Energy scheduling method for optical energy storage

Abstract

The embodiment of the invention provides an energy scheduling method for optical energy storage, and relates to the technical field of energy scheduling. The energy scheduling method for optical energy storage comprises the following steps: acquiring power generation power, component temperature, irradiance and environment temperature data of a photovoltaic component; according to the historical electricity utilization record and the real-time load monitoring information, predicting the electricity utilization demand of each piece of load equipment in the next period; detecting the residual capacity, the instantaneous charging and / or discharging power and the health state of the energy storage system; based on the photovoltaic power generation information, the load demand prediction result and the energy storage state, calculating the charging and discharging strategy of the next period by using an optimization algorithm to obtain an optimal energy scheduling scheme; and distributing and controlling the charging / discharging power of the energy storage system according to the optimal energy scheduling scheme. According to the invention, peak load shifting of the energy storage system can be effectively realized, and the power consumption cost is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of energy scheduling, and in particular to an energy scheduling method for photoelectric energy storage. Background Art

[0002] With the continuous development of renewable energy technology, the application scale of photovoltaic power generation is expanding globally. However, due to the fact that photovoltaic power generation is greatly affected by meteorological factors, its output is random and intermittent. In order to smooth out the fluctuations in photovoltaic output and achieve accurate matching of load-side power supply needs, energy storage systems have been widely used in photovoltaic grid-connected systems. By storing electrical energy during low-load or photovoltaic power generation peak periods and releasing the stored energy during high-load periods, it can effectively reduce peak loads and fill valleys, improve the operating economy of the system, and enhance the stability of the power grid.

[0003] In the existing technology, the energy scheduling of photovoltaic energy storage systems usually needs to comprehensively consider factors such as photovoltaic power generation forecast, energy storage system capacity and health status, load demand forecast on the load side, and peak and valley electricity prices of the power grid. Traditional scheduling methods often rely on fixed rules or simple optimization models, which are difficult to cope with complex and changeable meteorological conditions and load demands. In addition, the real-time and refinement of scheduling strategies have gradually become important factors affecting the efficiency and economic benefits of photovoltaic energy storage systems.

[0004] At present, in the field of solar energy storage energy scheduling, common technical means include:

[0005] Forecasting electricity demand and power generation based on time series or statistical models: For example, using autoregressive models, moving average models, or machine learning-based forecasting models (such as support vector machines, neural networks, etc.) to make short-term or medium- to long-term forecasts of load and photovoltaic power generation;

[0006] Energy scheduling using mathematical programming methods such as optimal power flow or linear programming: Optimize the charging and discharging strategy of the solar energy storage system under certain constraints to maximize energy utilization efficiency or minimize operating costs;

[0007] Adopt certain control strategies to protect the health of the energy storage system: for example, combine temperature management, charge and discharge rate control, overcharge and overdischarge protection, etc. to extend the service life of the energy storage battery;

[0008] Peak shaving and valley filling based on time-of-use electricity prices or demand response signals: better economic benefits can be achieved by discharging during peak electricity price periods and charging during valley electricity price periods.

[0009] However, with the expansion of the scale of photovoltaic energy storage systems and the iterative upgrading of technologies, existing scheduling methods still have certain limitations in terms of stability, precision and real-time performance. Especially in the face of multiple load conditions, uncertain meteorological factors and multiple operating modes (grid-connected / off-grid switching), traditional rules or simple models are difficult to take into account both the safety and economy of the system. In order to further improve the overall efficiency and reliability of photovoltaic energy storage systems, an energy scheduling method that can integrate multiple factors such as load forecasting, power generation forecasting, energy storage status, dynamic electricity prices and carbon emissions is needed to more effectively coordinate the management of photovoltaic power generation and energy storage systems, and achieve optimal scheduling of multiple objectives in different scenarios. It is based on this demand that an energy scheduling method for photovoltaic energy storage is proposed, in order to maximize the use of clean energy and improve the environmental benefits of overall operation while ensuring system safety and reducing operating costs. Summary of the invention

[0010] According to an embodiment of the present invention, a method for scheduling energy of a photovoltaic energy storage is provided to solve the problems raised by the above background technology.

