Charging and discharging control method and device for energy storage system

Abstract

The invention provides a charging and discharging control method and device for an energy storage system. The method comprises the following steps: acquiring a multivariate heterogeneous data set; constructing a load prediction model based on the multivariate heterogeneous data set, the environmental factor weights corresponding to the environmental parameters in the multivariate heterogeneous data set, the power grid load data weight, the energy storage state data weight and the deep learning model, wherein the power grid load data weight and the energy storage state data weight correspond to the power grid load data and the energy storage state data; and in response to the fact that the prediction confidence coefficient of the load prediction model is smaller than a confidence coefficient threshold value, a target charging and discharging instruction is generated based on a load prediction value predicted by the load prediction model, the real-time energy storage state data and the real-time power grid data, and the target charging and discharging instruction is issued to the energy storage system for real-time charging and discharging control. Therefore, the load prediction model is constructed through the coupling relation of the environment, the load and the energy storage, the load fluctuation is predicted through environment perception, and therefore the response speed, the control precision and the operation safety stability of the energy storage power station to the sudden load fluctuation of the power grid are improved.

Description

Technical Field

[0001] This invention relates to the field of energy storage system technology, and in particular to a charging and discharging control method and device for an energy storage system. Background Technology

[0002] Energy storage systems are key components in building new power systems, and the performance of their charging and discharging control strategies directly affects the power quality, operational safety, and economy of the power grid. Current technologies rely on static models such as neural networks based on historical power data for control, completely ignoring the sudden and nonlinear effects of changes in the energy storage system's operating environment. This results in a passive response and significant response delays. Summary of the Invention

[0003] The present invention aims to at least partially solve one of the technical problems in the related art.

[0004] Therefore, the first objective of this invention is to provide a charging and discharging control method for an energy storage system.

[0005] The second objective of this invention is to provide a charging and discharging control device for an energy storage system.

[0006] The third objective of this invention is to provide an electronic device.

[0007] The fourth objective of this invention is to provide a non-transitory computer-readable storage medium storing computer instructions.

[0008] To achieve the above objectives, a first aspect of the present invention proposes a charging and discharging control method for an energy storage system. The method includes: acquiring a multivariate heterogeneous dataset, wherein the multivariate heterogeneous dataset includes environmental parameters corresponding to the energy storage power station, grid load data, and energy storage status data; determining the environmental factor weights, grid load data weights, and energy storage status data weights corresponding to the environmental parameters, grid load data, and energy storage status data, respectively; constructing a load prediction model based on the environmental factor weights, grid load data weights, energy storage status data weights, the multivariate heterogeneous dataset, and a deep learning model; and generating a target charging and discharging command based on the load prediction value predicted by the load prediction model, real-time energy storage status data, and real-time grid data, wherein the target charging and discharging command is sent to the energy storage system for real-time charging and discharging control.

[0009] To achieve the above objectives, a second aspect of the present invention provides a charging and discharging control device for an energy storage system. The device includes: a first acquisition module for acquiring a multivariate heterogeneous dataset, wherein the multivariate heterogeneous dataset includes environmental parameters corresponding to the energy storage power station, grid load data, and energy storage status data; a determination module for determining the environmental factor weights, grid load data weights, and energy storage status data weights corresponding to the environmental parameters, grid load data, and energy storage status data, respectively; a construction module for constructing a load prediction model based on the environmental factor weights, grid load data weights, energy storage status data weights, the multivariate heterogeneous dataset, and a deep learning model; and a first generation module for generating a target charging and discharging command based on the load prediction value predicted by the load prediction model, real-time energy storage status data, and real-time grid data, in response to the load prediction confidence level of the load prediction model being less than a confidence threshold. The target charging and discharging command is then sent to the energy storage system for real-time charging and discharging control.

[0010] To achieve the above objectives, a third aspect of the present invention provides an electronic device comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the method described in the first aspect.

[0011] To achieve the above objectives, a fourth aspect of the present invention provides a non-transitory computer-readable storage medium storing computer instructions for causing the computer to perform the method described in the first aspect.

