Energy storage system ai collaborative scheduling method and system based on multi-objective optimization

Abstract

The application relates to the technical field of AI cooperative scheduling, and provides an energy storage system AI cooperative scheduling method and system based on multi-objective optimization, which solves the problems of energy storage system AI cooperative scheduling efficiency, operation stability and intelligent level. The application collects the charging time period distribution, historical power consumption and power price of the user side; constructs a power consumption preference vector, predicts a future to-be-scheduled load curve; through real-time monitoring of the temperature difference between the battery and the cabin environment in the target energy storage system by a temperature sensor, when the temperature difference reaches a preset threshold, the speed of the heat dissipation fan and the opening degree of the inert gas charging and discharging valve are adjusted by fuzzy PID control, so that the temperature difference is controlled within the threshold, and a device health report is generated; the optimization objective function with multiple optimization directions is matched with the to-be-scheduled load curve, and an initial charging and discharging plan is generated; based on reinforcement learning, the device health report and the plan of the adjacent energy storage system, the plan is corrected, and the energy storage system AI cooperative scheduling efficiency, operation stability and intelligent level are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of AI collaborative scheduling, in particular to an AI collaborative scheduling method and system for energy storage systems based on multi-objective optimization. BACKGROUND

[0002] In the scenario of distributed energy storage networks (such as user-side distributed energy storage, micro-grid cluster energy storage, etc.), the energy storage system needs to meet multiple technical requirements: on the one hand, it needs to balance the electricity cost, grid load smoothing (reduce peak-valley load fluctuations), and equipment operation safety (such as battery and cabin temperature difference control to avoid accelerated aging); on the other hand, as the number of energy storage systems increases, it needs to achieve multi-system collaborative scheduling to avoid local load overload caused by the concentrated charging and discharging of adjacent systems in the same period; in addition, it also needs to have dynamic response capability, which can adjust the scheduling strategy according to real-time load changes and equipment health status to adapt to complex and variable operating environments.

[0003] At present, a typical scheme for the above requirements is an energy storage scheduling scheme based on a single-objective optimization algorithm, which specifically includes: predicting the future load curve through historical load data, using an improved particle swarm optimization algorithm to generate the charging and discharging plan of each energy storage system with the single objective of "lowest electricity cost"; at the same time, adjusting the battery cabin heat dissipation equipment through traditional PID control to maintain the basic operation safety of the equipment; in terms of multi-system collaboration, only a fixed threshold (such as the upper limit of the charging and discharging power of each system) is used to avoid extreme load conflicts.

[0004] This scheme has three significant defects: first, it focuses on a single cost target, which can easily lead to excessive high-frequency charging and discharging to reduce costs, ignoring equipment health issues such as excessive battery temperature difference and continuous operation time exceeding the limit, thereby shortening the service life of the equipment; second, multi-system collaboration relies only on fixed threshold constraints, lacks a dynamic interaction mechanism, and cannot adjust its own strategy according to the real-time charging and discharging plan of adjacent systems, which can easily cause load overlap that does not cause extreme but affects the stability of the power grid; third, the scheduling plan lacks dynamic correction capability based on real-time data (such as equipment health reports and sudden load fluctuations), which is not adaptive enough to face prediction deviations or equipment state mutations, making it difficult to meet the multi-objective balancing requirements in complex scenarios. SUMMARY

[0005] The present application aims to provide an AI collaborative scheduling method and system for energy storage systems based on multi-objective optimization to solve the problems of single-objective imbalance, rigid collaborative mechanism, and lack of dynamic correction in existing technologies, which affect the efficiency, operation stability, and intelligence level of AI collaborative scheduling for energy storage systems.

[0006] To solve the above technical problems, in a first aspect, the present application provides an AI collaborative scheduling method for energy storage systems based on multi-objective optimization, which includes:

[0007] Collecting user side data through the edge node, the user side data including: charging period distribution, historical power consumption and power price;

[0008] According to the user side data, constructing a power consumption preference vector, and predicting a to-be-scheduled load curve of a future time period based on the power consumption preference vector;

[0009] Real-time monitoring temperature difference data between a battery and a cabin environment in the target energy storage system through a temperature sensor, when the temperature difference data is equal to or greater than a preset temperature difference threshold, adjusting a rotation speed of a cabin heat dissipation fan and an opening degree of an inert gas charging and discharging valve through a fuzzy PID control mode, so as to control the adjusted temperature difference data within the preset temperature difference threshold range, and generating a device health report according to a preset frequency during the adjustment process;

[0010] Matching a preset optimization objective function including a plurality of optimization directions with the to-be-scheduled load curve, to generate an initial charging and discharging plan of the target energy storage system;

[0011] Based on reinforcement learning, the device health report and a charging and discharging plan of an adjacent energy storage system, correcting the initial charging and discharging plan to realize AI collaborative scheduling of the energy storage system.

[0012] Optionally, when the temperature difference data is equal to or greater than the preset temperature difference threshold, the rotation speed of the cabin heat dissipation fan and the opening degree of the inert gas charging and discharging valve are adjusted through the fuzzy PID control mode, so that the adjusted temperature difference data is controlled within the preset temperature difference threshold range, and the device health report is generated according to the preset frequency during the adjustment process, including:

[0013] Converting the temperature difference data into a corresponding electrical signal, comparing the electrical signal with a reference signal corresponding to the preset temperature difference threshold, and starting an adjustment program when the electrical signal reaches or exceeds the reference signal;

[0014] Extracting a temperature difference change rate from the temperature difference data, adjusting a control parameter through the fuzzy PID control mode based on the adjustment program and the temperature difference change rate, adjusting the rotation speed of the cabin heat dissipation fan and the opening degree of the inert gas charging and discharging valve based on the adjusted control parameter, and obtaining the adjusted temperature difference data;

[0015] Converting the adjusted temperature difference data into a corresponding feedback signal based on a preset conversion rule;

[0016] When the feedback signal indicates that the adjusted temperature difference data is equal to or greater than the preset temperature difference threshold, dynamically adjusting the adjusted temperature difference data based on the feedback signal, so that the adjusted temperature difference data is maintained within the preset temperature difference threshold range;

[0017] In the process of dynamically adjusting the temperature difference data, the real-time rotating speed of the cabin heat dissipation fan, the real-time opening degree of the inert gas charging and discharging valve, and the temperature difference data at each time are recorded, and the real-time rotating speed, the real-time opening degree, and the temperature difference data at each time are integrated according to a preset frequency to form a device health report.

[0018] Optionally, based on the adjustment program and the temperature difference change rate, the control parameter is adjusted by a fuzzy PID control mode, the rotating speed of the cabin heat dissipation fan and the opening degree of the inert gas charging and discharging valve are adjusted based on the adjusted control parameter, and adjusted temperature difference data is obtained, including:

[0019] Based on the adjustment program, the operating direction corresponding to the cabin heat dissipation fan and the inert gas charging and discharging valve respectively and the numerical range of the temperature difference change rate are determined, the operating direction is associated with the numerical range, and an adjustment scheme for the control parameter is formed;

[0020] According to the increasing or decreasing trend of the temperature difference change rate, the numerical range of the temperature difference change rate is divided into different adjustment intervals, and the variation amplitude of the control parameter is set for the different adjustment intervals respectively;

[0021] Based on the adjustment scheme and the variation amplitude, the numerical value of the control parameter is adjusted;

[0022] Based on the adjusted control parameter, the rotating speed of the cabin heat dissipation fan is adjusted to the corresponding gear position and the opening and closing degree of the inert gas charging and discharging valve, and the adjusted temperature difference data is obtained according to the adjusted rotating speed and the opening and closing degree.

[0023] Optionally, the preset optimization objective function including a plurality of optimization directions is matched with the to-be-scheduled load curve to generate an initial charging and discharging plan of the target energy storage system, including:

[0024] A plurality of optimization directions are set, the plurality of optimization directions include: electricity cost, load fluctuation amplitude and device use time, and expressions corresponding to the plurality of optimization directions are fused to form an optimization objective function;

[0025] The to-be-scheduled load curve is divided into a plurality of time periods according to a preset time interval, and the load value of each time period is determined;

[0026] Based on the optimization objective function, a plurality of initial charging and discharging schemes are generated by using a multi-objective particle swarm optimization algorithm, each initial charging and discharging scheme includes charging and discharging values of each time period;

[0027] The charging and discharging values are compared with the load values of the corresponding time periods to obtain deviation values, and the corresponding scheme with the deviation values within a preset deviation range is retained as an initial charging and discharging plan of the target energy storage system.

[0028] Optionally, the multi-objective particle swarm optimization algorithm is used to generate a plurality of initial charging and discharging schemes based on the optimization objective function, each of the initial charging and discharging schemes including charging and discharging values of each time period, including:

[0029] determining specific requirements of each of the optimization directions in each time period, the specific requirements including: a cost limit corresponding to each time period of the electricity consumption cost, a numerical value change limit corresponding to adjacent time periods of the load fluctuation amplitude, and a time limit corresponding to continuous charging and discharging of the equipment use time length;

[0030] based on the cost limit, the numerical value change limit and the time limit, setting a minimum value and a maximum value allowed for the charging and discharging value of each time period;

[0031] between the minimum value and the maximum value of each time period, based on the optimization objective function, selecting a plurality of different charging and discharging values according to the influence degree of different optimization directions;

[0032] arranging the selected charging and discharging values in time sequence to form an initial charging and discharging scheme, and repeating the selection step and the arrangement step to generate a plurality of initial charging and discharging schemes by using the multi-objective particle swarm optimization algorithm.

[0033] Optionally, the initial charging and discharging plan is corrected based on reinforcement learning, the equipment health report and the charging and discharging plan of the adjacent energy storage system to realize AI collaborative scheduling of the energy storage system, including:

[0034] extracting the equipment operation state limit in the equipment health report, the equipment operation state limit including the time length limit of continuous operation of the equipment and the numerical value limit of single charging and discharging;

[0035] comparing the charging and discharging plan of the adjacent energy storage system with the initial charging and discharging plan according to the same time granularity to identify the load overlapping part in the same time period;

[0036] based on the load overlapping part, combining the time length limit and the numerical value limit to determine the time period in the initial charging and discharging plan that needs to be corrected and the corresponding correction range;

[0037] adjusting the charging and discharging value of the time period that needs to be corrected in the correction range, so that the adjusted charging and discharging value matches the charging and discharging value of the corresponding time period of the adjacent energy storage system;

[0038] integrating the adjusted charging and discharging values of each time period in time sequence to realize AI collaborative scheduling of the energy storage system.

[0039] Optionally, the adjusting the charge-discharge value of the time period in need of correction within the correction range to match the charge-discharge value of the corresponding time period of the adjacent energy storage system comprises:

[0040] setting an allowed difference range between the adjusted charge-discharge value and the charge-discharge value of the corresponding time period of the adjacent energy storage system;

[0041] calculating an adjustment amplitude value of the charge-discharge value of the time period in need of correction according to the charge-discharge value of the corresponding time period of the adjacent energy storage system and the allowed difference range;

[0042] adjusting the charge-discharge value of the corresponding time period according to the adjustment amplitude value within the correction range, and checking whether the adjusted charge-discharge value meets the numerical limit in the equipment operation state limit;

[0043] if the adjusted charge-discharge value meets the numerical limit, determining that the adjusted charge-discharge value matches the charge-discharge value of the corresponding time period of the adjacent energy storage system.