[0011] In a first aspect of the present invention, a method for scheduling energy of a photovoltaic energy storage is provided, comprising the following steps:

[0012] S1. Obtain the power generation, component temperature, irradiance and ambient temperature data of the photovoltaic module;

[0013] S2. Based on historical power consumption records and real-time load monitoring information, predict the power demand of each load device in the next cycle;

[0014] S3. Detect the remaining capacity, instantaneous charging and / or discharging power and health status of the energy storage system;

[0015] S4. Based on the photovoltaic power generation information, load demand forecast results and energy storage status, the optimization algorithm is used to calculate the charging and discharging strategy for the next cycle to obtain the optimal energy scheduling plan;

[0016] S5. Allocate and control the charging / discharging power of the energy storage system according to the optimal energy scheduling scheme. If the surplus photovoltaic power generation is greater than the power required by the energy storage, the surplus power will be delivered to the load or the power grid; if the photovoltaic power generation cannot meet the load demand, the energy storage system and / or the power grid will make up for it;

[0017] S6. In the process of implementing the scheduling plan, the energy scheduling strategy is dynamically modified according to the real-time data and prediction deviation, and the modified results are stored in the system database for use in the prediction and scheduling optimization of the next cycle.

[0018] Preferably, the load demand prediction in step S2 adopts a prediction algorithm based on time series analysis or a machine learning model, and the machine learning model includes one or more models such as a long short-term memory network LSTM, a support vector machine SVM or a random forest.

[0019] Preferably, the optimization algorithm in step S4 includes one or more of linear programming, mixed integer programming, genetic algorithm or particle swarm algorithm, and solves the objective function of the energy scheduling scheme according to constraints such as electricity prices, peak and valley electricity rates, demand response signals, etc.

[0020] Preferably, in the process of executing step S4, different scheduling schemes are further comprehensively evaluated in combination with the carbon emission factor, and energy scheduling strategies with lower carbon emissions are preferentially selected.

[0021] Preferably, the scheduling execution in step S5 is completed by a distributed control system, which includes a plurality of sub-control units for respectively controlling the switching action and power regulation of the photovoltaic grid-connected inverter, the energy storage converter and each load device.

[0022] Preferably, the feedback and correction in step S6 include fuzzy control or adaptive control algorithms, and the scheduling strategy is corrected in real time according to the power generation characteristics of the photovoltaic components and the changes in power consumption on the load side.

[0023] Preferably, when calculating the energy allocation strategy, the dispatchable load is time-sharingly controlled according to the current peak and valley electricity price information of the power grid, including but not limited to starting, pausing or adjusting the motor speed and air conditioning cooling capacity, so as to achieve the economic dispatching goal of peak shaving and valley filling.

[0024] Preferably, the method further includes the step of evaluating the health status of the energy storage system. When capacity attenuation or temperature abnormality is detected in the energy storage system, the safety protection strategy is triggered and the energy scheduling plan is derated or switched to protect the normal operation of the energy storage equipment.

[0025] Preferably, the energy scheduling method further comprises:

[0026] When abnormal power quality or grid failure is detected, it automatically switches to off-grid mode, giving priority to the power demand of key loads in off-grid mode;

[0027] When the power grid returns to normal, the system automatically switches back to the grid-connected mode and re-executes the optimization and scheduling of step S4 and step S5.

[0028] Preferably, it also includes photo-storage data management based on blockchain or edge computing technology, wherein the management includes encrypting and storing photovoltaic power generation, energy storage charging and discharging status, and power load data in a trusted ledger to achieve a traceable and auditable energy scheduling process when multiple entities collaborate.

[0029] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0030] The present invention can effectively realize peak shaving and valley filling of the energy storage system, reduce electricity costs, reduce power consumption in peak hours, and fully utilize low-priced electricity for energy storage in valley hours by comprehensively considering factors such as photovoltaic power generation prediction results, load demand prediction, peak-valley difference in electricity prices, and demand response signals in the energy scheduling method, thereby significantly improving the operating economy of the photovoltaic energy storage system.

