AI-based intelligent scheduling methods, devices, and storage media for energy storage power stations

Abstract

This invention relates to the fields of artificial intelligence, energy management, and smart grid technology, specifically to an AI-based intelligent scheduling method, device, and storage medium for energy storage power stations. The method includes: screening energy storage units based on the energy storage power station's operation plan and grid demand; selecting a pre-set number of energy storage units as priority scheduling units based on their capacity, efficiency, and status; calculating the optimal charging and discharging scheme for the energy storage power station during grid operation, as the scheduling ranking of the energy storage units; selecting the top-ranked energy storage unit from the priority scheduling units, performing optimized scheduling using an AI algorithm, adjusting the charging and discharging scheme in real time, and evaluating the scheduling effect; if the effect is unsatisfactory, then using the next-ranked energy storage unit to execute the scheduling strategy and evaluate the effect, repeating this process cyclically until the optimal energy storage unit is used to execute the scheduling strategy. This achieves efficient, flexible, and real-time optimized scheduling of energy storage power stations, improving the operating efficiency of energy storage power stations and the stability of the grid.

Description

Technical Field

[0001] This invention relates to the fields of artificial intelligence, energy management, and smart grid technology, and specifically to an intelligent scheduling method, device, and storage medium for energy storage power stations based on artificial intelligence. Background Technology

[0002] In the field of intelligent scheduling of energy storage power stations, existing technologies typically rely on traditional algorithms and human experience to select energy storage units and formulate charging and discharging schemes. These methods often suffer from low efficiency, slow response speed, and difficulty in adapting to rapid changes in grid demand. Especially when facing large-scale energy storage power stations and complex and ever-changing grid environments, traditional methods struggle to achieve optimal scheduling of energy storage units. Summary of the Invention

[0003] In view of this, the present invention provides an intelligent scheduling method, device and storage medium for energy storage power stations based on artificial intelligence, so as to solve the problem that the scheduling of energy storage units in the prior art is difficult to adapt to the grid demand.

[0004] In a first aspect, the present invention provides an intelligent scheduling method for energy storage power stations based on artificial intelligence. The method includes: screening energy storage units participating in scheduling in the energy storage power station according to the operation plan of the energy storage power station and the grid demand; sorting the screened energy storage units based on their efficiency and status, and selecting the first preset number of energy storage units after sorting as priority scheduling units; calculating the optimal charging and discharging scheme of each priority scheduling unit, and determining the scheduling order of the priority scheduling units based on the optimal discharging scheme; scheduling the first priority scheduling unit based on the scheduling order, adjusting the charging and discharging scheme of the scheduled priority scheduling unit using an artificial intelligence algorithm, and evaluating the scheduling effect; if the scheduling effect does not meet the preset requirements, scheduling the next priority scheduling unit, adjusting the charging and discharging scheme and evaluating the scheduling effect, until a priority scheduling unit that meets the scheduling effect is obtained for scheduling.

[0005] The intelligent scheduling method for energy storage power stations based on artificial intelligence provided in this invention achieves efficient operation of energy storage power stations and precise matching of grid demand through the screening, sorting, and calculation of optimal charging and discharging schemes for energy storage units, as well as the implementation of real-time optimized scheduling. Furthermore, by introducing artificial intelligence technology during scheduling, efficient and intelligent sorting and scheduling of energy storage units can be achieved. For example, artificial intelligence algorithms can be used to dynamically optimize and adjust charging and discharging schemes. This method not only improves the operating efficiency of energy storage power stations and reduces energy loss, but also enhances the stability and economy of the power grid.

[0006] In one optional implementation, the energy storage units participating in the scheduling of the selected energy storage power station include multiple combinations of energy storage units that meet the operation plan and grid requirements, with each combination including multiple energy storage units; the selected energy storage units are sorted based on their efficiency and status, and the top preset number of energy storage units after sorting are selected as priority scheduling units, including: weighting the importance of each energy storage unit in each combination to obtain a first score matrix reflecting the importance of the energy storage units; obtaining an efficiency status index reflecting the energy storage units based on the first parameter characteristics of the energy storage units and a pre-built efficiency status index prediction model; constructing a second score matrix reflecting the efficiency and status of the energy storage units based on the scores of the efficiency status index; and sorting the selected energy storage units according to a third score matrix obtained by weighted merging of the first and second score matrices, and selecting the top preset number of energy storage units after sorting as priority scheduling units.