[0044] In a second aspect, the present application provides an AI collaborative scheduling system for an energy storage system based on multi-objective optimization, comprising:

[0045] a collection module configured to collect user-side data through an edge node, the user-side data including: a charging time period distribution, a historical power consumption, and a power price;

[0046] a construction module configured to construct a power consumption preference vector according to the user-side data, and predict a to-be-scheduled load curve of a future time period based on the power consumption preference vector;

[0047] an adjustment module configured to monitor temperature difference data between a battery and a cabin environment in a target energy storage system in real time through a temperature sensor, and when the temperature difference data is equal to or greater than a preset temperature difference threshold, adjust a speed of a cabin cooling fan and an opening degree of an inert gas charge-discharge valve through a fuzzy PID control mode, so that the adjusted temperature difference data is controlled within a preset temperature difference threshold range, and generate an equipment health report at a preset frequency during the adjustment process;

[0048] a generation module configured to match a preset optimization objective function including a plurality of optimization directions with the to-be-scheduled load curve, and generate an initial charge-discharge plan of the target energy storage system;

[0049] a correction module configured to correct the initial charge-discharge plan based on reinforcement learning, the equipment health report, and a charge-discharge plan of an adjacent energy storage system, so as to realize AI collaborative scheduling of the energy storage system.

[0050] In a third aspect, the present application provides an electronic device, comprising:

[0051] a memory for storing a computer program;

[0052] a processor for implementing the steps of the multi-objective optimization based AI collaborative scheduling method of energy storage system according to the first aspect when executing the computer program.

[0053] In a fourth aspect, the present application provides a computer readable storage medium, wherein the computer readable storage medium stores a computer program, and the computer program can implement the steps of the multi-objective optimization based AI collaborative scheduling method of energy storage system according to the first aspect when executed by a processor.

[0054] In the present application, a multi-objective optimization based AI collaborative scheduling method of energy storage system is provided, which comprises: collecting user side data through an edge node, the user side data comprising: charging time period distribution, historical power consumption and power price; constructing a power consumption preference vector according to the user side data, and predicting a to-be-scheduled load curve of a future time period based on the power consumption preference vector; monitoring temperature difference data between a battery and a cabin environment in a target energy storage system in real time through a temperature sensor, and when the temperature difference data is equal to or greater than a preset temperature difference threshold, adjusting the speed of a cabin heat dissipation fan and the opening degree of an inert gas charge and discharge valve through a fuzzy PID control mode, so that the adjusted temperature difference data is controlled within the preset temperature difference threshold range, and generating a device health report at a preset frequency during the adjustment process; matching a preset optimization objective function comprising multiple optimization directions with the to-be-scheduled load curve, to generate an initial charge and discharge plan of the target energy storage system; and correcting the initial charge and discharge plan based on reinforcement learning, the device health report and the charge and discharge plan of an adjacent energy storage system, to realize AI collaborative scheduling of the energy storage system.

[0055] The present application has the following beneficial effects:

[0056] The multi-objective optimization based AI collaborative scheduling method of energy storage system provided by the present application can collect data such as charging time period distribution, historical power consumption and power price of the user side through an edge node, to provide basic information support for subsequent load prediction and scheduling plan generation; by constructing a power consumption preference vector and predicting a future to-be-scheduled load curve based thereon, the scheduling target and direction of the energy storage system can be determined; by monitoring the temperature difference in real time through a temperature sensor, and adjusting the speed of a heat dissipation fan and the opening degree of an inert gas charge and discharge valve through a fuzzy PID when the temperature difference exceeds the limit, and generating a device health report at a preset frequency, the operation safety of the battery device can be ensured and device state data can be provided; by matching an optimization objective function comprising multiple optimization directions with the to-be-scheduled load curve, to generate an initial charge and discharge plan, the charge and discharge strategy of the energy storage system can be initially planned; and by correcting the initial plan based on reinforcement learning, the device health report and the plan of an adjacent system, the dynamic collaborative scheduling of the energy storage system can be realized.

[0057] Further, the temperature difference data is converted into an electrical signal and compared with a reference signal to start the adjustment, the control parameters are adjusted based on the temperature difference change rate using fuzzy PID, the temperature difference is dynamically adjusted through feedback, the device state is recorded to generate a health report, and the association between the operation direction and the temperature difference change rate range during the adjustment of the control parameters is refined, the interval division and amplitude setting, the parameter adjustment, and the device adjustment process are refined. Through the refined signal processing, dynamic feedback mechanism and health report generation logic of the temperature difference adjustment, the accuracy and real-time performance of the temperature difference control are improved, and the device is ensured to operate within a safe temperature difference range; through the adjustment logic of the control parameters, the adaptability of the fuzzy PID control to complex temperature difference changes is enhanced, the device adjustment efficiency is improved, and reliable data support is provided for device health and subsequent scheduling correction.

[0058] These aspects or other aspects of the present application will be more apparent in the following description of the embodiments. BRIEF DESCRIPTION OF DRAWINGS

[0059] In order to more clearly illustrate the technical solutions of the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present application, and those skilled in the art can obtain other drawings according to these drawings without creative effort.

[0060] Figure 1 A flowchart of an AI collaborative scheduling method for a multi-objective optimization-based energy storage system according to an embodiment of the present application is shown in the figure.

[0061] Figure 2 A specific implementation diagram of an AI collaborative scheduling method for a multi-objective optimization-based energy storage system according to an embodiment of the present application is shown in the figure.

[0062] Figure 3 A specific implementation diagram of an AI collaborative scheduling method for a multi-objective optimization-based energy storage system according to an embodiment of the present application is shown in the figure.

[0063] Figure 4 A structure diagram of an AI collaborative scheduling system for a multi-objective optimization-based energy storage system according to an embodiment of the present application is shown in the figure. DETAILED DESCRIPTION

[0064] In order to solve the single target imbalance, the rigid coordination mechanism and the lack of dynamic correction in the prior art, the embodiment of the application provides a multi-target optimization-based AI collaborative scheduling method for a storage system, which adopts the following design concept: collecting the charging time distribution, past power consumption and electricity price and other data of users through edge nodes; predicting the power consumption habits of users according to these data, and further predicting the power consumption load in a future period of time; monitoring the temperature difference between the battery in the storage device and the surrounding environment in real time through a sensor, and when the temperature difference is too large, adjusting the heat dissipation device and the gas valve to control the temperature difference, and generating a device status report regularly; combining multiple targets (such as power consumption cost, device status, etc.) that need to be considered and the predicted power consumption load to formulate a preliminary charging and discharging plan; and adjusting the preliminary plan according to the device status report, the plans of other nearby storage systems and the experience accumulated through continuous learning, to realize the intelligent collaborative work of multiple storage systems.

[0065] In order to enable personnel in the technical field to better understand the application scheme, the application will be further described in detail below in combination with the drawings and specific embodiments. Obviously, the described embodiments are only part of the embodiments of the application, not all. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the application.

[0066] The core of the application is to provide a multi-target optimization-based AI collaborative scheduling method for a storage system, and a flowchart of a specific embodiment thereof is shown in Figure 1 The method comprises:

[0067] S11, collecting user side data through an edge node, the user side data comprising: charging time period distribution, historical power consumption and electricity price.

[0068] The edge node is an information collection device close to the user, which is used to directly obtain user power consumption related data; the user side data comprises charging time period distribution: user charging time interval record, historical power consumption: total amount of power consumption of the user in the past period of time, and electricity price: power charging standard in different time periods. These data form a user power consumption basic information set after being collected.

[0069] In the embodiment of the application, the edge node first establishes a connection with the user power consumption device (such as a charging pile and an electric meter), sets a fixed collection frequency (such as every 12 hours), and then collects the charging time period distribution (such as user charging from 17:30 to 19:00 every day), the historical power consumption (such as daily power consumption in the past week), and the electricity price (such as 0.6 yuan / degree during the day and 0.25 yuan / degree at night). For example, the edge node of a certain community aggregates the charging time of each household, the daily power consumption and the price of each period every 12 hours.

[0070] S12. Based on user-side data, construct an electricity preference vector and predict the load curve to be dispatched in future time periods based on the electricity preference vector.

[0071] Among them, the electricity preference vector is a set of information formed by integrating the distribution of charging time periods, historical electricity consumption, and electricity prices, which is used to reflect users' electricity consumption habits; the load curve to be dispatched is the trend of user electricity load changes in the future (such as 24 hours) predicted based on the electricity preference vector.

[0072] In this embodiment of the application, key features (such as charging peak, peak electricity consumption, and preference for low-priced electricity) are extracted from the data collected in S11 and integrated into an electricity preference vector. Combined with the load patterns of similar historical electricity consumption patterns, the load curve to be scheduled is predicted. For example, if a user prefers to charge between 18:00 and 20:00 and the electricity consumption is high at this time, it can be predicted that the load during this period will be higher than other periods in the future by combining historical data.

[0073] S13. Monitor the temperature difference between the battery and the cabin environment in the target energy storage system in real time through temperature sensors. When the temperature difference is equal to or greater than the preset temperature difference threshold, adjust the speed of the cabin cooling fan and the opening of the inert gas charging and discharging valve through fuzzy PID control so that the adjusted temperature difference is controlled within the preset temperature difference threshold range. During the adjustment process, generate an equipment health report according to the preset frequency.

[0074] Among them, the temperature difference data is the difference between the battery temperature and the ambient temperature of the cabin; the preset temperature difference threshold is the upper limit of the temperature difference to ensure battery safety (such as 5℃); fuzzy PID control is a method to flexibly adjust the equipment status; the cabin cooling fan is used to reduce the cabin temperature, and the inert gas charging and discharging valve regulates the cabin temperature by charging and discharging gas; the equipment health report is a periodic report that records the fan speed, valve opening, and temperature difference changes at a fixed frequency.

[0075] In the embodiments of this application, such as Figure 2 As shown, the temperature sensor monitors the temperature difference data in real time and compares it with the preset threshold. When the temperature difference is equal to or exceeds the threshold, the fan speed (the larger the temperature difference, the higher the speed) and valve opening are adjusted by fuzzy PID control (the opening is increased when the temperature difference exceeds the threshold by 1℃). During the adjustment, the fan speed, valve opening, and temperature difference data are recorded at a fixed frequency (e.g., every 30 minutes) and integrated to form an equipment health report. For example, when the temperature difference is 6℃ (threshold 5℃), the fan speed is adjusted from 800 rpm to 1200 rpm, and the valve opening is adjusted from 10% to 30%. Data is recorded every 30 minutes to generate a report.

[0076] S14. Match the preset optimization objective function containing multiple optimization directions with the load curve to be scheduled to generate the initial charging and discharging plan of the target energy storage system.