[0031] The present invention incorporates real-time evaluation and safety protection strategies for the health status of the energy storage system (such as capacity attenuation, temperature anomalies, etc.) into the scheduling algorithm of the optical energy storage, and can automatically switch to the off-grid mode to prioritize the power demand of key loads when abnormal power quality or failure of the power grid is detected; when the power grid returns to normal, it can automatically connect to the grid and resume normal scheduling. The above mechanism ensures the continuous and stable operation of the system in various emergency situations and improves the overall safety and reliability.

[0032] In the process of energy scheduling optimization, the present invention further combines carbon emission factors to evaluate and select different scheduling schemes, so that while meeting economic goals, it can effectively reduce carbon emissions caused by the use of high-carbon electricity. For scenarios with carbon trading or carbon emission quota management requirements, the present invention has more significant environmental benefits and policy adaptability.

[0033] The present invention introduces time series analysis or machine learning algorithms (such as LSTM, SVM, random forest, etc.) in the load demand forecasting link. By fully mining and learning the historical power consumption data and real-time monitoring information of the load, it can maintain a high prediction accuracy when facing complex and changeable load characteristics, providing a reliable basis for accurate decision-making of subsequent scheduling strategies.

[0034] In the energy scheduling strategy of the present invention, in addition to optimally allocating the charging and discharging schemes of the energy storage system, the present invention can also combine the peak and valley electricity prices of the power grid, demand response signals, etc., to start and stop or adjust the power of the dispatchable loads in a time-sharing manner, thereby giving full play to the potential of demand-side response, further reducing energy consumption during peak hours and improving the cost-effectiveness of electricity use during valley hours.

[0035] It should be understood that the contents described in the summary of the invention are not intended to limit the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. DETAILED DESCRIPTION

[0036] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described in combination with the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0037] In addition, the term "and / or" in this article is only a description of the association relationship between the associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship.

[0038] In a first aspect of the present invention, a method for scheduling energy of a photovoltaic energy storage is provided, comprising the following steps:

[0039] S1. Configure a data acquisition terminal in the photovoltaic module grid-connected system to detect the real-time output power of the photovoltaic module, the module operating temperature, the ambient irradiance and the ambient temperature;

[0040] The data acquisition terminal can communicate bidirectionally with the cloud server or local controller to ensure that the above data can be uploaded to the central control system in real time or periodically.

[0041] Filter, denoise and remove outliers on the collected raw data to reduce the impact of sampling errors and sensor failures;

[0042] The processed data is stored in the system database to provide basic support for subsequent prediction and scheduling strategies. The dynamic change characteristics of photovoltaic power generation can be grasped in a timely manner, providing reliable photovoltaic power generation information input for the entire energy scheduling process.

[0043] S2. Obtain the historical power consumption records and real-time operating power of each key load through a smart meter or load monitoring module;

[0044] Different data sampling periods and accuracies can be set for different types of loads (such as industrial production equipment, air conditioning and refrigeration systems, lighting equipment, etc.).

[0045] Load demand forecasting model

[0046] Time series analysis models: such as the autoregressive moving average model (ARMA) and the autoregressive integrated moving average model (ARIMA), which are used for short-term or medium- to long-term load forecasting;

[0047] Machine learning models: including long short-term memory networks (LSTM), support vector machines (SVM), or random forests;

[0048] You can select a suitable model based on the actual application scenario, or you can fuse multiple models to improve prediction accuracy.

[0049] The predicted electricity demand for the next cycle (such as the next hour or the next day) will serve as an important reference for subsequent energy storage charging and discharging strategies and load-side regulation strategies; if there is a sudden large load increase (such as starting and stopping large industrial equipment), dynamic scheduling can be performed in step S6 (scheduling correction).

[0050] S3. Detect the remaining capacity, instantaneous charge and / or discharge power and health status of the energy storage system. Obtain the remaining capacity (SoC), maximum / minimum allowable charge and discharge power (P_max / P_min) and temperature information of the energy storage battery (such as lithium-ion battery, lithium titanate battery, sodium-sulfur battery, etc.) through the energy storage management system (BMS) or battery management controller; detect the voltage, current, temperature, internal resistance and other parameters of the energy storage cell to evaluate the health status (SoH) of the energy storage system.

