor, reflecting the tendency of the load to change with temperature; This indicates the current hour, used to simulate the day-night cycle. These are all system self-tuning coefficients, dynamically adjusted based on historical accuracy.

5. The dynamic energy storage scheduling method for a modular photovoltaic system according to claim 4, characterized in that: The nonlinear prediction model used in the central coordination unit to predict the future load power of each module has a self-learning initialization mechanism, a rolling update mechanism, and a multi-period comparison mechanism in actual operation, specifically including: When the system is first run or a new module is added, the load prediction model builds an initial prediction template based on hourly load power data over the past three days. At the same time, it automatically associates the corresponding time period's light level, ambient temperature, and actual electricity price to form an initial combination of feature factors, and then fits the module-level electricity consumption behavior characteristics accordingly. Each time the model completes a prediction cycle, it triggers a parameter correction process, comparing the deviation between the predicted value and the actual load power of that cycle. If the deviation exceeds a set threshold, the weighting factor is automatically adjusted and some prediction factors are recalibrated. The system is equipped with a multi-cycle comparison mechanism. Every 24-hour operation cycle, the predicted trajectory of the day is dynamically aligned with the actual load curve, and the daily average relative error value and trend deviation are calculated as the benchmark for updating the initial template for the next day, so as to form a prediction correction closed loop on a daily basis.

6. The dynamic energy storage scheduling method for a modular photovoltaic system according to claim 1, characterized in that: In step three, the analytical calculation expression used by the central coordination unit is as follows: in, Indicates the total number of modules; This represents the photovoltaic power generation of the i-th module at time t; This represents the local load power of the i-th module at time t; This is a critical threshold jointly determined by the system capacity and the maximum access power of energy storage; it is reached when the sum of the total photovoltaic power generation of all modules in the system minus their total local load exceeds the threshold. This indicates that the system has entered the photovoltaic power surge zone.

7. The dynamic energy storage scheduling method for a modular photovoltaic system according to claim 6, characterized in that: The mechanism for determining whether the system has entered the photovoltaic power surge zone in step three adopts a combination of hierarchical judgment and dynamic threshold, specifically including: The preliminary judgment standard is that the total excess photovoltaic power of the modules exceeds 75% of the energy storage access capacity; Dynamic threshold Adjustments will be made in real time based on the following three factors: (1) The maximum charging power of all current energy storage units; (2) The ratio of predicted peak load to average load; (3) Permissible power access value at grid connection point and electricity price incentive coefficient; When the total excess power exceeds the dynamic threshold, it is considered to have entered the "photovoltaic power surge zone".

8. The dynamic energy storage scheduling method for a modular photovoltaic system according to claim 1, characterized in that: In step five, if a high-weight module fails to schedule due to a sudden anomaly or physical access bottleneck, the system immediately initiates an inter-module power redundancy migration mechanism, which includes: Introducing a "real-time power reception capability query interface," each module uploads its maximum power reception value every cycle. When the system detects that the original target module is unavailable, it will adjust the current modules accordingly. The value is used to redistribute redundant power; The original redundant power Divided into several sub-power units Prioritize distributing the power in batches according to the remaining power margin that the subordinate modules can safely receive; If the total receiving capacity of all modules is still less than [a certain value] in the current scheduling period The system starts the "power abstraction scheduling cache pool" to temporarily record the remaining redundant power and prioritizes the allocation of newly added modules when starting the next scheduling cycle. If multiple modules request to accept the same redundant power within the synchronization time window, the coordination unit will sort the conflicting modules according to their most recent scheduling participation frequency, energy storage health factor, and response speed.

Images