[0012] The energy storage system charging and discharging control method, device, electronic device, and storage medium provided in this invention acquire a multivariate heterogeneous dataset. Based on the multivariate heterogeneous dataset, the environmental factor weights, grid load data weights, and energy storage status data weights corresponding to environmental parameters, grid load data, and energy storage status data within the multivariate heterogeneous dataset, and a deep learning model, a load prediction model is constructed. In response to the load prediction model's prediction confidence level being less than a confidence threshold, a target charging and discharging command is generated based on the load prediction value, real-time energy storage status data, and real-time grid data. This target charging and discharging command is then sent to the energy storage system for real-time charging and discharging control. Thus, the load prediction model constructed through the coupling relationship between environment, load, and energy storage enables environmental perception and load fluctuation prediction, thereby improving the energy storage power station's response speed, control accuracy, and operational safety and stability to sudden grid load fluctuations.

[0013] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0014] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein: Figure 1 This is a schematic flowchart of a charging and discharging control method for an energy storage system provided in an embodiment of the present invention; Figure 2 This is a schematic diagram illustrating the principle of a charging and discharging control method for an energy storage system provided in an embodiment of the present invention. Figure 3 This is a schematic diagram of the structure of a charging and discharging control device for an energy storage system provided in an embodiment of the present invention. Detailed Implementation

[0015] Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain the present invention, and should not be construed as limiting the present invention.

[0016] It should be noted that the acquisition, storage, use, and processing of data in the technical solution of this invention all comply with the relevant provisions of national laws and regulations.

[0017] The following description, with reference to the accompanying drawings, outlines an energy storage system charging and discharging control method, apparatus, electronic device, and storage medium according to embodiments of the present invention.

[0018] Figure 1 This is a schematic flowchart of a charging and discharging control method for an energy storage system provided in an embodiment of the present invention.

[0019] like Figure 1 As shown, the method includes the following steps: Step 101: Obtain a multivariate heterogeneous dataset, which includes environmental parameters, grid load data, and energy storage status data corresponding to the energy storage power station.

[0020] In some possible implementations, to ensure the accuracy of the multivariate heterogeneous dataset of the energy storage system, as an example, multivariate heterogeneous data is acquired, which includes the initial environmental parameters, initial grid load data, and initial energy storage status data corresponding to the energy storage power station; the multivariate heterogeneous data is preprocessed to obtain the environmental parameters, grid load data, and energy storage status data corresponding to the energy storage power station; wherein, the preprocessing is to eliminate noise, dimensional differences, and time misalignment of the multivariate heterogeneous data.

[0021] As an example, environmental parameters (e.g., temperature, humidity, wind speed, precipitation, irradiance, etc.), grid load data, and energy storage status data (e.g., state of charge (SOC), state of health (SOH), charge / discharge power, and battery temperature) can be collected in real time through devices such as micro weather stations, programming interfaces (satellite APIs) provided by satellite communication networks, supervisory control and data acquisition (SCADA) systems, and smart meters.

[0022] Specifically, environmental data can be collected by deploying high-density micro weather stations and satellite weather API interfaces, integrating satellite weather API data, surrounding weather station data, and local sensor data (core temperature and humidity sensors) to collect multi-dimensional meteorological data such as temperature, humidity, wind speed, precipitation, and irradiance, thus reconstructing a complete environmental field. Load data can be collected from grid-side SCADA data (e.g., 15-minute intervals) and user-side smart meter data (minute intervals). Energy storage data can be collected from the local controller of the energy storage system, including SOC, SOH, charging and discharging power, and battery temperature (second intervals). Finally, to facilitate the management of diverse and heterogeneous datasets, data can be stored in a format that includes timestamps, parameter names, raw values, and device identifiers.

[0023] In this embodiment of the invention, multi-source heterogeneous data can be collected synchronously to eliminate noise, dimensional differences, and time misalignment in the multi-source heterogeneous data, and output standardized time-series data to provide high-quality input for modeling. Specifically, noise elimination in multi-source heterogeneous data can be achieved through data cleaning, dimensional differences can be achieved through standardization, and time misalignment can be achieved through timestamp alignment, thereby obtaining a standardized and aligned time-series multi-source heterogeneous dataset (time granularity is uniformly 1 minute, data range is normalized to [0,1], and there are no outliers or missing values).