[0051] When it is detected that the energy storage system has excessive capacity attenuation, abnormal temperature, or overvoltage / undervoltage, a fault warning is triggered and corresponding protection operations are performed, such as limiting the charge and discharge rate or cutting off the connection between the battery and the load; this information can also be considered in the calculation of subsequent scheduling strategies to ensure the safety and life of the energy storage system.

[0052] S4. Based on the photovoltaic power generation information, load demand forecast results and energy storage status, the optimization algorithm is used to calculate the charging and discharging strategy for the next cycle to obtain the optimal energy scheduling plan.

[0053] Scheduling objectives and constraints:

[0054] Economic goal: reduce operating costs, mainly reflected in maximizing the benefits of peak-shaving and valley-filling electricity charges or minimizing the cost per kilowatt-hour;

[0055] Safety goal: Ensure the safe operation of the energy storage system (avoid overcharging, overdischarging, and abnormal temperature) and meet the power demand of the load;

[0056] Carbon emission target: In further optimization of this implementation, carbon emission factors can be comprehensively introduced to give priority to scheduling strategies with lower carbon emissions;

[0057] Constraints include energy storage capacity limitations, photovoltaic output power limitations, peak and valley electricity price settings, demand response signals, and other system-level safety constraints.

[0058] The optimization algorithm can be but is not limited to:

[0059] Mathematical programming algorithms: including linear programming, mixed integer programming, etc.;

[0060] Intelligent optimization algorithms: such as genetic algorithm, particle swarm algorithm, ant colony algorithm, etc.;

[0061] The above algorithms can be selected or combined to obtain a better scheduling solution based on the system scale, computational complexity and real-time requirements.

[0062] The photovoltaic power generation forecast value (obtained from step S1) and the load demand forecast value (obtained from step S2) are used as input, combined with the energy storage system status (obtained from step S3) and the electricity price and demand response information, to build a scheduling optimization model; under a given number of iterations or convergence conditions, by solving the optimal solution that satisfies all constraints in the feasible domain, the optimal charging and discharging strategy for the next cycle and the control plan for the dispatchable part of the load side (if there is a demand side response mechanism) are output.

[0063] Furthermore, the carbon emission factor is combined to comprehensively evaluate different dispatching schemes. grid 、Photovoltaic power generation carbon emission coefficient α pv , Energy storage manufacturing and operation carbon emission coefficient α ess ; Add carbon emission target (C_carbon) to economic target (C_cost) and find Pareto optimal solution through weighted or multi-objective optimization;

[0064] If the single objective comprehensive function: According to the comprehensive cost F = w1 × C cost +w2×C carbon The optimal solution is selected based on the size of; if multi-objective optimization: according to the Pareto frontier solution, a scheduling solution with a good balance between economic benefits and emission reduction effects is selected. Among them, w1 is the weight factor.

[0065] S5. Allocate and control the charging / discharging power of the energy storage system according to the optimal energy scheduling plan. If the surplus photovoltaic power generation is greater than the power required for energy storage, the surplus electric energy will be delivered to the load or the power grid. If the photovoltaic power generation cannot meet the load demand, it will be supplemented by the energy storage system and / or the power grid.

[0066] Furthermore, based on the optimal energy dispatching instruction outputted in step S4, the energy storage converter / inverter is controlled to perform corresponding charging and discharging operations; when the photovoltaic power generation is greater than the charging power required by the energy storage system, the excess power is first supplied to the local load, and if there is any surplus, it is sent back to the grid; when photovoltaic power generation cannot fully meet the load demand, the energy storage system outputs power to supplement it, and if the energy storage is still insufficient, the grid supplies power. Through a distributed controller or a central control platform, the operating status of the photovoltaic grid-connected inverter, the energy storage system, and the load-side dispatchable equipment (such as variable speed motors, air conditioning loads, etc.) is uniformly coordinated to ensure that the optimal energy flow distribution is achieved on the basis of meeting safety and stability.

[0067] S6. In the process of implementing the scheduling plan, the energy scheduling strategy is dynamically modified according to the real-time data and prediction deviation, and the modified results are stored in the system database for use in the prediction and scheduling optimization of the next cycle.