[0007] In this embodiment, a first score matrix is ​​determined based on importance, a second score matrix is ​​determined based on efficiency and indicators, and a third score matrix, which is determined by weighting the first and second score matrices, is used to determine the priority scheduling unit, thus ensuring the scientificity and rationality of the scheduling scheme.

[0008] In one optional implementation, a combination of multiple energy storage units is determined using an optimization algorithm. The efficiency status indicators include charge / discharge efficiency and response time. A second fractional matrix reflecting the efficiency and status of the energy storage units is constructed based on the scores of the efficiency status indicators. This includes: constructing an indicator matrix by arranging the efficiency status indicators in descending order; constructing a fractional matrix of efficiency status based on cluster analysis; and multiplying the indicator matrix and the fractional matrix, and combining the energy storage unit numbers to obtain the second fractional matrix.

[0009] In this embodiment, an optimization algorithm is used to determine the combination of energy storage units, which can enhance the flexibility and adaptability of the scheduling scheme. When determining the second fractional matrix, it is determined by multiplying the index matrix and the fractional matrix, making the determined second fractional matrix more objective and achieving accurate evaluation of the energy storage unit efficiency status indicators.

[0010] In one optional implementation, calculating the optimal charging and discharging scheme for each priority scheduling unit includes: training a machine learning model based on the second parameter features of the energy storage unit and the optimal charging and discharging scheme to obtain a charging and discharging prediction model; and predicting the optimal charging and discharging scheme for each priority scheduling unit based on the current second parameter features of each priority scheduling unit and the charging and discharging prediction model.

[0011] In this embodiment, the optimal charging and discharging scheme is predicted by training a charging and discharging prediction model, which provides a data foundation for the scheduling and efficient scheduling of energy storage units.

[0012] In one alternative implementation, the second parameter features include the location, capacity, efficiency, current status, and matching degree with grid demand of the energy storage unit, and the optimal charging and discharging scheme includes the charging and discharging time point, power magnitude, and duration.

[0013] In this embodiment, by determining the second parameter characteristics, including the location, capacity, efficiency, current state, and matching degree with grid demand of the energy storage unit, the optimal charging and discharging scheme includes the charging and discharging time point, power magnitude, and duration, thus achieving the best matching between the energy storage unit and grid demand.

[0014] In one optional implementation, the charging and discharging scheme of the priority scheduling unit is adjusted using an artificial intelligence algorithm, including: adjusting the charging and discharging scheme of the priority scheduling unit based on a preset objective function and using a deep reinforcement learning algorithm according to the running context data.

[0015] In one optional implementation, the deep reinforcement learning algorithm uses a policy network and a value network to adjust the charging and discharging scheme; the policy network is used to obtain the charging and discharging scheme based on the operating scenario data; the value network is used to obtain the operating scenario data and the charging and discharging scheme, determine the matching degree between the energy storage power station and the grid demand and the operating efficiency of the energy storage unit as rewards, generate a reward function based on the rewards, and use the reward function to optimize the policy network.

[0016] In this embodiment, the dynamic adjustment and optimization of the energy storage unit's charging and discharging scheme can be achieved through the collaborative work of the policy network and the value network.