[0077] The optimization objective function is a calculation method of comprehensively optimizing multiple optimization directions (such as reducing power consumption, reducing load fluctuation, and prolonging equipment use time). The function can be a linear superposition formula or a nonlinear formula, and the specific expression of the function is not limited in the embodiment.

[0078] In the embodiment of the application, a preset optimization objective function with multiple optimization directions is included. As shown in the formula, the function is matched with the to-be-scheduled load curve of S12, and multiple charging and discharging schemes are generated in combination with the load peak / valley period. Figure 3 As shown in the formula, the function is matched with the to-be-scheduled load curve of S12, and multiple charging and discharging schemes are generated in combination with the load peak / valley period. The schemes that meet all the optimization directions are selected as the initial plan. For example, the load curve shows that 18:00-20:00 is a peak, the optimization objective function requires less discharging at this time and preferentially charging at the low-price night period, and the initial plan arranges charging at 22:00-6:00 and discharging at 18:00-20:00.

[0079] S15, based on reinforcement learning, a device health report, and a charging and discharging plan of a neighboring energy storage system, the initial charging and discharging plan is corrected to realize AI collaborative scheduling of the energy storage system.

[0080] The reinforcement learning is a method of optimizing behavior by accumulating experience and is used to optimize the charging and discharging plan. The charging and discharging plan of the neighboring energy storage system is the charging and discharging arrangement of other systems in a similar location. The initial plan is corrected according to the device state and the neighboring plan to realize collaborative operation.

[0081] In the embodiment of the application, the device health report (such as single discharging for no more than 3 hours) of S13 and the neighboring system plan (such as discharging of the neighboring system at 18:00-20:00) are obtained. Based on the historical experience of reinforcement learning, the conflict between the initial plan and the device limit and the neighboring plan (such as discharging for 4 hours at 18:00-20:00 in the initial plan) is analyzed. The plan is corrected to avoid the conflict, for example, the discharging period is adjusted to 18:00-21:00, and is divided into two segments, each for 2 hours, which meets the device limit and avoids the neighboring peak.

[0082] The application provides the following specific examples: in the A cell energy storage system, the edge node is installed in the power distribution room, and 50 households of data are collected every 12 hours: user B charges every day from 17:30 to 19:00, the average daily electricity consumption in the past week is 15 degrees, the electricity price is 0.6 yuan / degree from 8:00 to 21:00 on weekdays, and the electricity price is 0.25 yuan / degree during the rest of the time. Based on these data, the constructed electricity consumption preference vector shows that user B is in the charging peak (8 degrees of electricity consumption) from 17:30 to 19:00, and prefers low-price electricity at night, and accordingly, it is predicted that the load in this period in the next week will be high and stable at night. At the same time, the temperature sensor of the battery cabin body monitors that the battery temperature is 25℃ and the cabin environment is 19℃ (temperature difference 6℃, threshold 5℃), and the fuzzy PID control is started: the fan speed is adjusted from 800 rpm to 1200 rpm, and the valve opening is adjusted from 10% to 30%, and the data is recorded every 30 minutes (for example, the first record temperature difference is 5.5℃, and the second record temperature difference is 4.8℃), and the equipment health report is generated (showing that the longest single discharge is 3 hours). Combined with the predicted load curve (the highest load from 18:00 to 20:00), the initial plan is generated according to the optimization target (lowest cost, change of charging and discharging ≤2 degrees per hour, single charging ≤5 hours): charging from 22:00 to 5:00 the next day (divided into two sections of 22:00-3:00 and 3:00-5:00), and discharging 3 degrees per hour from 18:00 to 20:00. Finally, referring to the equipment health report and the adjacent B cell plan (discharge from 18:00 to 20:00), the plan is corrected based on the reinforcement learning experience: the discharge of the A cell from 18:00 to 20:00 is adjusted to two sections of 18:00-20:00 and 20:00-21:00 (each 2 hours), avoiding conflict with the B cell and meeting the equipment limit.

[0083] By performing S11-S15, the application embodiment collects data through the edge node to provide a basis for subsequent analysis; constructs the electricity consumption preference vector and predicts the load curve to clearly define the future electricity demand; monitors the temperature difference in real time and adjusts the equipment to ensure the safety of the battery and generate a health report; generates an initial plan based on an optimization target function to balance the cost, load and safety; corrects the plan by combining reinforcement learning, equipment status and adjacent plans to realize multi-system collaboration. The whole process forms a closed loop from data collection to dynamic scheduling, improves the adaptability, safety and collaborative efficiency of the energy storage system, and meets the actual electricity consumption demand of users.

[0084] In a possible embodiment, S13, when the temperature difference data is equal to or greater than the preset temperature difference threshold, the speed of the cabin body heat dissipation fan and the opening of the inert gas charging and discharging valve are adjusted by the fuzzy PID control mode, so that the adjusted temperature difference data is controlled within the preset temperature difference threshold range, and in the adjustment process, an equipment health report is generated at a preset frequency, including:

[0085] Step 131, convert the temperature difference data into corresponding electrical signals, compare the electrical signals with the reference signal corresponding to the preset temperature difference threshold, and start the adjustment program when the electrical signals reach or exceed the reference signal.

[0086] Wherein, the temperature difference data is the difference between the battery temperature and the cabin environment temperature in the target energy storage system, the electrical signal is the electronic signal converted from the temperature difference data for information transmission between devices, the reference signal corresponding to the preset temperature difference threshold is the electrical signal standard matched with the upper limit of the safety temperature difference, and the adjustment program is a set of instructions for starting the device adjustment action. Through the processing of these elements, the signal for starting the adjustment is finally generated.

[0087] In the embodiments of the present application, first, the signal conversion device converts the temperature difference data into corresponding electrical signals, such as converting the temperature difference of 6℃ into an electrical signal of 4.8V according to the proportion of 1℃ corresponding to 0.8V, and second, comparing the converted electrical signal with the reference signal corresponding to the preset temperature difference threshold, assuming that the reference signal corresponding to the preset temperature difference threshold of 5℃ is 4V, when the converted electrical signal reaches or exceeds 4V, it means that the temperature difference exceeds the safety range, at this time, the adjustment program is started, for example, the converted electrical signal is 4.8V which exceeds 4V, the system immediately starts the adjustment program.

[0088] Step 132, extract the temperature difference change rate from the temperature difference data, adjust the control parameters based on the adjustment program and the temperature difference change rate through the fuzzy PID control method, and adjust the speed of the cabin heat dissipation fan and the opening of the inert gas charge and discharge valve based on the adjusted control parameters to obtain the adjusted temperature difference data.

[0089] Wherein, the temperature difference change rate is the speed of the temperature difference changing with time, such as the degree of temperature difference change per minute, the fuzzy PID control method is a method that can flexibly adjust the device parameters according to the temperature difference change, the control parameter is the specific value of adjusting the fan speed and the valve opening, and the adjusted temperature difference data is the new temperature difference value obtained after the device adjustment. These processes finally generate the adjusted temperature difference data.

[0090] In the embodiments of the present application, first, the temperature difference change rate is calculated from the continuously monitored temperature difference data, such as the temperature difference rising from 6℃ to 8℃ in the past 2 minutes, the change rate is calculated as (8-6)℃ ÷ 2 minutes = 1℃ / minute, second, based on the adjustment program started in step 131 and the calculated temperature difference change rate, the control parameters are adjusted through the fuzzy PID control method, for example, according to the change rate of 1℃ / minute, the control parameter of the fan speed is adjusted from 1000 revolutions / minute to 1300 revolutions / minute, and the control parameter of the valve opening is adjusted from 20% to 30%, third, based on the adjusted control parameters, the device is adjusted to obtain the adjusted temperature difference data, such as the adjusted temperature difference from 8℃ to 7℃.

[0091] Step 133: Based on the preset conversion rules, convert the adjusted temperature difference data into the corresponding feedback signal.

[0092] The adjusted temperature difference data is the new temperature difference value obtained after the equipment is adjusted in step 132. The preset conversion rule is a fixed method for converting the temperature difference data into a feedback signal, such as the corresponding ratio of temperature difference to voltage. The feedback signal is an electrical signal that reflects the adjusted temperature difference state and is used to transmit it to the control system, ultimately generating a feedback signal that reflects the adjusted temperature difference state.

[0093] In this embodiment of the application, the adjusted temperature difference data in step 132 is first obtained, for example, the adjusted temperature difference is 6℃. Then, the temperature difference data is converted into the corresponding feedback signal according to the preset conversion rule. Assuming that the preset conversion rule is 1℃ corresponds to 0.7V, the calculation process of converting a temperature difference of 6℃ into a feedback signal is 6×0.7V=4.2V. For example, if the adjusted temperature difference is 5℃, according to the rule that 1℃ corresponds to 0.6V, the feedback signal is calculated to be 5×0.6V=3V.

[0094] Step 134: When the feedback signal indicates that the adjusted temperature difference data is equal to or greater than the preset temperature difference threshold, the adjusted temperature difference data is dynamically adjusted based on the feedback signal so that the adjusted temperature difference data is maintained within the preset temperature difference threshold range.

[0095] The feedback signal is the electrical signal generated in step 133 that reflects the temperature difference state after adjustment. Dynamic adjustment is the process of continuously optimizing the equipment's operating state based on the feedback signal. Maintaining the adjusted temperature difference data within the preset temperature difference threshold range means that the temperature difference is kept below the safe upper limit through continuous adjustment, ultimately ensuring that the temperature difference is within a safe range.

[0096] In this embodiment, the feedback signal generated in step 133 is first received, and it is determined whether the adjusted temperature difference data it represents is equal to or greater than the preset temperature difference threshold. For example, if the feedback signal is 4.5V, the corresponding temperature difference is 5.6℃, while the preset threshold is 5℃, indicating that the temperature difference still exceeds the standard. Secondly, based on the feedback signal, dynamic adjustment is performed to further adjust the fan speed and valve opening. For example, the fan speed is adjusted from 1300 rpm to 1500 rpm, and the valve opening is adjusted from 30% to 40%. The adjustment process is repeated until the temperature difference data represented by the feedback signal drops to within the preset threshold range. For example, after adjustment, the temperature difference is 4.8℃, which meets the safety requirements.

[0097] Step 135: During the process of dynamically adjusting the temperature difference data, record the real-time speed of the cabin cooling fan, the real-time opening degree of the inert gas charging and discharging valve, and the temperature difference data at each moment. Integrate the real-time speed, real-time opening degree, and temperature difference data at each moment according to the preset frequency to form an equipment health report.

[0098] Wherein, the real-time speed of the cabin heat dissipation fan is the instantaneous rotating speed of the fan in the adjustment process, the real-time opening of the inert gas charging and discharging valve is the instantaneous opening and closing ratio of the valve in the adjustment process, the temperature difference data at each time is the temperature difference value recorded at different time points, the preset frequency is the fixed time interval for recording and integrating data, and the equipment health report is a periodic report reflecting the equipment running state formed after integrating these data.