[0068] Furthermore, the photovoltaic power generation output, actual load power consumption and energy storage operation status are continuously collected and compared with the predicted and calculated values ​​of step S4; if a large deviation is found between the predicted value and the actual value (such as sudden changes in meteorological conditions, abnormal changes in power load), the scheduling correction mechanism is triggered. Using online algorithms (such as those based on rolling time domain optimization or adaptive control algorithms), the charging and discharging power of the energy storage system and the energy distribution on the load side are dynamically adjusted while ensuring system safety and basic constraints remain unchanged; a rapid response mechanism can be triggered for large disturbance scenarios (such as sudden changes in weather and power grid failures) to achieve priority protection of important loads or emergency protection of energy storage systems. The triggering cause, correction amplitude and effect evaluation of each scheduling correction are stored in the database; in the next scheduling cycle, the prediction accuracy of photovoltaic power generation and load demand is improved by updating the model parameters or correcting the prediction algorithm, and the scheduling strategy is further optimized.

[0069] In this embodiment, the load demand forecast in step S2 adopts a forecasting algorithm based on time series analysis and / or machine learning to improve the forecasting accuracy and adaptability of the load demand in the next cycle; wherein the machine learning model includes (but is not limited to) long short-term memory network (LSTM), support vector machine (SVM) or random forest (RF), etc., which can be flexibly selected or used in combination according to different load types and operating characteristics.

[0070] In this embodiment, the optimization algorithm in step S4 includes one or more of linear programming, mixed integer programming, genetic algorithm or particle swarm algorithm, and solves the objective function of the energy scheduling plan according to constraints such as electricity prices, peak and valley electricity rates, demand response signals, etc.

[0071] In this embodiment, during the execution of step S4, different scheduling schemes are further comprehensively evaluated in combination with the carbon emission factor, and energy scheduling strategies with lower carbon emissions are preferentially selected.

[0072] In this embodiment, the scheduling execution in step S5 is completed by a distributed control system, which includes a plurality of sub-control units for respectively controlling the switching action and power regulation of the photovoltaic grid-connected inverter, the energy storage converter and each load device.

[0073] Furthermore, the distributed control system is composed of several sub-control units, each of which can be distributed near the photovoltaic grid-connected inverter, energy storage converter and various load devices; each sub-control unit has data collection, instruction execution and local fault diagnosis functions to ensure basic independent control when network communication is delayed or partially interrupted; under the unified coordination of the central controller or energy management system (EMS), each sub-control unit can exchange operating status and power demand information with each other to achieve coordinated management of the entire system.

[0074] According to the optimal dispatching scheme outputted in step S4, the active and reactive power outputs of the grid-connected inverter are dynamically adjusted, such as limiting the grid-connected power or improving the power factor; according to the charging and discharging instructions issued by the central controller, the energy storage system is controlled to execute the corresponding power output or absorption within a safe range, and the status of the energy storage battery (voltage, temperature, etc.) is monitored in real time; it is responsible for the switching action or power regulation of various energy-consuming equipment (such as motors, air conditioners, lighting, etc.); for dispatchable loads, such as variable speed motors, they can be started and stopped or the frequency can be adjusted according to the time-of-use electricity price and dispatching instructions; for air-conditioning and refrigeration equipment, the cooling capacity or operating frequency can be adjusted in real time according to the load demand and the data of indoor and outdoor temperature sensors.

[0075] After obtaining the optimal energy scheduling plan in step S4, the central controller sends specific scheduling commands to each sub-control unit, including: grid-connected inverter output limit, energy storage charging and discharging power reference value, schedulable load start and stop or power level, etc.; each sub-control unit executes the assigned task based on the current equipment operating status and safety thresholds (such as battery SoC upper limit, equipment rated power, etc.) without violating safety constraints; if an emergency abnormality (such as excessive battery temperature, load failure) occurs, the sub-control unit will immediately start local protection and send an alarm signal to the central controller.

[0076] In this embodiment, the feedback and correction in step S6 include fuzzy control or adaptive control algorithms, and the scheduling strategy is corrected in real time according to the power generation characteristics of the photovoltaic components and the changes in power consumption on the load side.

[0077] Furthermore, the distributed control system continuously monitors the power generation of the photovoltaic modules, the status of the energy storage system and the actual power consumption on the load side, and compares them with the expected value of the scheduling strategy in step S4; if the deviation exceeds the pre-set threshold (such as the difference between the predicted power and the actual power is too large, the energy storage charging and discharging state changes abnormally, etc.), the scheduling correction process is triggered.