[0024] It should be noted that data cleaning: for data such as energy storage status data and grid load data, this can be achieved by combining the Laida criterion (…). Outliers are removed by associating them with environmental thresholds. For missing data, time-series linear interpolation is used to fill in the missing data. For example, if there are 3 or fewer consecutive missing data points, cubic spline interpolation is used to fill in the missing data; if there are more than 3 consecutive missing data points, historical data backtracking (e.g., load curves under the same environmental conditions for the last 7 days) is triggered.

[0025] Standardization: The Z-score standardization method can be used to normalize multi-source heterogeneous data to the [0,1] interval, eliminating dimensional differences. The normalization formula is as follows: ;in, For the multi-source heterogeneous data to be detected, The sample mean. This represents the sample standard deviation.

[0026] Timestamp alignment: The clocks of all devices can be synchronized via the Network Time Protocol (NTP) to perform timestamp alignment on discrete sampled multi-source heterogeneous data. For non-1-minute data (such as 10-minute satellite data and 15-minute SCADA data), cubic spline interpolation is used to unify it into a 1-minute time series.

[0027] Step 102: Determine the environmental factor weights, grid load data weights, and energy storage status data weights corresponding to the environmental parameters, grid load data, and energy storage status data, respectively.

[0028] In some possible implementations, in order to accurately determine the environmental factor weights, the environmental factor weights are obtained by the following steps: calculating the real-time correlation score between environmental parameters and power grid load data; and calculating the environmental factor weights based on the real-time correlation score.

[0029] As an example, environmental factor weights It is calculated by the softmax function, and the calculation formula is: , Environmental parameters With grid load data The real-time relevance score can be calculated using the Pearson correlation coefficient over a sliding time window. Environmental factors with higher real-time relevance scores have greater weights, allowing for differentiated attention to environmental factor weights. For example, during periods of high summer temperatures... They show a strong positive correlation, and the calculation yields... During precipitation They show a weak negative correlation, and the calculation yields... The energy storage state weight can be calibrated using aging curves provided by battery manufacturers.

[0030] Step 103: Construct a load forecasting model based on environmental factor weights, power grid load data weights, energy storage status data weights, multivariate heterogeneous datasets, and deep learning models.

[0031] As an example, the attention LSTM in a deep learning model can be used to dynamically focus the weights of environmental factors, grid load data, and energy storage status data, giving the load forecasting model strong regional and seasonal adaptability and significantly improving the prediction and control accuracy under different climatic conditions. Among these improvements, the attention LSTM... ) Construct a load forecasting model The method is as follows: .

[0032] in, For the future Time-based load forecasts (e.g., ultra-short-term) For 10 minutes to 1 hour, short-term (From 1 hour to 24 hours) For environmental parameters, Temperature (T), humidity (RH), wind speed (V), precipitation (P), and irradiance (G); For energy storage status data, State of charge (SOC), State of health (SOH), battery temperature ; For power grid load data, , This represents the power grid load data at time ti.

[0033] Therefore, multi-dimensional heterogeneous datasets at the minute level can be... The input is fed into an attention LSTM to output load forecast curves for the ultra-short term (10 minutes to 1 hour) and short term (1 hour to 24 hours).

[0034] It should be noted that, For example, a long short-term memory network with an attention mechanism, such as 64 hidden layer nodes, 100 iterations, and a learning rate of 0.001.

[0035] Step 104: In response to the load forecasting model's prediction confidence being less than the confidence threshold, a target charge / discharge command is generated based on the load forecast value predicted by the load forecasting model, real-time energy storage status data, and real-time grid data. The target charge / discharge command is then sent to the energy storage system for real-time charge / discharge control.