[0099] In the embodiment of the present application, first, in the dynamic adjustment process of step 134, the speed of the fan (such as 1200 rpm, 1400 rpm), the opening of the valve (such as 25%, 35%), and the temperature difference data at different times (such as 5.3℃, 4.9℃) are recorded in real time, and second, these real-time data are integrated according to a preset frequency (such as every 30 minutes), such as summarizing the highest speed, average opening, temperature difference trend information within 30 minutes, forming an equipment health report, for example, integrating data once every 30 minutes, the report will show "10:00-10:30, the highest speed of the fan is 1400 rpm, the average opening of the valve is 30%, and the temperature difference decreases from 5.3℃ to 4.9℃".

[0100] The present application provides the following specific examples: In the A cell energy storage system, the temperature sensor monitors that the temperature difference between the battery and the cabin is 7℃, according to the conversion rule that 1℃ corresponds to 0.9V, the calculated electric signal is 7x0.9V=6.3V, the preset temperature difference threshold is 5℃ corresponding to the reference signal 4.5V, after comparison, 6.3V exceeds 4.5V, the system starts the adjustment program. Then, continuous monitoring finds that the temperature difference rises from 5℃ to 9℃ within 3 minutes, the temperature difference rate is calculated as (9-5)℃ ÷ 3 minutes ≈ 1.33℃ / minute, based on the adjustment program and the rate, the fan speed is adjusted from 900 rpm to 1400 rpm and the valve opening is adjusted from 15% to 35% through fuzzy PID control, and the temperature difference decreases to 7℃ after adjustment. Then, according to the preset conversion rule that 1℃ corresponds to 0.8V, 7℃ is converted into a feedback signal, and the calculation process is 7x0.8V=5.6V. Since the feedback signal corresponding to the temperature difference of 7℃ is still higher than the preset threshold of 5℃, the system adjusts the fan speed to 1500 rpm and the valve opening to 40%, and the temperature difference decreases to 4.8℃ after adjustment, which is converted into a feedback signal of 4.8x0.8V=3.84V, meeting the safety requirements. During the whole dynamic adjustment process, the system records data every 30 minutes, during 9:00-9:30, the fan speed is 1100 rpm, 1300 rpm, and 1500 rpm in turn, the valve opening is 20%, 35%, and 40% in turn, and the temperature difference data is 5.5℃, 5.2℃, and 4.8℃ in turn, and finally these data are integrated to form an equipment health report.

[0101] By performing steps 131-135, the embodiment of the application can start the adjustment program in time by converting the temperature difference data into an electrical signal and comparing it with the reference signal, provide a trigger basis for subsequent adjustment; calculate the temperature difference change rate and adjust the control parameter, flexibly adjust the equipment according to the temperature difference change trend, and improve the adjustment accuracy; convert the adjusted temperature difference into a feedback signal, which can provide a clear basis for judging whether further adjustment is needed; dynamic adjustment based on the feedback signal can ensure that the temperature difference is finally controlled within a safe range, avoiding equipment risks; recording and integrating data to generate a health report can comprehensively reflect the equipment state and adjustment effect, and provide a reliable basis for subsequent charge and discharge plan correction. The whole process works together to achieve effective control of the temperature difference, ensuring the safe operation of the equipment, and providing data support for system scheduling.

[0102] In one possible embodiment, step 132 adjusts the control parameter by fuzzy PID control method based on the adjustment program and the temperature difference change rate, adjusts the speed of the cabin heat dissipation fan and the opening of the inert gas charge and discharge valve based on the adjusted control parameter, and obtains the adjusted temperature difference data, including:

[0103] a1, based on the adjustment program, determining the operation direction corresponding to the cabin heat dissipation fan and the inert gas charge and discharge valve respectively, and the numerical range of the temperature difference change rate, associating the operation direction with the numerical range to form an adjustment scheme for the control parameter.

[0104] Wherein, the adjustment program is a set of instructions for guiding the adjustment of the equipment, the operation direction is the adjustment direction of the speed of the cabin heat dissipation fan and the opening of the inert gas charge and discharge valve (such as increasing or decreasing), the numerical range of the temperature difference change rate is the interval of the change speed of the temperature difference with time (such as 0.5-1℃ / min), and the adjustment scheme is the control parameter adjustment plan formed by associating the operation direction with the numerical range, which is used to determine the adjustment mode under different temperature differences, and finally form a scheme for guiding parameter adjustment.

[0105] In the embodiment of the application, first, the operation direction of the fan and the valve is determined based on the adjustment program (such as increasing the speed and increasing the opening when the temperature difference increases), and the numerical range of the temperature difference change rate is determined (such as 0-0.5℃ / min, 0.5-1℃ / min), second, the operation direction is associated with these ranges to form an adjustment scheme, for example, the 0.5-1℃ / min interval corresponds to an increase of 10% in fan speed and an increase of 15% in valve opening, for example, the adjustment program stipulates that the heat dissipation needs to be enhanced when the temperature difference rises, then the scheme in this interval will clearly indicate that the speed is increased and the opening is increased.

[0106] a2, according to the increasing or decreasing trend of the temperature difference change rate, the numerical range of the temperature difference change rate is divided into different adjustment intervals, and the variation amplitude of the control parameter is set for different adjustment intervals.

[0107] The increasing trend of the temperature difference change rate is the trend of the temperature difference increasing over time, the decreasing trend is the trend of the temperature difference decreasing, the adjustment interval is a subinterval of the numerical range (such as 0.5-0.7℃ / min), and the change range is the specific adjustment amount of the control parameter in each interval (such as increasing by 10%), which is used to quantify the adjustment strength, and finally determines the adjustment strength corresponding to each interval.

[0108] In the embodiment of the present application, first, the numerical range in a1 is divided into adjustment intervals according to the increasing or decreasing trend of the temperature difference change rate, for example, 0.5-1℃ / min is divided into 0.5-0.7℃ / min and 0.7-1℃ / min in the increasing trend, and second, the change range is set for each interval, for example, the former corresponds to an increase of 10% in speed and an increase of 15% in opening, and the latter corresponds to an increase of 15% in speed and an increase of 20% in opening, for example, the adjustment strength of the 0.5-0.7℃ / min interval is less than that of the 0.7-1℃ / min interval.

[0109] a3, adjusting the numerical value of the control parameter based on the adjustment scheme and the change range.

[0110] The adjustment scheme is the association plan of the operation direction and the range in a1, the change range is the adjustment amount of each interval in a2, the numerical value of the control parameter is the specific value of the fan speed and the valve opening (such as 1000 revolutions / minute, 20%), the adjusted control parameter is the new value obtained by combining the scheme and the range, which is used to guide the equipment adjustment, and finally the parameter directly used for adjustment is obtained.

[0111] In the embodiment of the present application, first, the interval in which the current temperature difference change rate is located and the corresponding adjustment scheme and change range are determined, and second, the control parameter is adjusted according to the change range, for example, the original speed is 1200 revolutions / minute and the opening is 20%, in the 0.5-0.7℃ / min interval (change range 10%, 15%), the adjusted speed is 1200+1200x10%=1320 revolutions / minute, and the opening is 20%+20%x15%=23%, for example, the parameter adjustment is ensured to meet the interval requirements through calculation.

[0112] a4, based on the adjusted control parameter, adjusting the speed of the cabin heat dissipation fan to the corresponding gear and the opening and closing degree of the inert gas charge and discharge valve, and obtaining the adjusted temperature difference data according to the adjusted speed and the opening and closing degree.

[0113] The adjusted control parameter is the new value obtained by a3, the corresponding gear is the fixed gear of the fan speed (such as 1320 revolutions / minute corresponding to 3 gears), the opening and closing degree is the opening ratio of the valve (such as 23%), and the adjusted temperature difference data is the new temperature difference measured after the equipment runs, which is used to reflect the adjustment effect, and finally the data reflecting the adjustment effect is obtained.

[0114] In the embodiments of the present application, first, the fan is adjusted to the corresponding gear and the valve is adjusted to the corresponding opening degree according to the adjusted control parameters, for example, 1320 rpm (3 gears) and 23% opening degree, and second, the temperature difference is re-measured after the equipment runs for a period of time to obtain the adjusted temperature difference data, for example, the original temperature difference is 6.5°C, and after running for 15 minutes, it decreases to 5.3°C, for example, the adjustment effect is verified by actual measurement.

[0115] The present application provides the following specific examples: In the A cell energy storage system, the adjustment program stipulates that the heat dissipation needs to be enhanced when the temperature difference rises, and accordingly, the a1 step determines that the operation direction of the fan and the valve is to increase the rotating speed and increase the opening degree, and the temperature difference change rate numerical range is divided into 0-0.5°C / min, 0.5-1°C / min and 1°C / min or more, and is associated to form an adjustment scheme, wherein 0.5-1°C / min corresponds to a 10% increase in rotating speed and a 15% increase in opening degree. In the a2 step, the range shows an increasing trend and is divided into 0.5-0.7°C / min and 0.7-1°C / min intervals, the former is set to a 10% (rotating speed) and 15% (opening degree) variation range, and the latter is set to a 15% and 20% variation range. When the current temperature difference change rate is 0.6°C / min, it is in the 0.5-0.7°C / min interval, in the a3 step, the original control parameters are a rotating speed of 1200 rpm and an opening degree of 20%, and the calculation shows that the adjusted rotating speed is 1200+1200×10%=1320 rpm and the opening degree is 20%+20%×15%=23%. In the a4 step, the fan is adjusted to 1320 rpm (3 gears) and the valve is adjusted to 23%, and after running for 15 minutes, the battery temperature is 26°C, the cabin environment temperature is 20.7°C, and the calculated temperature difference is 26-20.7=5.3°C, which is the adjusted temperature difference data.

[0116] By executing a1-a4, the adjustment scheme formed by the a1 step of the embodiments of the present application clearly determines the adjustment direction under different temperature differences, providing a basis for subsequent operation; the interval division and variation range setting of the a2 step make the adjustment intensity match the temperature difference change trend, improving the pertinence; the a3 step adjusts the parameters based on the scheme and the range, ensuring the accuracy and reasonableness of the adjustment amount; and the a4 step reflects the adjustment effect intuitively through equipment adjustment and temperature difference measurement. The whole process works together to make the temperature difference adjustment more accurate and meet the actual needs, providing protection for the safe operation of the equipment and reliable temperature difference data support for subsequent system scheduling.

[0117] In a possible embodiment, S14, a preset optimization objective function containing multiple optimization directions is matched with the to-be-scheduled load curve to generate an initial charge and discharge plan of the target energy storage system, including:

[0118] Step 141, set multiple optimization directions, the multiple optimization directions include: electricity cost, load fluctuation amplitude and equipment use time, and expressions corresponding to the multiple optimization directions are fused to form an optimization objective function.

[0119] wherein the optimization direction is a target to be considered when formulating the charging and discharging plan, including the electricity cost: the cost of power use, the load fluctuation amplitude: the change size of the electricity load in different time periods, and the equipment use time length: the continuous working time of the equipment, each direction has a corresponding quantitative expression, and the optimization objective function is a comprehensive evaluation standard formed by fusing the expressions, used for judging the pros and cons of the charging and discharging scheme, and finally obtaining a function that can comprehensively evaluate the scheme.