[0078] When the system detects uncertainty or strong random disturbances in meteorological conditions or load demand, fuzzy control algorithms can be used to flexibly adjust photovoltaic power generation, energy storage output and load scheduling; when system demand or external environment presents nonlinear and time-varying characteristics, adaptive control algorithms (such as adaptive fuzzy, online parameter identification, etc.) can be introduced to timely update control parameters and perform closed-loop corrections; by analyzing fuzzy rules or adaptive laws of multi-dimensional variables such as input deviation, rate of change, and energy storage / load, output instructions (such as energy storage charging and discharging power, air conditioning cooling capacity, and load start and stop timing) can be dynamically adjusted.

[0079] During the feedback correction process, the system can perform rolling forecasts at the same time: continuously update the photovoltaic output and load demand for the next time period (such as 15 minutes or 30 minutes later), re-call or simplify the optimization algorithm in step S4, and perform rolling updates on the recent scheduling strategy; if the cause of the deviation is confirmed to be a long-term or structural change (such as seasonal load changes), the parameters of the time series model or machine learning model are synchronously iterated and updated to ensure the accuracy of subsequent scheduling forecasts.

[0080] In this embodiment, when the energy allocation strategy is calculated, the dispatchable load is time-sharingly controlled according to the current peak-valley electricity price information of the power grid, including but not limited to starting, pausing or adjusting the motor speed and air conditioning cooling capacity, so as to achieve the economic dispatching goal of peak shaving and valley filling.

[0081] Obtain the latest time-of-use electricity price plan from the power company or electricity price supervision platform, including the electricity price level and corresponding time periods for peak, flat and valley periods; input this information into the dispatch optimization model of step S4, and include it in the calculation together with the photovoltaic power generation forecast value, energy storage available capacity and load demand. According to the load type and its process requirements, identify which loads can start and stop or adjust the power to a certain extent (such as some non-continuous processes in industrial production, motor speed variable equipment, air conditioning and refrigeration systems, etc.); during the peak electricity price period, moderately reduce or suspend the power output of the adjustable load to reduce the power load during the high-price period; during the valley electricity price period, moderately increase the load operating power or perform energy storage charging to take advantage of the electricity fee to complete the "peak shaving and valley filling". When performing the above-mentioned time-of-use regulation operation, fully consider factors such as the safety of the energy storage system, the load process requirements, the user experience (the impact on life or production) and the expected economic benefits; ultimately achieve an effective reduction in electricity costs, ensure that the normal operation of key loads is not affected while shaving the peak and filling the valley, and match the high and low peaks of photovoltaic power generation to improve the overall economy of the system. The time-sharing control process of the dispatchable load and the economic benefit evaluation results are recorded in the system database; in the subsequent feedback correction in step S6 or the optimization calculation of the next dispatch cycle, the load control plan is further corrected or improved by referring to the data.

[0082] This implementation can make real-time corrections through fuzzy control or adaptive control algorithms under the coordinated execution of distributed control systems, and perform time-sharing regulation of dispatchable loads based on the peak and valley electricity prices of the power grid, effectively achieving the goals of peak shaving and valley filling and economic dispatch. At the same time, this method fully utilizes the output of renewable energy and the price difference of electricity on the basis of ensuring the safe operation of the solar energy storage system, further improving the reliability and economy of the overall operation.

[0083] In this embodiment, the step of evaluating the health status of the energy storage system is further included. When capacity attenuation or temperature abnormality is detected in the energy storage system, the safety protection strategy is triggered and the energy scheduling plan is derated or switched to protect the normal operation of the energy storage equipment.

[0084] The key parameters of energy storage battery cells and battery packs (including voltage, current, temperature, internal resistance, capacity decay rate, etc.) are monitored in real time through sensors; the key parameters are collected regularly and compared with factory reference values ​​or historical records to identify possible capacity decay, temperature anomalies or internal resistance abnormality trends. Health assessment methods based on statistical models or machine learning, such as Kalman filtering, support vector machines (SVM) or random forests, can be used to calculate the remaining service life (RUL) and health state (SoH) of the battery based on long-term monitoring data; when the capacity decay rate exceeds the threshold or the temperature is continuously in an unsafe range, an alarm is triggered and the next safety protection strategy is implemented.