[0036] In some possible implementations, rolling calibration and prediction confidence levels enable the model to quickly adapt to the climate characteristics and sudden weather events of different regions and seasons. The prediction confidence level is determined based on the real-time and predicted values ​​of environmental parameters. In response to the load forecasting model's prediction confidence level being greater than or equal to a preset confidence threshold, a multivariate heterogeneous dataset within a set real-time time period is acquired. The weights of environmental factors, grid load data, and energy storage status data are adjusted using this multivariate heterogeneous dataset to obtain the real-time load forecasting model corresponding to a prediction confidence level less than the confidence threshold. Thus, by continuously optimizing using real-time rolling multivariate heterogeneous datasets, model errors are overcome, ensuring the long-term stable and reliable operation of the energy storage system in complex real-world environments.

[0037] For example, predict confidence level One possible calculation method is: ,in, (Number of environmental parameter categories) , These are the real-time and predicted values ​​for the environmental parameters, respectively.

[0038] Optionally, the confidence threshold can be 15%, such as... (If the prediction deviation is greater than or equal to 15%, the load forecasting model is not adaptable enough) and adjustments will be made. Additionally, if environmental parameters change abruptly (e.g., a sudden temperature change), adjustments will be made. Sudden precipitation This can also trigger adjustments to the load forecasting model.

[0039] In this embodiment of the invention, when the confidence level is greater than or equal to a preset confidence threshold, only the most recent hour's multivariate heterogeneous dataset (such as power grid load data and environmental parameter correlation samples during periods of environmental abrupt changes) can be retained to avoid interference from old data, thereby triggering online fine-tuning and updating of the load forecasting model. The weighting parameters are adjusted to adapt to environmental changes in order to obtain a real-time load forecasting model.

[0040] As an example, in response to the load forecasting model's prediction confidence level being less than a confidence threshold, a multi-objective optimization function and constraints for generating corresponding charge / discharge commands for the energy storage power station are determined. Based on the load forecast values ​​predicted by the load forecasting model, real-time energy storage status data, and real-time grid data, target charge / discharge commands that satisfy the multi-objective optimization function and constraints are generated. Simultaneously optimizing grid performance indicators (e.g., frequency deviation) and energy storage intrinsic indicators (e.g., SOC, health status, lifetime degradation) achieves global optimization, balancing economy and safety.

[0041] It should be noted that the multi-objective optimization function includes minimizing the grid frequency deviation function, the energy storage state fluctuation function, and the energy storage system operating cost function. The constraints include energy storage power boundary constraints, energy storage state safety range constraints, energy storage power change rate constraints, and grid dispatch command constraints corresponding to the energy storage system.

[0042] For example, with the goals of grid security and energy storage economy, a multi-objective optimization function and constraints are established. The optimal target charging and discharging command is generated through rolling, real-time optimization using multi-dimensional heterogeneous datasets, balancing multiple objective requirements. Specifically: The objective function for multi-objective optimization can be: , These are the objective weights for minimizing the grid frequency deviation function, the energy storage state fluctuation function, and the energy storage system operating cost function, respectively. It can be dynamically configured.

[0043] F is the grid frequency deviation cost function. It calculates the impact of future grid load changes on the frequency based on load forecasts, thereby optimizing the energy storage system's charging and discharging strategy to minimize frequency deviation. The greater the frequency deviation from the rated value, the higher the cost. , It is the actual power grid frequency (derived from load forecast values). It is the rated frequency of the power grid.

[0044] S is the energy storage state of flux function, such as SOC fluctuation cost or predicted load value. The future charging and discharging requirements are determined by the power allocation, which needs to minimize SOC fluctuations, avoid frequent SOC fluctuations or approaching the safety margin, and extend battery life. , This is the predicted SOC value for the battery. It is the target SOC value within the safe range.

[0045] C represents the operating cost function (primarily referring to electricity costs). By combining peak-valley electricity pricing and renewable energy output, the charging and discharging timing of the energy storage system is optimized to reduce user electricity costs. ,in: It is the predicted power demand of the power grid. It's the real-time electricity price.