[0017] Secondly, the present invention provides an intelligent dispatching device for an energy storage power station based on artificial intelligence. The device includes: a screening module for screening energy storage units participating in dispatching in the energy storage power station according to the operation plan of the energy storage power station and the grid demand; a priority dispatching determination module for sorting the screened energy storage units based on their efficiency and status, and selecting the first preset number of energy storage units after sorting as priority dispatching units; a dispatching sorting module for calculating the optimal charging and discharging scheme of each priority dispatching unit, and determining the dispatching order of the priority dispatching units based on the optimal discharging scheme; and a sorting module for dispatching the first priority dispatching unit based on the dispatching order, adjusting the charging and discharging scheme of the dispatched priority dispatching unit using an artificial intelligence algorithm, and evaluating the dispatching effect. If the dispatching effect does not meet the preset requirements, the next priority dispatching unit is dispatched, the charging and discharging scheme is adjusted, and the dispatching effect is evaluated, until a priority dispatching unit that meets the dispatching effect is obtained for dispatching.

[0018] Thirdly, the present invention provides a computer device, comprising: a memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the computer instructions to perform the artificial intelligence-based intelligent scheduling method for energy storage power stations described in the first aspect or any corresponding embodiment.

[0019] Fourthly, the present invention provides a computer-readable storage medium storing computer instructions, which are used to cause a computer to execute the artificial intelligence-based intelligent scheduling method for energy storage power stations described in the first aspect or any of its corresponding embodiments.

[0020] Fifthly, the present invention provides a computer program product, including computer instructions, which are used to cause a computer to execute the artificial intelligence-based intelligent scheduling method for energy storage power stations described in the first aspect or any corresponding embodiment above. Attached Figure Description

[0021] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0022] Figure 1 This is a flowchart illustrating an intelligent scheduling method for energy storage power stations based on artificial intelligence, according to an embodiment of the present invention.

[0023] Figure 2 This is a flowchart illustrating another intelligent scheduling method for energy storage power stations based on artificial intelligence according to an embodiment of the present invention;

[0024] Figure 3 This is a structural block diagram of an intelligent dispatching device for an energy storage power station based on artificial intelligence, according to an embodiment of the present invention.

[0025] Figure 4 This is a schematic diagram of the hardware structure of a computer device according to an embodiment of the present invention. Detailed Implementation

[0026] As described in the background section, existing technologies struggle to adapt energy storage unit scheduling to grid demands, particularly when facing large-scale energy storage power plants and complex, ever-changing grid environments. Traditional methods often fail to achieve optimal scheduling of energy storage units. Specifically, firstly, traditional scheduling methods lack comprehensive consideration of the multi-dimensional characteristics of energy storage units, such as capacity, efficiency, and status, potentially leading to suboptimal scheduling schemes. Secondly, existing technologies suffer from insufficient computational power and real-time performance when processing large-scale and real-time data, making it difficult to meet the rapid response requirements of modern power grids. Furthermore, existing scheduling systems often lack self-learning and adaptive capabilities, failing to dynamically adjust and optimize based on the actual operating conditions of the grid. These issues limit the potential of energy storage power plants in improving grid stability and economic efficiency.

[0027] In view of this, this embodiment provides an intelligent scheduling method for energy storage power stations based on artificial intelligence. The method selects energy storage units based on the power station's operation plan, grid demand, and the efficiency and status of the energy storage units. Simultaneously, it prioritizes and schedules units based on optimal charging and discharging schemes, employing artificial intelligence algorithms to achieve scheduling and adjustment of charging and discharging schemes. This improves the potential for enhancing grid stability and economic efficiency.

[0028] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0029] According to an embodiment of the present invention, an embodiment of an intelligent scheduling method for energy storage power stations based on artificial intelligence is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0030] This embodiment provides an intelligent scheduling method for energy storage power stations based on artificial intelligence, which can be used in electronic devices such as computers, mobile phones, and tablets. Figure 1 This is a flowchart of an artificial intelligence-based intelligent scheduling method for energy storage power stations according to an embodiment of the present invention, such as... Figure 1 As shown, the process includes the following steps:

[0031] Step S101: Based on the operation plan of the energy storage power station and the grid demand, select energy storage units from the energy storage power station that will participate in the dispatch. Specifically, the types of energy storage units can be various forms such as battery energy storage systems, pumped hydro storage, and compressed air energy storage. When selecting energy storage units, the operation plan and grid demand can be specified, that is, the specific indicator requirements of the operation plan and grid demand can be determined, such as load demand, peak power requirements, and power quality standards. Then, based on the specific indicator requirements, select energy storage units from the energy storage power station that meet the requirements.