[0085] When capacity attenuation or temperature abnormality is detected in the energy storage system, the system will automatically initiate a safety protection strategy:

[0086] Limit the maximum charge and discharge power or charge and discharge rate;

[0087] Adjust the target SoC range of energy storage to reduce the additional loss of batteries caused by deep charging and discharging;

[0088] If necessary, terminate the connection between the energy storage system and the load or photovoltaic panels to prevent the fault from expanding.

[0089] At the same time, the energy scheduling plan is derated or switched, for example:

[0090] If the energy storage system is no longer able to fully meet the peak load demand, it will prioritize critical loads and delay or reduce the operation of some dispatchable loads;

[0091] If the energy storage temperature is too high, it is necessary to temporarily reduce the charging and discharging power until the temperature returns to the normal range.

[0092] In this embodiment, the energy scheduling method further includes:

[0093] When abnormal power quality or grid failure is detected, it automatically switches to off-grid mode, giving priority to the power demand of key loads in off-grid mode;

[0094] When the power grid returns to normal, the system automatically switches back to the grid-connected mode and re-executes the optimization and scheduling of step S4 and step S5.

[0095] By installing power quality detection devices at the grid connection point, real-time monitoring of parameters such as voltage, current, frequency and harmonic content is carried out; once the grid voltage or frequency deviates from the allowable range, or a grid fault (such as a large-scale power outage) is detected, an abnormal signal is immediately sent to the central controller. When the grid has a serious abnormality or fault, the central controller triggers the off-grid protection logic:

[0096] Disconnect from the power grid to prevent grid failure from spreading to the local system;

[0097] Use energy storage systems and photovoltaic power generation to provide power support for local microgrids or important loads;

[0098] In off-grid mode, the system prioritizes the basic electricity needs of key loads (such as medical equipment, core equipment of production lines, etc.), and at the same time reduces or queues other loads based on the actual photovoltaic power generation capacity and remaining energy storage capacity.

[0099] When the grid is detected to be back to normal, the central controller gradually resumes grid connection according to the safe grid connection process:

[0100] Synchronize phase with grid voltage and frequency, and close the circuit breaker after confirming that grid parameters are stable;

[0101] Re-execute the energy scheduling optimization and execution process in step S4 and step S5, including updating the energy storage system charging and discharging strategy and load scheduling;

[0102] The operating data of this fault and off-grid / grid switching are recorded for reference in subsequent strategy evaluation and improvement.

[0103] In this embodiment, it also includes photoelectric energy storage data management based on blockchain or edge computing technology, which includes encrypting and storing photovoltaic power generation, energy storage charging and discharging status, and power load data in a trusted ledger to achieve a traceable and auditable energy scheduling process when multiple subjects collaborate.

[0104] Deploy blockchain nodes or edge computing devices in the system to encrypt and store photovoltaic power generation, energy storage charging and discharging status, and power load data and record them in distributed ledgers; blockchain nodes can be deployed in different operating entities (such as power grid companies, energy storage operators, and electricity customers) or key physical locations to achieve data sharing and trusted records under multi-party collaboration. Encrypt information such as photovoltaic power generation, remaining capacity of energy storage systems, charging and discharging behaviors, and load requirements and write them into the blockchain ledger; each node relies on consensus mechanisms (such as PoW, PoS, PBFT, etc.) to complete the verification and update of the distributed ledger to ensure that the data cannot be tampered with and is traceable.

[0105] The smart contract function of blockchain can realize automated interaction between multiple entities, such as:

[0106] When a charging or discharging dispatch is completed, the corresponding "transaction record" or "emission accounting" can be automatically generated and archived to the blockchain; when the energy storage operator or the demand-side response participant needs to reconcile or settle the energy transaction, the complete and reliable dispatch record can be obtained from the blockchain account book. If combined with the carbon emission trading mechanism, the blockchain account book can also publicly and auditably record the emission reduction contribution (or emission) of each dispatch behavior, which is convenient for subsequent carbon asset settlement or regulatory inspection.