[0046] In this embodiment of the invention, energy storage power boundary constraints are used to avoid extreme temperature overload; wherein, the upper limit of the charging and discharging power of the energy storage system varies with the battery temperature. Dynamically adjusted, therefore the constraint conditions for the energy storage power boundary constraints can be: , Let t be the charging and discharging power (kW) of the energy storage system at time t; Let t be the average battery temperature (°C). This represents the minimum / maximum charge / discharge power of the energy storage system.

[0047] In this embodiment of the invention, the state of energy storage safety range constraint is used to prevent overcharging and over-discharging, that is, the state of energy storage (SOC) must always be within the safety boundary in the prediction time domain: ,in, This represents the battery's minimum / maximum charge.

[0048] In this embodiment of the invention, the energy storage power change rate constraint is used to avoid grid impact, that is, to limit the sudden changes in the charging and discharging power of the energy storage system, such as: , This represents the maximum permissible change in charge and discharge power.

[0049] In this embodiment of the invention, the grid dispatch command constraint corresponding to the energy storage system is used to respond to the upper-level dispatch. That is, if the energy storage system also needs to respond to the dispatch command of the upper-level grid (such as AGC), this constraint can ensure that the total output is within the upper limit of the dispatch requirement, such as: ,in, The power specified by the superior power grid dispatch command (e.g., the dispatch requires the discharge to not exceed 80kW).

[0050] Furthermore, in some possible implementations, feedback data is acquired after the target charge / discharge command is issued to the energy storage system for real-time charge / discharge control. This feedback data includes actual grid load data, actual energy storage status data, and corresponding actual environmental parameters. Based on the feedback data, the multi-objective optimization function and constraints are calibrated to obtain calibrated multi-objective optimization functions and constraints, and a target charge / discharge command satisfying the calibrated multi-objective optimization function and constraints is generated. Through this real-time calibration mechanism, continuous external disturbances are mitigated, ensuring the long-term stable and reliable operation of the energy storage system under complex real-world environments.

[0051] The calibration includes, but is not limited to, feedforward calibration, feedback calibration, and adaptive calibration of safety boundaries.

[0052] As an example, feedforward calibration involves detecting sudden changes in environmental parameters (such as a sharp drop in irradiance due to cloud cover) at edge nodes and processing the signal of these changes. As a feedforward disturbance, the weights in the multi-objective optimization function are adjusted to compensate for the impact of the disturbance on the power grid load data in advance.

[0053] Feedback calibration: This is achieved by calculating actual grid load data. Compared with load forecast deviation , ,Will As a feedback quantity, the weights in the multi-objective optimization function are adjusted, the optimization problem is solved again, and the impact of compensation disturbances on grid load data is corrected.

[0054] Safety boundary adaptive: This refers to adjusting the energy storage power boundary constraints in the constraint conditions, for example, by adjusting them according to the battery temperature. Dynamically adjust the upper limit of the charging and discharging power of the energy storage system to avoid overcharging and over-discharging at high or low temperatures.

[0055] The energy storage system charging and discharging control method of this invention acquires a multivariate heterogeneous dataset; based on the multivariate heterogeneous dataset, the environmental factor weights, grid load data weights, and energy storage status data weights corresponding to the environmental parameters, grid load data, and energy storage status data in the multivariate heterogeneous dataset, and a deep learning model, a load prediction model is constructed; in response to the load prediction model's prediction confidence being less than a confidence threshold, a target charging and discharging command is generated based on the load prediction value predicted by the load prediction model, real-time energy storage status data, and real-time grid data, wherein the target charging and discharging command is sent to the energy storage system for real-time charging and discharging control. Thus, the load prediction model constructed through the coupling relationship between environment, load, and energy storage enables environmental perception and load fluctuation prediction, thereby improving the energy storage power station's response speed, control accuracy, and operational safety and stability to sudden load fluctuations in the grid.