[0120] In the embodiments of the present application, first, the three optimization directions of the electricity cost, the load fluctuation amplitude and the equipment use time length are set, expressions are established for each direction, for example, the electricity cost expression is "the sum of the electricity quantity in each time period multiplied by the corresponding electricity price", the load fluctuation amplitude expression is "the sum of the difference values of the load values of adjacent time periods", and the equipment use time length expression is "the cumulative value of the continuous charging and discharging time"; then the expressions are fused according to certain weights to form the optimization objective function, for example, the three expressions are multiplied by the weights of 0.4, 0.3 and 0.3 respectively and then added together, if the electricity cost of a certain scheme is 50 yuan, the load fluctuation amplitude is 10, and the equipment use time length is 3 hours, the calculated value of the optimization objective function is 50*0.4+10*0.3+3*0.3=20+3+0.9=23.9.

[0121] Step 142, dividing the to-be-scheduled load curve into multiple time periods according to a preset time interval, and determining the load value of each time period.

[0122] wherein the to-be-scheduled load curve is the predicted future electricity load change trend, the preset time interval is the fixed time length for dividing the time periods, the multiple time periods are the time periods obtained by dividing the curve according to the interval, and the load value of each time period is the predicted electricity load quantity in the time period, and finally the load data corresponding to each time period is obtained.

[0123] In the embodiments of the present application, first, the preset time interval is determined, and the to-be-scheduled load curve is divided into multiple time periods according to the interval; then the load value of each time period is read from the curve, for example, the load value of the 0:00-1:00 time period is 30, the load value of the 1:00-2:00 time period is 25, and the values of all time periods are determined in turn.

[0124] Step 143, generating multiple initial charging and discharging schemes by using a multi-objective particle swarm optimization algorithm based on the optimization objective function, and each initial charging and discharging scheme contains the charging and discharging values of each time period.

[0125] wherein the optimization objective function is the comprehensive evaluation standard formed in step 141, the multi-objective particle swarm optimization algorithm is a method for finding the optimal solution of multiple objectives, the initial charging and discharging scheme is a possible charging and discharging arrangement generated based on the optimization objective function and the load curve, each scheme contains the charging and discharging values (charging or discharging quantity) of each time period, and finally multiple possible charging and discharging schemes are obtained.

[0126] In the embodiment of the present application, first, the optimization objective function of step 141 is taken as the evaluation basis, and the multi-objective particle swarm optimization algorithm is used to generate multiple initial charging and discharging schemes in combination with the load values of each period in step 142; in the generation process, the algorithm constantly adjusts the charging and discharging values of each period to optimize the objective function, for example, a scheme of charging 20 from 0:00 to 1:00, discharging 5 from 1:00 to 2:00, and charging 15 from 2:00 to 3:00, to form a scheme containing the values of all periods.

[0127] Step 144, compare the charging and discharging values with the load values of the corresponding period to obtain the deviation value, and retain the corresponding scheme with the deviation value within the preset deviation range as the initial charging and discharging plan of the target energy storage system.

[0128] Wherein, the charging and discharging value is the charging and discharging amount of each period in the initial scheme, the load value is the load of each period in step 142, the deviation value is the difference between the two, the preset deviation range is the allowed deviation interval, the initial charging and discharging plan is the scheme with the deviation value within the range, and finally the plan meeting the load demand is obtained.

[0129] In the embodiment of the present application, first, the deviation value of the charging and discharging value of each period in each initial scheme and the corresponding load value is calculated, for example, the charging and discharging value of a scheme from 0:00 to 1:00 is 25, the corresponding load is 30, and the deviation value is 25-30=-5; then the deviation value is compared with the preset range (such as-10 to 10), and the scheme with the deviation value of all periods within the range is retained as the initial charging and discharging plan.

[0130] The application provides the following specific examples: in the A cell energy storage system, step 141 sets the optimization direction as the electricity cost, the load fluctuation amplitude and the equipment use time length, the electricity cost expression is "the sum of the electricity price of each period charging and discharging amount", the load fluctuation amplitude is "the sum of the absolute value of the adjacent period difference", the equipment use time length is "the total sum of the continuous charging and discharging time", and the weights of the three are 0.5, 0.3 and 0.2. If the electricity cost of a scheme is 60 yuan, the load fluctuation amplitude is 12, and the equipment use time length is 4 hours, the optimization objective function is 60*0.5+12*0.3+4*0.2=30+3.6+0.8=34.4. Step 142 divides the future 24-hour to-be-scheduled load curve into 24 time periods according to 1-hour intervals, and determines the load values of 0:00-1:00, 1:00-2:00, 2:00-3:00 and the like. Step 143 generates five initial schemes based on the optimization objective function and the load values by using a multi-objective particle swarm optimization algorithm, wherein the charging and discharging values of one of the schemes are as follows: 0:00-1:00 charging 28, 1:00-2:00 charging 23, 2:00-3:00 charging 18 and the like. Step 144 takes -10 to 10 as the preset deviation range, calculates the deviation values of each time period of the scheme: 0:00-1:00 28-30=-2, 1:00-2:00 23-25=-2, 2:00-3:00 18-20=-2, all in the range, and therefore the scheme is selected as the initial charging and discharging plan.

[0131] By performing steps 141-144, the embodiment of the application realizes the comprehensive balance of the electricity cost, the load stability and the equipment use by setting multiple optimization directions and fusing them into an optimization objective function; the load curve is divided into time periods and the load values are determined, thereby providing specific time and load basis for the charging and discharging scheme; the optimization algorithm is used to generate multiple initial schemes, thereby providing diversified choices; the deviation screening is used to ensure that the plan matches the actual load demand. The synergistic effect of the steps makes the generated initial charging and discharging plan more reasonable and feasible, and can effectively support the scheduling demand of the energy storage system.

[0132] In a possible embodiment, step 143 generates multiple initial charging and discharging schemes by using a multi-objective particle swarm optimization algorithm based on the optimization objective function, and each initial charging and discharging scheme contains the charging and discharging values of each time period, including:

[0133] b1, determining the specific requirements of each optimization direction in each time period, and the specific requirements include: the cost limit corresponding to the electricity cost of each time period, the numerical value change limit corresponding to the adjacent time period of the load fluctuation amplitude and the time limit corresponding to the continuous charging and discharging of the equipment use time length.

[0134] The optimization direction includes electricity cost, load fluctuation amplitude and equipment use time length, and specific requirements are constraint conditions of each optimization direction in each time period. The cost limit is the maximum amount of electricity cost in each time period, the numerical value change limit is the maximum allowed difference value of charging and discharging in adjacent time periods, and the time limit is the longest time of continuous charging and discharging of the equipment. These requirements jointly regulate the range and change of the charging and discharging value, and finally form the constraint standard of each time period.

[0135] In the embodiments of the present application, first, for electricity cost, the cost limit is determined in combination with the electricity price of each time period, the cost limit is low when the peak period electricity price is high, and the cost limit is high when the valley period electricity price is low; second, for the load fluctuation amplitude, the numerical value change limit is determined according to the grid stability demand, such as a difference of not more than 5; third, for the equipment use time length, the continuous charging and discharging time limit is determined according to the equipment parameters, such as not more than 4 hours. For example, the A area peak period 9:00-18:00 electricity price is 0.6 yuan / unit, the cost limit is 20 yuan / hour, the adjacent time period difference limit is 5, and the continuous charging and discharging limit is 4 hours.

[0136] b2, based on the cost limit, the numerical value change limit and the time limit, set the minimum value and the maximum value allowed for the charging and discharging value of each time period.

[0137] The cost limit, the numerical value change limit and the time limit are the constraint conditions determined by b1, the minimum value is the minimum charging and discharging amount allowed in each time period, the maximum value is the maximum charging and discharging amount allowed in each time period, and the two constitute the selectable range of the charging and discharging value, which is used to regulate the value selection boundary.

[0138] In the embodiments of the present application, first, the maximum value is calculated based on the cost limit, such as a cost limit of 15 yuan in a certain time period and an electricity price of 0.5 yuan / unit, the maximum value is 15÷0.5=30; second, in combination with the numerical value change limit, the maximum value of the current time period is adjusted by referring to the maximum value of the previous time period, such as a previous maximum value of 25 and a change limit of 4, the current maximum value is not more than 29; third, according to the time limit, if the equipment is close to the upper limit of continuous operation, set a smaller minimum and maximum value. For example, the minimum is 10 and the maximum is 30 after comprehensive calculation in a certain time period.

[0139] b3, between the minimum value and the maximum value of each time period, based on the optimization objective function, select a plurality of different charging and discharging values according to the influence degree of different optimization directions.

[0140] The minimum value and the maximum value are the time period value range determined by b2, the optimization objective function is a standard for comprehensive evaluation of the scheme, the influence degree is the importance of each optimization direction, and the charging and discharging value is the actual selected charging and discharging amount of each time period. Finally, a plurality of values meeting the constraint and optimization requirements are generated.

[0141] In the embodiments of the present application, firstly, the influence degree of each optimization direction is determined, such as the electricity cost is the most important (0.5), the load fluctuation is the second (0.3), and the equipment time length is lower (0.2); secondly, in the numerical range, a plurality of values are selected in combination with the influence degree, a larger charging value is selected in a low price period, a value close to the value is selected in adjacent periods to meet the fluctuation limit, and the limit is adjusted. For example, the previous period is 20, the change limit is 5, the current value is selected in the range of 15-25, and the value is 18, 20, and 22.

[0142] b4, the selected charging and discharging values are arranged in time sequence to form an initial charging and discharging scheme, and the multi-objective particle swarm optimization algorithm is used to repeat the selection and arrangement steps to generate a plurality of initial charging and discharging schemes.

[0143] Among them, the selected charging and discharging values are the specific values of the time period determined by b3, the initial charging and discharging scheme is the value combination arranged by time, the multi-objective particle swarm optimization algorithm generates a plurality of schemes by repeating selection and arrangement, and finally obtains a plurality of charging and discharging schemes meeting the constraints.

[0144] In the embodiments of the present application, firstly, the values selected by b3 are arranged by time to form a scheme, such as 0:00-1:00 selected 20, 1:00-2:00 selected 15; secondly, the multi-objective particle swarm optimization algorithm is used to repeat the selection and arrangement to generate a plurality of schemes, such as the second time selected 25, 18, the third time selected 18, 12, each scheme meets the constraints. For example, 5 different initial schemes are generated.