[0032] Step S102: Sort the selected energy storage units based on their efficiency and status, and select the top preset number of energy storage units as priority scheduling units. Specifically, for the selected energy storage units, further screening can be performed based on parameters such as their efficiency and status, including charging and discharging efficiency and power status.

[0033] Step S103: Calculate the optimal charging / discharging scheme for each priority scheduling unit, and determine the scheduling order of the priority scheduling units based on the optimal discharging scheme. Specifically, after determining the priority scheduling units, since there may be multiple priority scheduling units, it is necessary to determine the scheduling order of multiple priority scheduling units. In this embodiment, the scheduling order of the priority scheduling units is determined by the optimal charging / discharging scheme of each priority scheduling unit. The optimal charging / discharging scheme specifically includes relevant parameters for achieving optimal charging / discharging, and the scheduling order can be determined by comparing these relevant parameters.

[0034] Step S104: Based on the scheduling ranking, the highest-ranking priority scheduling unit is scheduled. An artificial intelligence algorithm is used to adjust the charging and discharging schemes of the priority scheduling unit and evaluate the scheduling effect. If the scheduling effect does not meet the preset requirements, the next priority scheduling unit is scheduled, and its charging and discharging scheme is adjusted and its scheduling effect evaluated, until a priority scheduling unit that meets the scheduling effect is selected for scheduling. Specifically, during scheduling, the highest-ranking energy storage unit in the priority scheduling unit is selected, and an artificial intelligence algorithm is used for optimized scheduling, adjusting the charging and discharging scheme in real time and evaluating the scheduling effect. If the effect is unsatisfactory, the next highest-ranking energy storage unit is used to execute the scheduling strategy and evaluate the effect, and this process is repeated until the optimal energy storage unit is used to execute the scheduling strategy. The artificial intelligence algorithm can be an intensity chemistry algorithm. Using an artificial intelligence algorithm, the charging and discharging strategy can be dynamically adjusted based on the real-time operating data of the energy storage power station and the grid demand to maximize energy storage efficiency and meet grid demand.

[0035] The intelligent scheduling method for energy storage power stations based on artificial intelligence provided in this invention achieves efficient operation of energy storage power stations and precise matching of grid demand through the screening, sorting, and calculation of optimal charging and discharging schemes for energy storage units, as well as the implementation of real-time optimized scheduling. Furthermore, by introducing artificial intelligence technology during scheduling, efficient and intelligent sorting and scheduling of energy storage units can be achieved. For example, artificial intelligence algorithms can be used to dynamically optimize and adjust charging and discharging schemes. This method not only improves the operating efficiency of energy storage power stations and reduces energy loss, but also enhances the stability and economy of the power grid.

[0036] This embodiment provides an intelligent scheduling method for energy storage power stations based on artificial intelligence, such as... Figure 2 As shown, the process includes the following steps:

[0037] Step S201: Based on the operation plan of the energy storage power station and the grid demand, select the energy storage units in the energy storage power station that will participate in the dispatch. Specifically, the selected energy storage units that will participate in the dispatch include multiple combinations of energy storage units that meet the operation plan and grid demand, with each combination including multiple energy storage units; the multiple combinations of energy storage units are determined using an optimization algorithm.

[0038] The search for energy storage unit combinations can be achieved using optimization algorithms such as genetic algorithms, particle swarm optimization, or other heuristic algorithms. These algorithms can search for multiple combination schemes of energy storage units while meeting grid demand and energy storage power station operation plans. When using optimization algorithms to screen energy storage unit combinations, evaluation indicators can first be determined based on the operation plan and grid demand. Then, the grid demand and energy storage power station operation constraints are transformed into optimization problems, and a mathematical model is established. Optimization algorithms such as genetic algorithms and particle swarm optimization can then be used to process the mathematical model, such as mapping energy storage unit combination schemes to individual representations in the algorithm. Finally, the selected energy storage power station combinations are obtained through iterative processing.