[0056] To clearly illustrate the above embodiment, Figure 2 This is a schematic diagram illustrating the principle of a charging and discharging control method for an energy storage system provided in an embodiment of the present invention. Specifically, the multi-source data acquisition module is used to acquire environmental parameters of the energy storage power station through micro weather stations, satellite APIs, and local sensors; acquire grid load data of the energy storage power station through grid SCADA systems, smart meters, and other devices; and also to acquire energy storage status data of the energy storage power station, including SOC, SOH, charging and discharging power, and battery temperature. The edge preprocessing module is used to preprocess the multi-source heterogeneous data, including data cleaning, standardization, and timestamp alignment. The fusion modeling and decision-making module includes a fusion modeling module and a control decision-making module. The fusion modeling module receives the preprocessed multi-source heterogeneous dataset and constructs a load prediction model based on environmental factor weights, grid load data weights, energy storage status data weights, the multi-source heterogeneous dataset, and a deep learning model. It then determines whether to adjust and calibrate the load prediction model based on the prediction confidence of the load prediction model. The control decision-making module generates target charging and discharging commands that satisfy multi-objective optimization functions and constraints based on the load prediction values ​​predicted by the load prediction model, real-time energy storage status data, and real-time grid data. The calibration module executes the target charge / discharge commands and calibrates the multi-objective optimization function and constraints based on the feedback data after execution (feedforward calibration, feedback calibration, and adaptive calibration of the safety boundary) to regenerate the target charge / discharge commands that satisfy the calibrated multi-objective optimization function and constraints. Thus, by predicting load fluctuations through environmental perception, the control action is transformed from reactive remediation to proactive prevention, reducing the response delay from seconds to milliseconds, significantly improving the ability to suppress high-frequency fluctuations.

[0057] To achieve the above embodiments, the present invention also proposes a charging and discharging control device for an energy storage system.

[0058] Figure 3This is a schematic diagram of the structure of a charging and discharging control device for an energy storage system provided in an embodiment of the present invention.

[0059] like Figure 3 As shown, the energy storage system charge and discharge control device 30 includes: a first acquisition module 31, a determination module 32, a construction module 33, and a first generation module 34.

[0060] The system comprises the following modules: a first acquisition module 31, used to acquire a multivariate heterogeneous dataset, which includes environmental parameters, grid load data, and energy storage status data corresponding to the energy storage power station; a determination module 32, used to determine the environmental factor weights, grid load data weights, and energy storage status data weights corresponding to the environmental parameters, grid load data, and energy storage status data; a construction module 33, used to construct a load prediction model based on the environmental factor weights, grid load data weights, energy storage status data weights, the multivariate heterogeneous dataset, and a deep learning model; and a first generation module 34, used to generate a target charge / discharge command based on the load prediction value predicted by the load prediction model, real-time energy storage status data, and real-time grid data, in response to the load prediction confidence level of the load prediction model being less than a confidence threshold. The target charge / discharge command is then sent to the energy storage system for real-time charge / discharge control.

[0061] Furthermore, in one possible implementation of this invention, the first acquisition module 31 is specifically used to: acquire multi-dimensional heterogeneous data, which includes initial environmental parameters, initial grid load data, and initial energy storage status data corresponding to the energy storage power station; preprocess the multi-dimensional heterogeneous data to obtain the environmental parameters, grid load data, and energy storage status data corresponding to the energy storage power station; wherein, the preprocessing is to eliminate noise, dimensional differences, and time misalignment of the multi-dimensional heterogeneous data.

[0062] Furthermore, in one possible implementation of this invention, the determining module 32 is specifically used to: calculate the real-time correlation score between the environmental parameters and the power grid load data; and calculate the environmental factor weights based on the real-time correlation score.

[0063] Furthermore, in one possible implementation of this invention, the prediction confidence level is determined based on the real-time value and predicted value corresponding to the environmental parameters. The device further includes: a second acquisition module, used to acquire a multivariate heterogeneous dataset within a real-time set time period in response to the prediction confidence level of the load prediction model being greater than or equal to a preset confidence level threshold; and an adjustment module, used to adjust the weights of the environmental factors, the power grid load data, and the energy storage status data using the multivariate heterogeneous dataset within the real-time set time period, so as to obtain a real-time load prediction model corresponding to when the real-time prediction confidence level is less than the confidence level threshold.