[0145] The application provides the following specific examples: In a cell energy storage system, the b1 step determines: in terms of electricity cost, 7:00-10:00 and 17:00-20:00 are peak periods (electricity price 0.5 yuan / unit), the cost limit is 15 yuan / period, the cost limit is 25 yuan in the remaining period (electricity price 0.3 yuan / unit); the numerical change limit of load fluctuation is 4; the continuous charging and discharging limit of equipment is 3 hours. In the b2 step, the 7:00-8:00 period is calculated according to the cost limit of 15 yuan, the maximum value is 15÷0.5=30, the maximum value of the previous period (6:00-7:00) is 25, and the current maximum adjustment is 29 due to the change limit of 4, and the minimum is 5; 17:00-18:00, similarly, the maximum is 29, and the minimum is 5. In the b3 step, 7:00-8:00 is in the range of 5-29, combined with the influence degree of high electricity cost, 25, 27 and 29 (the difference with the previous period 25 is less than or equal to 4) are selected; 17:00-18:00, 24, 26 and 29 are selected. In the b4 step, the values are arranged according to time to form the first scheme: 7:00-8:0025, 8:00-9:0027, 9:00-10:0029, 17:00-18:0024…; repeat the selection and arrangement to generate five initial schemes including "7:00-8:0027, 17:00-18:0026" and the like.

[0146] By performing b1-b4, the embodiments of the application determine the boundary by b1, convert the constraint into a numerical range by b2, select the value that meets the optimization requirement in the range by b3, and generate diversified schemes by b4. The cost, power grid stability and equipment safety are comprehensively considered, so that the generated initial charging and discharging scheme not only meets various constraints, but also has diversity, thereby providing a feasible basis for reasonable scheduling of the energy storage system.

[0147] In a possible embodiment, S15, based on reinforcement learning, device health report and charging and discharging plan of adjacent energy storage systems, the initial charging and discharging plan is corrected to realize AI collaborative scheduling of the energy storage system, including:

[0148] Step 151, extracting the device running state limit in the device health report, the device running state limit including the length limit of continuous operation of the device and the numerical limit of single charging and discharging.

[0149] Among them, the device health report is a report recording the running state of the energy storage device, and the device running state limit is a constraint condition for safe operation of the device, including the length limit of continuous operation of the device (the longest time for uninterrupted work at a time) and the numerical limit of single charging and discharging (the maximum amount of charging or discharging at a time). After these limits are extracted from the report, they are used as safety constraint standards for regulating the adjustment of the charging and discharging plan.

[0150] In the embodiments of the present application, first, the device health report is obtained, from which the constraint information related to the device running state is screened out, second, the time limit of continuous operation of the device is identified, for example, the record of "the continuous charging and discharging of the device cannot exceed 4 hours" in the report, and finally the numerical limit of single charging and discharging is extracted, for example, the content of "the single charging and discharging capacity cannot exceed 50 units", for example, from the device health report of the energy storage system in area A, the continuous operation time limit is extracted as 3 hours, and the single charging and discharging numerical limit is extracted as 40 units.

[0151] Step 152, compare the charging and discharging plan of the adjacent energy storage system with the initial charging and discharging plan according to the same time granularity to identify the load overlapping part in the same period.

[0152] Wherein, the adjacent energy storage system is other energy storage system close to the location of the target energy storage system, the charging and discharging plan is the arrangement of charging or discharging in each period, the same time granularity is to divide the time periods of the two plans into the same length (such as 1 hour), and the load overlapping part is the part in which the charging and discharging capacity of the two systems is high in the same period. By comparing and identifying these parts, the scheduling conflict period that needs to be coordinated can be determined.

[0153] In the embodiments of the present application, first, the same time granularity is determined, such as dividing the charging and discharging plans of the two systems into each hour, second, the charging and discharging plan of the adjacent energy storage system is split according to the time granularity, and then the charging and discharging capacity of each period is compared one by one to identify the part in which the charging and discharging capacity of the two systems exceeds a certain standard in the same period, that is, the load overlapping part, for example, the A community energy storage system discharges 30 units from 8:00 to 9:00, and the adjacent B community system discharges 25 units in the same period, which is the load overlapping part.

[0154] Step 153, based on the load overlapping part, combined with the time limit and the numerical limit, determine the period in the initial charging and discharging plan that needs to be corrected and the corresponding correction range.

[0155] Wherein, the load overlapping part is the part in which the charging and discharging capacity is high in the same period identified in step 152, the time limit is the longest time of continuous operation of the device, the numerical limit is the maximum amount of single charging and discharging, the period that needs to be corrected is the period in the initial plan that has load overlapping with the adjacent system and may exceed the device limit, and the correction range is the interval in which the charging and discharging value in the period can be adjusted. These contents jointly constitute the specific object and boundary of plan correction.

[0156] In the embodiments of the present application, firstly, the time periods corresponding to the load overlapping parts are analyzed to determine whether the initial charging and discharging plans of these time periods may exceed the device operating state limit, and secondly, in combination with the time limit, if a certain overlapping time period belongs to the fourth hour of continuous operation of the device (exceeding the 3-hour limit), the time period needs to be corrected, and in combination with the numerical limit, if the charging and discharging capacity of the time period is 45 units (exceeding the 40-unit limit), it is determined that the correction range is an interval of not more than 40 units and close to the numerical value of the adjacent system, for example, the 8:00-9:00 time period of the load overlap, the initial plan has reached 3 hours of continuous operation and the charging and discharging capacity is 45 units, and the correction range is set to 20-40 units.

[0157] Step 154, adjusting the charging and discharging numerical value of the time period that needs to be corrected in the correction range, so that the adjusted charging and discharging numerical value matches the charging and discharging numerical value of the adjacent energy storage system corresponding to the time period.

[0158] wherein the correction range is the interval of the charging and discharging numerical value that can be adjusted in step 153, the charging and discharging numerical value is the amount of charging or discharging in the time period, and the adjusted numerical value matching the numerical value of the adjacent system corresponding to the time period means that the two numerical values are close to avoid high load in the same time period, and this matching can be achieved by adjusting in the correction range to form the coordinated charging and discharging numerical value.

[0159] In the embodiments of the present application, firstly, the correction range of the time period that needs to be corrected is determined, such as 20-40 units, and secondly, the charging and discharging numerical value of the adjacent energy storage system corresponding to the time period is referred to, such as 30 units, and then a charging and discharging numerical value close to the numerical value is selected in the correction range, for example, 28 units are selected from 20-40 units, to ensure that the adjusted numerical value is in the correction range and matches the numerical value of the adjacent system, for example, the 8:00-9:00 time period that needs to be corrected, the correction range is 20-40 units, the numerical value of the adjacent system is 25 units, and the adjustment is 26 units.

[0160] Step 155, integrating the adjusted charging and discharging numerical values of each time period in time sequence to realize AI collaborative scheduling of the energy storage system.

[0161] wherein the adjusted charging and discharging numerical values of each time period are the charging and discharging capacities of each time period corrected in step 154, and the integration in time sequence is to arrange these numerical values in time sequence to form a complete plan, and the AI collaborative scheduling is that multiple energy storage systems coordinate work through the adjusted plan to realize stable overall operation, and finally a complete charging and discharging plan that can realize collaborative operation is formed.

[0162] In the embodiments of the present application, firstly, the adjusted charge and discharge values of each time period that needs to be corrected in step 154 are collected, secondly, the original charge and discharge values of the uncorrected time periods and the adjusted values are arranged in chronological order to form a complete charge and discharge plan, and finally, the target energy storage system and the adjacent system are coordinated to run through the plan, for example, the adjusted 8:00-9:00 time period of 29 units and the adjusted 18:00-19:00 time period of 31 units of the A cell are arranged in chronological order with the values of other time periods to form a coordinated scheduling plan.

[0163] The present application provides the following specific examples: In the A cell energy storage system, step 151 extracts the device operation state limit from the device health report: continuous operation time is 3 hours, and single charge and discharge value is 40 units. Step 152 compares the initial charge and discharge plan of the A cell with the plan of the adjacent B cell at a granularity of 1 hour, and identifies that 8:00-9:00 (A discharges 42 units, and B discharges 28 units) and 18:00-19:00 (A charges 38 units, and B charges 32 units) are load overlapping parts. Step 153 analyzes and finds that the 8:00-9:00 time period has been continuously operated for 3 hours (reaching the time limit) and the charge and discharge amount is 42 units (exceeding the 40 unit value limit), and the 18:00-19:00 time period will be continuously operated for 4 hours (exceeding the 3 hour time limit), so it is determined that these two time periods are time periods that need to be corrected, and the correction range is 20-40 units. Step 154 adjusts in the range: 8:00-9:00 time period of A from 42 units to 29 units (close to B's 28 units), and 18:00-19:00 time period from 38 units to 31 units (close to B's 32 units). Step 155 arranges the adjusted values and the values of other time periods (such as 7:00-8:00 30 units, 9:00-10:00 25 units, etc.) in chronological order to form a complete coordinated scheduling plan.

[0164] By performing steps 151-155, the embodiments of the present application define a safe boundary for plan adjustment by extracting device operation limits, ensure the safety of the device; accurately locate the conflict time period with the adjacent system by identifying the load overlapping part; determine the correction time period and range in combination with the device limit, so that the adjustment has a clear direction and boundary; adjust the value in the correction range and match with the adjacent system to effectively solve the load conflict; finally integrate to form a coordinated plan to realize the coordinated operation of multiple energy storage systems. These steps work together to improve the stability and overall efficiency of the energy storage system operation, and ensure the safety of the device and the balance of the power grid load.

[0165] In a possible embodiment, step 154 adjusts the charge and discharge values of the time periods that need to be corrected in the correction range, so that the adjusted charge and discharge values match the charge and discharge values of the corresponding time periods of the adjacent energy storage system, comprising:

[0166] c1, set the allowed difference range between the adjusted charge and discharge value and the charge and discharge value of the adjacent energy storage system in the corresponding period.

[0167] The adjusted charge and discharge value is the corrected charge or discharge amount of each period of the target energy storage system, the charge and discharge value of the adjacent energy storage system in the corresponding period is the charge or discharge amount of the same period of the other system in the same period, and the allowed difference range is the acceptable maximum difference of the charge and discharge value of the two systems in the same period, which is used to avoid conflicts caused by too large difference, and finally forms the standard of value matching between systems.

[0168] In the embodiments of the present application, first, according to the power grid load bearing capacity and system coordination demand, the allowed difference range between the adjusted charge and discharge value and the value of the adjacent system in the corresponding period is set, for example, considering the system stability, the range is set to ±3 units, that is, the value of the target system can be 3 units higher or lower than that of the adjacent system, for example, the allowed difference range of the A area energy storage system is set to ±4 units.

[0169] c2, according to the charge and discharge value of the adjacent energy storage system in the corresponding period and the allowed difference range, calculate the adjustment amplitude value of the charge and discharge value of the period to be corrected.

[0170] The charge and discharge value of the adjacent energy storage system in the corresponding period is the charge or discharge amount of the same period of the other system, the allowed difference range is the acceptable maximum difference set in step c1, and the adjustment amplitude value is the difference between the current value of the target system and the target value to be adjusted, which is used to determine the specific amount of adjustment, and finally obtain the quantitative adjustment basis.

[0171] In the embodiments of the present application, first, the charge and discharge value of the adjacent system in the corresponding period (such as 28 units) is obtained, and the allowed difference range (such as ±3 units) of step c1 is combined to determine the reasonable interval (such as 25-31 units) of the adjusted value of the target system, and then the current value (such as 42 units) of the target system is subtracted from the target value (such as 28 units) of the interval to calculate the adjustment amplitude value (42-28=14 units), that is, 14 units need to be reduced to enter the allowed range, for example, the current value is 35 units, and the adjustment amplitude value is 3 units.