[0039] Meanwhile, during the search process, historical operating data of energy storage units and real-time changes in grid demand can be combined to dynamically adjust the search strategy, thereby improving the accuracy and adaptability of the search. Furthermore, in addition to basic optimization algorithms, more advanced optimization techniques such as simulated annealing or tabu search can be introduced to improve search efficiency and solution quality.

[0040] Step S202: Sort the selected energy storage units based on their efficiency and status, and select the first preset number of energy storage units after sorting as priority scheduling units.

[0041] Specifically, step S202 includes:

[0042] Step S2021: Weight the importance of each energy storage unit in each energy storage unit combination to obtain the first score matrix reflecting the importance of the energy storage units.

[0043] Specifically, the first fractional matrix can be determined through the following steps:

[0044] Step a1 involves ranking and assigning scores to the energy storage units in each energy storage unit combination according to their importance. Specifically, the importance of each energy storage unit can be determined based on factors such as its capacity, charge / discharge efficiency, charge / discharge speed, response time, and geographical location, and a corresponding score can be assigned. These factors can be adjusted based on the type, scale, and historical performance of the energy storage unit. Furthermore, a multi-objective optimization approach can be used to determine the importance of each energy storage unit, considering multiple factors simultaneously.

[0045] Step a2 involves accumulating the scores of each energy storage unit within each combination to determine its overall importance. Specifically, this accumulation can be achieved by summing the scores of each energy storage unit across different combinations to obtain a score reflecting its overall importance. Weighted averaging or other mathematical methods can be used to process the scores of energy storage units in different combinations during the accumulation process, thus more accurately reflecting their comprehensive importance. Furthermore, weighted scoring can fully consider the diversity and complementarity of energy storage units, making the determined scores more objective.

[0046] Step a3 involves mapping the scores of each energy storage unit to their corresponding numbers to form a first score matrix. Specifically, the first score matrix is ​​constructed by associating the scores of each energy storage unit with their numbers, creating a matrix that allows for quantitative comparison. Furthermore, data visualization techniques can be used to present the first score matrix, making the results more intuitive and easier to understand.

[0047] Step S2022: Based on the first parameter characteristics of the energy storage unit and the pre-built efficiency status index prediction model, an efficiency status index reflecting the energy storage unit is obtained.

[0048] Specifically, step S2022 above includes:

[0049] Step b1 involves selecting the first parameter features of the energy storage unit as input to the first machine learning model for calculating the energy storage unit's efficiency status index. Specifically, the first parameter features can be determined using feature selection techniques such as correlation coefficient, information gain, and chi-square test to identify the most predictive features, i.e., the features most critical to the energy storage unit's efficiency evaluation index, thereby reducing model complexity and improving computational efficiency. For example, the first parameter features may include the energy storage unit's capacity, charge / discharge efficiency, current status, historical operating data, and environmental factors. Capacity can be the energy storage unit's rated energy or actual usable energy; charge / discharge efficiency relates to the energy conversion efficiency of the energy storage unit during charging and discharging; current status includes the energy storage unit's charge level, temperature, and health status; historical operating data covers past charge / discharge records and fault records; and environmental factors include external conditions such as temperature, humidity, and load changes. These parameters can be acquired in real time through sensors or a pre-set data acquisition system.

[0050] After obtaining these features, they can be processed using necessary data cleaning and preprocessing techniques such as normalization and standardization to transform feature values ​​of different dimensions and ranges to a uniform scale, thereby ensuring data quality and consistency, while improving the convergence speed and prediction accuracy of the model.

[0051] The first machine learning model can be a Support Vector Machine (SVM), decision tree, neural network, random forest, etc. These models can effectively process and analyze the multidimensional data of energy storage units. Simultaneously, this first machine learning model can calculate efficiency state indicators through training, such as the charging and discharging efficiency and response time of the energy storage unit under different states.