[0064] Furthermore, in one possible implementation of this invention, the first generation module 34 is specifically configured to: in response to the load forecasting model having a prediction confidence level less than a confidence level threshold, determine a multi-objective optimization function and constraints for generating a charging and discharging command corresponding to the energy storage power station; and generate a target charging and discharging command that satisfies the multi-objective optimization function and constraints based on the load forecast value predicted by the load forecasting model, real-time energy storage status data, and real-time power grid data.

[0065] Furthermore, in one possible implementation of this invention, the multi-objective optimization function includes minimizing the grid frequency deviation function, the energy storage state fluctuation function, and the energy storage system operating cost function, and the constraints include energy storage power boundary constraints, energy storage state safety range constraints, energy storage power change rate constraints, and grid dispatch command constraints corresponding to the energy storage system.

[0066] Furthermore, in one possible implementation of this invention, the apparatus further includes: a third acquisition module, configured to acquire feedback data after the target charge / discharge command is sent to the energy storage system for real-time charge / discharge control, the feedback data including actual grid load data, actual energy storage status data, and corresponding actual environmental parameters; and a second generation module, configured to calibrate the multi-objective optimization function and constraints based on the feedback data to obtain calibrated multi-objective optimization function and constraints, and generate a target charge / discharge command that satisfies the calibrated multi-objective optimization function and constraints.

[0067] It should be noted that the foregoing explanation of the method embodiments also applies to the apparatus of this embodiment, and will not be repeated here.

[0068] The energy storage system charge / discharge control device of this invention extracts the highest temperature of the day from the operating temperature curve of the energy storage system; within a predetermined time window, it calculates the time-series temperature data corresponding to the highest temperature of the characteristic equipment of the energy storage system using a sliding time window method, and performs curve fitting on the linear function of the time-series temperature data and time to obtain the temperature rise trend of the characteristic equipment; based on the time-series temperature data and the temperature rise trend, it predicts the over-temperature time of the characteristic equipment within the prediction time window, as well as the temperature trend; based on the over-temperature time and the temperature trend, it predicts the health status of the characteristic equipment and generates a corresponding early warning detection strategy. Thus, based on the time-series temperature data and the temperature rise trend, it predicts the over-temperature time of the characteristic equipment of the energy storage system, and performs inspection and maintenance according to the generated early warning detection strategy to avoid failures.

[0069] To achieve the above embodiments, the present invention also proposes an electronic device, comprising: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the aforementioned method.

[0070] To implement the above embodiments, the present invention also proposes a non-transitory computer-readable storage medium storing computer instructions for causing the computer to perform the aforementioned method.

[0071] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0072] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this invention, "a plurality of" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0073] Any process or method description in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing custom logic functions or processes, and the scope of preferred embodiments of the invention includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as will be understood by those skilled in the art to which embodiments of the invention pertain.

[0074] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include: an electrical connection having one or more wires (electronic device), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Alternatively, the computer-readable medium may be paper or other suitable media on which the program can be printed, since the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in a computer memory.

[0075] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware as in another embodiment, it can be implemented using any of the following techniques known in the art, or a combination thereof: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0076] Those skilled in the art will understand that all or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.

[0077] Furthermore, the functional units in the various embodiments of the present invention can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.

[0078] The storage medium mentioned above can be a read-only memory, a disk, or an optical disk, etc. Although embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present invention.

Claims

1. A charging and discharging control method for an energy storage system, characterized in that, The method includes: Obtain a multi-dimensional heterogeneous dataset, wherein the multi-dimensional heterogeneous dataset includes environmental parameters, grid load data, and energy storage status data corresponding to the energy storage power station; Determine the environmental factor weights, grid load data weights, and energy storage status data weights corresponding to the environmental parameters, grid load data, and energy storage status data, respectively. Based on the environmental factor weights, power grid load data weights, energy storage status data weights, multivariate heterogeneous datasets, and deep learning models, a load prediction model is constructed. In response to the load forecasting model having a prediction confidence level less than a confidence level threshold, a target charge / discharge command is generated based on the load forecast value predicted by the load forecasting model, real-time energy storage status data, and real-time grid data. The target charge / discharge command is then sent to the energy storage system for real-time charge / discharge control.