[0172] c3, adjust the charge and discharge value of the corresponding period in the correction range according to the adjustment amplitude value, and check whether the adjusted charge and discharge value meets the value limit in the equipment operation state limit.

[0173] Wherein, the correction range is a preset charge-discharge value adjustable interval, the adjustment amplitude value is the adjustment difference calculated in step c2, the adjusted charge-discharge value is the specific charge or discharge amount after adjustment, and the value limit in the equipment operation state limit is the maximum amount of single charge-discharge of the equipment, which is used to ensure that the adjusted value does not harm the equipment, and finally an adjusted value meeting the constraint is obtained.

[0174] In the embodiments of the present application, first, the charge-discharge value of the corresponding period is adjusted in the correction range (such as 20-40 units) according to the adjustment amplitude value (such as 14 units) of step c2, for example, 42 units is reduced by 14 units to obtain 28 units, and then it is checked whether the adjusted value (28 units) meets the equipment value limit (such as 40 units). Since 28 units is less than 40 units, it meets the requirement, for example, the adjusted value 35 units meets the requirement of the value limit 40 units.

[0175] c4, if the adjusted charge-discharge value meets the value limit, determining that the adjusted charge-discharge value matches the charge-discharge value of the corresponding period of the adjacent energy storage system.

[0176] Wherein, the adjusted charge-discharge value is the specific charge or discharge amount obtained in step c3, the value limit is the maximum amount of single charge-discharge of the equipment, the charge-discharge value of the corresponding period of the adjacent energy storage system is the amount of the same period of other systems, and the matching means that the adjusted value is within the allowed difference range and meets the value limit, which is used to confirm the coordination and safety between systems, and finally the coordination state between systems is confirmed.

[0177] In the embodiments of the present application, first, it is judged whether the adjusted value (such as 28 units) of step c3 meets the equipment value limit (such as 40 units). Since 28 units is less than 40 units, it meets the requirement, and then it is checked whether the value is within the allowed difference range (such as 28 units of the adjacent system, ±3 units) of step c1. 28 units is within the interval of 25-31 units, which meets the range, and finally it is determined that the adjusted value matches the value of the corresponding period of the adjacent system, for example, the adjusted value 31 units, the adjacent system 28 units, the difference 3 units, which meets the range and the value limit, and it is determined that it matches.

[0178] The application provides the following specific examples: the coordination process of a cell energy storage system and an adjacent B cell energy storage system is as follows: step c1 sets the allowed difference range as ±3 units; in step c2, the charge and discharge value of the B cell during the 8:00-9:00 period is 28 units, the current value of the A cell during the period is 42 units, the calculation adjustment amplitude value is 42-28=14 units (14 units need to be reduced); in step c3, the A cell correction range during the period is 20-40 units, and 28 units are obtained after adjustment according to the amplitude value, and it is found that 28 units are less than the equipment value limit 40 units, which meets the requirements; in step c4, the difference between 28 units and 28 units of the B cell is 0, which is within the ±3 unit range and meets the value limit, and it is determined that the two match.

[0179] By performing c1-c4, the application embodiment clearly defines the standard for matching the values between systems by setting the allowed difference range, provides a quantitative adjustment basis by calculating the adjustment amplitude value, adjusts within the correction range and checks the value limit to ensure the rationality of the adjustment and the safety of the equipment, and finally confirms the matching to realize the coordinated operation of the target system and the adjacent system. These steps work together to effectively reduce the load conflict between systems and ensure the safety of the equipment and the stability of the power grid.

[0180] Figure 4 A structure diagram of an energy storage system AI collaborative scheduling system based on multi-objective optimization provided by the application embodiment is shown in Figure 4 The system comprises:

[0181] The acquisition module 41 is configured to acquire user side data through an edge node, and the user side data comprises a charging period distribution, historical power consumption and power prices.

[0182] The construction module 42 is configured to construct a power consumption preference vector according to the user side data, and predict a to-be-scheduled load curve of a future time period based on the power consumption preference vector.

[0183] The adjustment module 43 is configured to monitor temperature difference data between a battery and a cabin environment in a target energy storage system in real time through a temperature sensor, and when the temperature difference data is equal to or greater than a preset temperature difference threshold, adjust the speed of a cabin heat dissipation fan and the opening degree of an inert gas charge and discharge valve through a fuzzy PID control mode, so that the adjusted temperature difference data is controlled within the preset temperature difference threshold range, and a device health report is generated at a preset frequency during the adjustment process.

[0184] The generation module 44 is configured to match a preset optimization objective function comprising a plurality of optimization directions with the to-be-scheduled load curve, and generate an initial charge and discharge plan of the target energy storage system.

[0185] The correction module 45 is configured to correct the initial charging and discharging plan based on reinforcement learning, a device health report, and a charging and discharging plan of a neighboring energy storage system, so as to realize AI collaborative scheduling of the energy storage system.

[0186] The AI collaborative scheduling system for the energy storage system based on multi-objective optimization according to the embodiments of the present application is used to realize the AI collaborative scheduling method for the energy storage system based on multi-objective optimization described above, and therefore the specific embodiments of the AI collaborative scheduling system for the energy storage system based on multi-objective optimization can refer to the embodiments of the AI collaborative scheduling method for the energy storage system based on multi-objective optimization described above, and the specific embodiments can refer to the descriptions of the respective embodiments, which will not be described herein again.

[0187] The present application also provides an electronic device, comprising: a memory configured to store a computer program; and a processor configured to execute the computer program to implement the steps of the AI collaborative scheduling method for the energy storage system based on multi-objective optimization described above.

[0188] The present application also provides a computer readable storage medium, wherein the computer readable storage medium stores a computer program, and the computer program is executed by a processor to implement the steps of the AI collaborative scheduling method for the energy storage system based on multi-objective optimization described above.

[0189] In an exemplary embodiment, the computer readable storage medium described above can include, but is not limited to, a U disk, a read-only memory, a random access memory, a mobile hard disk, a magnetic disk or an optical disk, and various media that can store computer programs.

[0190] The embodiments of the present application also provide a computer program product, wherein the computer program product comprises a computer program, and the computer program is executed by a processor to implement the steps of the AI collaborative scheduling method for the energy storage system based on multi-objective optimization described above.

[0191] Those skilled in the art can further realize that the units and algorithm steps of the examples described in conjunction with the embodiments disclosed herein can be realized in electronic hardware, computer software or a combination of both. In order to clearly illustrate the interchangeability of hardware and software, the components and steps of the examples have been described in a general manner in the above description. Whether the functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0192] The above describes in detail a multi-objective optimization-based energy storage system AI collaborative scheduling method and system, an electronic device, and a storage medium provided by the present application. In this paper, specific examples are used to explain the principles and implementation modes of the present application. The above description of the embodiments is only used to help understand the method of the present application and its core idea. It should be pointed out that for ordinary skilled persons in the technical field, some improvements and modifications can be made to the present application without departing from the principles of the present application, and these improvements and modifications also fall within the protection scope of the present application.

Claims

1. A multi-objective optimization-based AI collaborative scheduling method for energy storage systems, characterized in that, The method comprises the following steps: Collecting user side data through an edge node, the user side data comprising: charging period distribution, historical power consumption and power price; According to the user side data, constructing a power consumption preference vector, and predicting a to-be-scheduled load curve of a future time period based on the power consumption preference vector; Real-time monitoring of temperature difference data between a battery and a cabin environment in a target energy storage system through a temperature sensor, and when the temperature difference data is equal to or greater than a preset temperature difference threshold, adjusting the rotation speed of a cabin heat dissipation fan and the opening degree of an inert gas charging and discharging valve through a fuzzy PID control method, so that the adjusted temperature difference data is controlled within the preset temperature difference threshold range, and generating a device health report at a preset frequency during the adjustment process; Matching a preset optimization objective function comprising multiple optimization directions with the to-be-scheduled load curve to generate an initial charging and discharging plan of the target energy storage system; Based on reinforcement learning, the device health report and the charging and discharging plan of the adjacent energy storage system, correcting the initial charging and discharging plan to realize AI collaborative scheduling of the energy storage system; The matching of the preset optimization objective function comprising multiple optimization directions with the to-be-scheduled load curve to generate the initial charging and discharging plan of the target energy storage system comprises: Setting multiple optimization directions, the multiple optimization directions comprising: power consumption, load fluctuation amplitude and device usage time, and fusing expressions corresponding to the multiple optimization directions to form an optimization objective function; Dividing the to-be-scheduled load curve into multiple periods according to a preset time interval, and determining the load value of each period; Based on the optimization objective function, generating multiple initial charging and discharging schemes using a multi-objective particle swarm optimization algorithm, each initial charging and discharging scheme comprising charging and discharging values of each period; Comparing the charging and discharging values with the load values of the corresponding periods to obtain deviation values, and retaining the corresponding scheme with the deviation values within a preset deviation range as the initial charging and discharging plan of the target energy storage system; The correction of the initial charging and discharging plan based on reinforcement learning, the device health report and the charging and discharging plan of the adjacent energy storage system to realize AI collaborative scheduling of the energy storage system comprises: Extracting device operating state limits in the device health report, the device operating state limits comprising a device continuous operating time limit and a single charging and discharging value limit; Comparing the charging and discharging plan of the adjacent energy storage system with the initial charging and discharging plan according to the same time granularity to identify the load overlapping part in the same period; Based on the load overlapping part, combining the time limit and the value limit to determine the period in the initial charging and discharging plan that needs to be corrected and the corresponding correction range; Adjusting the charging and discharging values of the period that needs to be corrected within the correction range, so that the adjusted charging and discharging values match the charging and discharging values of the corresponding period of the adjacent energy storage system; Integrating the adjusted charging and discharging values of each period in chronological order to realize AI collaborative scheduling of the energy storage system.

2. The multi-objective optimization-based energy storage system AI collaborative scheduling method according to claim 1, characterized in that, The temperature difference data is converted into a corresponding electrical signal, and the electrical signal is compared with a reference signal corresponding to the preset temperature difference threshold value. When the electrical signal reaches or exceeds the reference signal, a regulation program is started. The temperature difference data is converted into a corresponding electrical signal, and the electrical signal is compared with a reference signal corresponding to the preset temperature difference threshold value. When the electrical signal reaches or exceeds the reference signal, a regulation program is started. The temperature difference data is converted into a corresponding electrical signal, and the electrical signal is compared with a reference signal corresponding to the preset temperature difference threshold value. When the electrical signal reaches or exceeds the reference signal, a regulation program is started. The temperature difference data is converted into a corresponding electrical signal, and the electrical signal is compared with a reference signal corresponding to the preset temperature difference threshold value. When the electrical signal reaches or exceeds the reference signal, a regulation program is started. During the dynamic adjustment of the temperature difference data, the real-time speed of the cabin heat dissipation fan, the real-time opening degree of the inert gas charging and discharging valve, and the temperature difference data at each time are recorded. The real-time speed, the real-time opening degree, and the temperature difference data at each time are integrated according to a preset frequency to form a device health report. The temperature difference data is converted into a corresponding electrical signal, and the electrical signal is compared with a reference signal corresponding to the preset temperature difference threshold value. When the electrical signal reaches or exceeds the reference signal, a regulation program is started.