[0052] Step b2 involves training and parameter tuning of the first machine learning model. Specifically, during training, the first parameter features can be used as input, with efficiency and state indicators as output. That is, the input data for the first machine learning model includes multi-dimensional features such as the energy storage unit's capacity, charging and discharging efficiency, current state, historical operating data, and environmental factors, while the output is the efficiency and state indicators. During training and parameter tuning, optimization algorithms such as batch gradient descent and stochastic gradient descent can be used to adjust the model parameters, and techniques such as cross-validation can be used to evaluate the model's performance. Regularization techniques such as L1 or L2 regularization can also be introduced to prevent overfitting and improve the model's predictive ability on unknown data. Furthermore, ensemble learning methods, such as random forests or gradient boosters, can be considered to improve the overall predictive ability by combining multiple weak predictive models. In practical applications, the weights of the input features can be dynamically adjusted according to the specific needs of the energy storage power station and the real-time status of the power grid to adapt to different scheduling strategies and optimization objectives.

[0053] Step b3 involves inputting the parameter features of the energy storage unit into the trained model to calculate the efficiency and state indices. Specifically, in practical applications, a multi-task learning method can be used to simultaneously predict multiple efficiency and state indices of the energy storage unit, thereby improving the practicality of the model.

[0054] Step S2023: Construct a second score matrix reflecting the efficiency and status of the energy storage unit based on the scores of the efficiency status indicators.

[0055] In an optional implementation, step S2023 includes:

[0056] Step b1: Arrange the efficiency status indicators in descending order to construct an indicator matrix.

[0057] Step b2: Construct a score matrix of efficiency status based on cluster analysis.

[0058] Step b3: Multiply the index matrix and the fractional matrix, and combine them with the energy storage unit number to obtain the second fractional matrix.

[0059] Specifically, when determining the score matrix, techniques such as cluster analysis can be used to classify the efficiency and state of energy storage units to better understand and analyze their performance. Then, the efficiency and state index column vectors of each energy storage unit, arranged in descending order, can be multiplied by the score matrix to obtain a score reflecting the efficiency and state of each unit. The scores of each energy storage unit are then mapped to their corresponding unit numbers to form a second score matrix. Furthermore, when calculating the scores, weighted averaging or other comprehensive scoring mechanisms can be introduced to ensure that the score matrix not only reflects the level of a single indicator but also comprehensively considers the balance of multiple indicators, thereby providing a more comprehensive evaluation of the energy storage unit's performance.

[0060] Step S2024: The selected energy storage units are sorted according to the third fractional matrix obtained by weighted merging of the first and second fractional matrices. A predetermined number of energy storage units are selected as priority scheduling units. Specifically, the number of priority scheduling units selected can be determined according to actual needs. This embodiment does not impose specific limitations on this.

[0061] Step S203: Calculate the optimal charging / discharging scheme for each priority scheduling unit, and determine the scheduling order of the priority scheduling units based on the optimal discharging scheme; for details, please refer to [link to relevant documentation]. Figure 1 Step S103 of the illustrated embodiment will not be described again here.

[0062] Step S204: Based on the scheduling order, the highest priority scheduling unit is scheduled. An artificial intelligence algorithm is used to adjust the charging and discharging schemes of the prioritized scheduling unit, and the scheduling effect is evaluated. If the scheduling effect does not meet the preset requirements, the next priority scheduling unit is scheduled, and the charging and discharging scheme is adjusted and the scheduling effect is evaluated again, until a priority scheduling unit that meets the scheduling effect is obtained for scheduling. For details, please refer to [link to details]. Figure 1 Step S104 of the illustrated embodiment will not be described again here.

[0063] This embodiment provides an intelligent scheduling method for energy storage power stations based on artificial intelligence, which includes the following steps:

[0064] Step S301: Based on the operation plan of the energy storage power station and the grid demand, select the energy storage units in the energy storage power station to participate in the dispatch; for details, please refer to [link to relevant documentation]. Figure 1 Step S101 of the illustrated embodiment will not be described again here.

[0065] Step S302: Based on the efficiency and status of the energy storage units, sort the selected energy storage units, and select the first preset number of energy storage units after sorting as priority scheduling units; for details, please refer to [link to relevant documentation]. Figure 1 Step S102 of the illustrated embodiment will not be described again here.