[0107] For frequent and massive data interaction scenarios, edge computing nodes can be used to quickly process and lightweight store local data, reducing dependence on the central cloud or main chain; only key results or summary information are uploaded to the blockchain network, thereby taking into account both real-time and security, and avoiding transmission congestion or data overload.

[0108] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. An energy scheduling method for optical energy storage, characterized in that: The following steps are involved: S1. Obtain the power generation, component temperature, irradiance and ambient temperature data of the photovoltaic module; S2. Based on historical power consumption records and real-time load monitoring information, predict the power demand of each load device in the next cycle; S3. Detect the remaining capacity, instantaneous charging and / or discharging power and health status of the energy storage system; S4. Based on the photovoltaic power generation information, load demand forecast results and energy storage status, the optimization algorithm is used to calculate the charging and discharging strategy for the next cycle to obtain the optimal energy scheduling plan; S5. Allocate and control the charging / discharging power of the energy storage system according to the optimal energy scheduling scheme. If the surplus photovoltaic power generation is greater than the power required by the energy storage, the surplus power will be delivered to the load or the power grid; if the photovoltaic power generation cannot meet the load demand, the energy storage system and / or the power grid will make up for it; S6. In the process of implementing the scheduling plan, the energy scheduling strategy is dynamically modified according to the real-time data and prediction deviation, and the modified results are stored in the system database.

2. The energy scheduling method of optical energy storage according to claim 1, characterized in that: The load demand prediction in step S2 adopts a prediction algorithm based on time series analysis or a machine learning model, and the machine learning model includes one or more models such as a long short-term memory network LSTM, a support vector machine SVM or a random forest.

3. The energy scheduling method of optical energy storage according to claim 1 or 2, characterized in that: The optimization algorithm in step S4 includes one or more of linear programming, mixed integer programming, genetic algorithm or particle swarm algorithm, and solves the objective function of the energy scheduling scheme according to constraints such as electricity price, peak and valley electricity price, demand response signal, etc.

4. The energy scheduling method of optical energy storage according to claim 1, characterized in that: In the process of executing step S4, different scheduling schemes are further comprehensively evaluated in combination with the carbon emission factor, and energy scheduling strategies with lower carbon emissions are preferentially selected.

5. The energy scheduling method of optical energy storage according to claim 1, characterized in that: The scheduling execution in step S5 is completed by a distributed control system, which includes a plurality of sub-control units for respectively controlling the switching actions and power regulation of the photovoltaic grid-connected inverter, the energy storage converter and each load device.

6. The energy scheduling method of optical energy storage according to claim 1, characterized in that: The feedback and correction in step S6 include fuzzy control or adaptive control algorithms, which perform real-time corrections on the scheduling strategy based on the power generation characteristics of the photovoltaic modules and the changes in power consumption on the load side.

7. The energy scheduling method of optical energy storage according to claim 1, characterized in that: When calculating the energy allocation strategy, the dispatchable load is time-sharingly regulated according to the current peak-valley electricity price information of the power grid, including but not limited to starting, pausing or adjusting the motor speed and air conditioning cooling capacity, so as to achieve the economic dispatching goal of peak shaving and valley filling.

8. The energy scheduling method of optical energy storage according to claim 1, characterized in that: The method further includes the step of evaluating the health status of the energy storage system. When capacity attenuation or temperature abnormality is detected in the energy storage system, the safety protection strategy is triggered and the energy scheduling plan is derated or switched to protect the normal operation of the energy storage equipment.

9. The energy scheduling method of optical energy storage according to claim 1, characterized in that: The energy scheduling method further comprises: When abnormal power quality or grid failure is detected, it automatically switches to off-grid mode, giving priority to the power demand of key loads in off-grid mode; When the power grid returns to normal, the system automatically switches back to the grid-connected mode and re-executes the optimization and scheduling of step S4 and step S5.

10. The energy scheduling method of optical energy storage according to claim 1, characterized in that: It also includes photovoltaic energy storage data management based on blockchain or edge computing technology, which includes encrypting and storing photovoltaic power generation, energy storage charging and discharging status, and power load data in a trusted ledger to achieve a traceable and auditable energy scheduling process when multiple entities collaborate.