3. The multi-objective optimization based energy storage system AI co-scheduling method according to claim 2, characterized in that, The temperature difference data is converted into a corresponding electrical signal, and the electrical signal is compared with a reference signal corresponding to the preset temperature difference threshold value. When the electrical signal reaches or exceeds the reference signal, a regulation program is started. The temperature difference data is converted into a corresponding electrical signal, and the electrical signal is compared with a reference signal corresponding to the preset temperature difference threshold value. When the electrical signal reaches or exceeds the reference signal, a regulation program is started. The temperature difference data is converted into a corresponding electrical signal, and the electrical signal is compared with a reference signal corresponding to the preset temperature difference threshold value. When the electrical signal reaches or exceeds the reference signal, a regulation program is started. The temperature difference data is converted into a corresponding electrical signal, and the electrical signal is compared with a reference signal corresponding to the preset temperature difference threshold value. When the electrical signal reaches or exceeds the reference signal, a regulation program is started. The temperature difference data is converted into a corresponding electrical signal, and the electrical signal is compared with a reference signal corresponding to the preset temperature difference threshold value. When the electrical signal reaches or exceeds the reference signal, a regulation program is started.

4. The multi-objective optimization based energy storage system AI co-scheduling method according to claim 1, wherein, The temperature difference data is converted into a corresponding electrical signal, and the electrical signal is compared with a reference signal corresponding to the preset temperature difference threshold value. When the electrical signal reaches or exceeds the reference signal, a regulation program is started. The temperature difference data is converted into a corresponding electrical signal, and the electrical signal is compared with a reference signal corresponding to the preset temperature difference threshold value. When the electrical signal reaches or exceeds the reference signal, a regulation program is started. The temperature difference data is converted into a corresponding electrical signal, and the electrical signal is compared with a reference signal corresponding to the preset temperature difference threshold value. When the electrical signal reaches or exceeds the reference signal, a regulation program is started. The temperature difference data is converted into a corresponding electrical signal, and the electrical signal is compared with a reference signal corresponding to the preset temperature difference threshold value. When the electrical signal reaches or exceeds the reference signal, a regulation program is started. The temperature difference data is converted into a corresponding electrical signal, and the electrical signal is compared with a reference signal corresponding to the preset temperature difference threshold value. When the electrical signal reaches or exceeds the reference signal, a regulation program is started. The temperature difference data is converted into a corresponding electrical signal, and the electrical signal is compared with a reference signal corresponding to the preset temperature difference threshold value. When the electrical signal reaches or exceeds the reference signal, a regulation program is started. The temperature difference data is converted into a corresponding electrical signal, and the electrical signal is compared with a reference signal corresponding to the preset temperature difference threshold value. When the electrical signal reaches or exceeds the reference signal, a regulation program is started. The temperature difference data is converted into a corresponding electrical signal, and the electrical signal is compared with a reference signal corresponding to the preset temperature difference threshold value. When the electrical signal reaches or exceeds the reference signal, a regulation program is started. The temperature difference data is converted into a corresponding electrical signal, and the electrical signal is compared with a reference signal corresponding to the preset temperature difference threshold value. When the electrical signal reaches or exceeds the reference signal, a regulation program is started. The temperature difference data is converted into a corresponding electrical signal, and the electrical signal is compared with a reference signal corresponding to the preset temperature difference threshold value. When the electrical signal reaches or exceeds the reference signal, a regulation program is started. The temperature difference data is converted into a corresponding electrical signal, and the electrical signal is compared with a reference signal corresponding to the preset temperature difference threshold value. When the electrical signal reaches or exceeds the reference signal, a regulation program is started. The temperature difference data is converted into a corresponding electrical signal, and the electrical signal is compared with a reference signal corresponding to the preset temperature difference threshold value. When the electrical signal reaches or exceeds the reference signal, a regulation program is started. The temperature difference data is converted into a corresponding electrical signal, and the electrical signal is compared with a reference signal corresponding to the preset temperature difference threshold value. When the electrical signal reaches or exceeds the reference signal, a regulation program is started. The temperature difference data is converted into a corresponding electrical signal, and the electrical signal is compared with a reference signal corresponding to the preset temperature difference threshold value. When the electrical signal reaches or exceeds the reference signal, a regulation program is started. The temperature difference data is converted into a corresponding electrical signal, and the electrical signal is compared with a reference signal corresponding to the preset temperature difference threshold value. When the electrical signal reaches or exceeds the reference signal, a regulation program is started. The temperature difference data is converted into a corresponding electrical signal, and the electrical signal is compared with a reference signal corresponding to the preset temperature difference threshold value. When the electrical signal reaches or exceeds the reference signal, a regulation program is started. The temperature difference data is converted into a corresponding electrical signal, and the electrical signal is compared with a reference signal corresponding to the preset temperature difference threshold value. When the electrical signal reaches or exceeds the reference signal, a regulation program is started. The temperature difference data is converted into a corresponding electrical signal, and the electrical signal is compared with a reference signal corresponding to the preset temperature difference threshold value. When the electrical signal reaches or exceeds the reference signal, a regulation program is started. The temperature difference data is converted into a corresponding electrical signal, and the electrical signal is compared with a reference signal corresponding to the preset temperature difference threshold value. When the electrical signal reaches or exceeds the reference signal, a regulation program is started. The temperature difference data is converted into a corresponding electrical signal, and the electrical signal is compared with a reference signal corresponding to the preset temperature difference threshold value. When the electrical signal reaches or exceeds the reference signal, a regulation program is started. The temperature difference data is converted into a corresponding electrical signal, and the electrical signal is compared with a reference signal corresponding to the preset temperature difference threshold value. When the electrical signal reaches or exceeds the reference signal, a regulation program is started. The temperature difference data is converted into a corresponding electrical signal, and the electrical signal is compared with a reference signal corresponding to the preset temperature difference threshold value. When the electrical signal reaches or exceeds the reference signal, a regulation program is started. The temperature difference data is converted into a corresponding electrical signal, and the electrical signal is compared with a reference signal corresponding to the preset temperature difference threshold value. When the electrical signal reaches or exceeds the reference signal, a regulation program is started. The temperature difference data is converted into a corresponding electrical signal, and the electrical signal is compared with a reference signal corresponding to the preset temperature difference threshold value. When the electrical signal reaches or exceeds the reference signal, a regulation program is started. The temperature difference data is converted into a corresponding electrical signal, and the electrical signal is compared with a reference signal corresponding to the preset temperature difference threshold value. When the electrical signal reaches or exceeds the reference signal, a regulation program is started. The temperature difference data is converted into a corresponding electrical signal, and the electrical signal is compared with a reference signal corresponding to the preset temperature difference threshold value. When the electrical signal reaches or exceeds the reference signal, a regulation program is started. The temperature difference data is converted into a corresponding electrical signal, and the electrical signal is compared with a reference signal corresponding to the preset temperature difference threshold value. When the electrical signal reaches or exceeds the reference signal, a regulation program is started. The temperature difference data is converted into a corresponding electrical signal, and the electrical signal is compared with a reference signal corresponding to the preset temperature difference threshold value. When the electrical signal reaches or exceeds the reference signal, a regulation program is started. The temperature difference data is converted into a corresponding electrical signal, and the electrical signal is compared with a reference signal corresponding to the preset temperature difference threshold value. When the electrical signal reaches or exceeds the reference signal, a regulation program is started. The temperature difference data is converted into a corresponding electrical signal, and the electrical signal is compared with a reference signal corresponding to the preset temperature difference threshold value. When the electrical signal reaches or exceeds the reference signal, a regulation program is started. The temperature difference data is converted into a corresponding electrical signal, and the electrical signal is compared with a reference signal corresponding to the preset temperature difference threshold value. When the electrical signal reaches or exceeds the reference signal, a regulation program is started setting a minimum value and a maximum value of an allowed value for each period of charge and discharge value based on the cost limit, the numerical value change limit and the time limit; selecting a plurality of different charge and discharge values between the minimum value and the maximum value of each period based on the optimization objective function according to the influence degree of different optimization directions; arranging the selected charge and discharge values in time sequence to form an initial charge and discharge scheme, and repeatedly selecting and arranging to generate a plurality of initial charge and discharge schemes by using a multi-objective particle swarm optimization algorithm.

5. The multi-objective optimization based energy storage system AI co-scheduling method according to claim 1, wherein, The adjusting the charge and discharge value of the period to be corrected within the correction range to match the adjusted charge and discharge value with the charge and discharge value of the corresponding period of the adjacent energy storage system comprises: setting an allowed difference value range between the adjusted charge and discharge value and the charge and discharge value of the corresponding period of the adjacent energy storage system; calculating an adjustment amplitude value of the charge and discharge value of the period to be corrected according to the charge and discharge value of the corresponding period of the adjacent energy storage system and the allowed difference value range; adjusting the charge and discharge value of the corresponding period according to the adjustment amplitude value within the correction range, and checking whether the adjusted charge and discharge value meets the numerical value limit in the equipment operation state limit; if the adjusted charge and discharge value meets the numerical value limit, it is determined that the adjusted charge and discharge value matches the charge and discharge value of the corresponding period of the adjacent energy storage system.

6. An energy storage system AI collaborative scheduling system based on multi-objective optimization, characterized in that, A method for performing the multi-objective optimization-based energy storage system AI collaborative scheduling method according to any one of claims 1 to 5, comprising: a collection module for collecting user-side data through an edge node, the user-side data including: charging period distribution, historical power consumption, and power price; a construction module for constructing a power consumption preference vector based on the user-side data, and predicting a to-be-scheduled load curve of a future period based on the power consumption preference vector; an adjustment module for monitoring temperature difference data between a battery and a cabin environment in a target energy storage system in real time through a temperature sensor, and adjusting the speed of a cabin cooling fan and the opening degree of an inert gas charging and discharging valve through a fuzzy PID control method when the temperature difference data is equal to or greater than a preset temperature difference threshold, so that the adjusted temperature difference data is controlled within the preset temperature difference threshold range, and generating a device health report at a preset frequency during the adjustment process; a generation module for matching a preset optimization objective function including a plurality of optimization directions with the to-be-scheduled load curve to generate an initial charge and discharge plan of the target energy storage system; a correction module for correcting the initial charge and discharge plan based on reinforcement learning, the device health report, and the charge and discharge plan of the adjacent energy storage system to realize AI collaborative scheduling of the energy storage system.

7. An electronic device, comprising: comprising: a memory for storing a computer program; a processor for executing the computer program to implement the steps of the multi-objective optimization-based energy storage system AI collaborative scheduling method according to any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that, The computer readable storage medium stores a computer program, and the computer program is executed by the processor to implement the multi-objective optimization based energy storage system AI collaborative scheduling method in any one of claims 1 to 5.

Images