2. The method according to claim 1, characterized in that, The acquisition of multivariate heterogeneous datasets includes: Acquire diverse and heterogeneous data, including initial environmental parameters, initial grid load data, and initial energy storage status data corresponding to the energy storage power station; The multivariate heterogeneous data is preprocessed to obtain the environmental parameters, grid load data, and energy storage status data corresponding to the energy storage power station; wherein, the preprocessing is to eliminate noise, dimensional differences, and time misalignment of the multivariate heterogeneous data.

3. The method according to claim 1, characterized in that, The environmental factor weights are obtained using the following steps: Calculate the real-time correlation score between the environmental parameters and the power grid load data; The environmental factor weights are calculated based on the real-time correlation scores.

4. The method according to claim 1, characterized in that, The prediction confidence level is determined based on the real-time values ​​and predicted values ​​corresponding to the environmental parameters, and the method further includes: In response to the prediction confidence of the load prediction model being greater than or equal to a preset confidence threshold, a multivariate heterogeneous dataset within a real-time set time period is obtained. The weights of environmental factors, power grid load data, and energy storage status data are adjusted using a multivariate heterogeneous dataset within a set real-time time period to obtain a real-time load prediction model when the real-time prediction confidence is less than a confidence threshold.

5. The method according to claim 4, characterized in that, In response to the load forecasting model's prediction confidence level being less than a confidence threshold, a target charge / discharge command is generated based on the load forecast value predicted by the load forecasting model, real-time energy storage status data, and real-time grid data. This target charge / discharge command is then sent to the energy storage system for real-time charge / discharge control, including: In response to the load forecasting model having a prediction confidence level less than a confidence threshold, a multi-objective optimization function and constraints for generating the corresponding charging and discharging commands for the energy storage power station are determined. Based on the load forecast values ​​predicted by the load forecast model, as well as real-time energy storage status data and real-time power grid data, target charging and discharging commands that satisfy the multi-objective optimization function and constraints are generated.

6. The method according to claim 5, characterized in that, in, The multi-objective optimization function includes minimizing the grid frequency deviation function, the energy storage state fluctuation function, and the energy storage system operating cost function. The constraints include energy storage power boundary constraints, energy storage state safety range constraints, energy storage power change rate constraints, and grid dispatch command constraints corresponding to the energy storage system.

7. The method according to claim 5, characterized in that, The method further includes: The system obtains feedback data after the target charge / discharge command is sent to the energy storage system for real-time charge / discharge control. The feedback data includes actual grid load data, actual energy storage status data, and corresponding actual environmental parameters. The multi-objective optimization function and constraints are calibrated based on the feedback data to obtain the calibrated multi-objective optimization function and constraints, and a target charge / discharge command that satisfies the calibrated multi-objective optimization function and constraints is generated.

8. A charging and discharging control device for an energy storage system, characterized in that, The device includes: The first acquisition module is used to acquire a multi-dimensional heterogeneous dataset, wherein the multi-dimensional heterogeneous dataset includes environmental parameters, grid load data and energy storage status data corresponding to the energy storage power station; The determination module is used to determine the environmental factor weights, grid load data weights, and energy storage status data weights corresponding to the environmental parameters, grid load data, and energy storage status data, respectively. The construction module is used to build a load prediction model based on the environmental factor weights, power grid load data weights, energy storage status data weights, multivariate heterogeneous datasets, and deep learning models. The first generation module is used to generate a target charge / discharge command based on the load forecast value predicted by the load forecast model, real-time energy storage status data, and real-time power grid data, in response to the load forecast confidence level of the load forecast model being less than the confidence level threshold. The target charge / discharge command is sent to the energy storage system for real-time charge / discharge control.

9. An electronic device, characterized in that, include: At least one processor; as well as A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-7.

10. A non-transitory computer-readable storage medium storing computer instructions, characterized in that, The computer instructions are used to cause the computer to perform the method according to any one of claims 1-7.

Images