[0066] Step S303: Calculate the optimal charging and discharging scheme for each priority scheduling unit, and determine the scheduling order of the priority scheduling units based on the optimal discharging scheme.

[0067] Specifically, step S303 includes:

[0068] Step S3031: The machine learning model is trained based on the second parameter features of the energy storage unit and the optimal charging and discharging scheme to obtain the charging and discharging prediction model.

[0069] Specifically, the second parameter features include the location, capacity, efficiency, current status, and grid demand matching of the energy storage unit. Location information may affect the response speed and effectiveness of the energy storage unit to different areas of the grid; capacity and efficiency are key performance indicators of the energy storage unit; current status includes power level, health status, etc.; and grid demand matching involves the energy storage unit's ability to meet grid demands. These parameters can be obtained through real-time monitoring and data acquisition systems, and necessary data preprocessing and feature engineering can be performed to adapt to the model's requirements. The optimal charging and discharging scheme includes the charging and discharging time points, power levels, and durations to achieve the best match between the energy storage unit and grid demand.

[0070] The purpose of constructing a charge / discharge prediction model is to optimize the charging and discharging schemes of energy storage units in grid operation. Therefore, the machine learning model used for training can employ sequence models from deep learning, such as Long Short-Term Memory networks or Recurrent Neural Networks (RNNs), which are well-suited for processing time-series data and predicting charging and discharging strategies. For example, convolutional neural networks or recurrent neural networks from deep learning can be used to process grid demand data with time-series characteristics, and graph neural networks can be used to consider the spatial relationships of energy storage units.

[0071] The training and hyperparameter tuning of the machine learning model can be achieved through an automated machine learning pipeline, including steps such as data cleaning, feature selection, and hyperparameter optimization. During training, the second parameter features and the charging and discharging records of the energy storage units can be used as the training set to train the model to predict the optimal charging and discharging scheme through supervised learning. Simultaneously, during model training, regularization terms can be introduced to prevent overfitting, and early stopping can be used to avoid excessively long training times. Hyperparameter tuning can be accomplished using techniques such as grid search or Bayesian optimization to discover the optimal model parameter configuration.

[0072] To improve the practicality and adaptability of the model, online learning and incremental learning methods can be employed, allowing the model to self-update and optimize based on real-time power grid operation data. Simultaneously, the introduction of model interpretability tools, such as feature importance analysis or Local Interpretable Model-agnostic Explanations (LIME), can be considered to enhance model interpretability and help operators better understand the model's predictions.

[0073] In practical applications, a multi-task learning framework can be employed to simultaneously predict multiple output features related to charging and discharging schemes, thereby improving the model's predictive power and flexibility. Furthermore, reinforcement learning techniques can be introduced to optimize the generation process of charging and discharging schemes. Within this framework, the model can act as an agent, learning the optimal charging and discharging strategy through interaction with the power grid environment. Additionally, federated learning techniques can be used to process data from distributed energy storage units, allowing for model training and optimization while protecting data privacy. In actual operation, the input feature weights of the model can be dynamically adjusted based on the actual operating conditions of the power grid and changes in the performance of the energy storage units to adapt to different scheduling needs and optimization objectives.

[0074] In addition, after the model training is completed, the model performance can be evaluated using model evaluation metrics, such as mean squared error or mean absolute error.

[0075] Step S3032: Based on the current second parameter characteristics of each priority scheduling unit and the charging / discharging prediction model, the optimal charging / discharging scheme for each priority scheduling unit is predicted. Specifically, the acquired second parameter characteristics can be input into the charging / discharging prediction model to obtain the optimal charging / discharging scheme corresponding to each priority scheduling unit. The calculation results of these schemes will be used for actual energy storage unit scheduling and ranking to guide the operation of the energy storage power station.

[0076] Step S304: Based on the scheduling order, the highest priority scheduling unit is scheduled. The charging and discharging scheme of the priority scheduling unit is adjusted using an artificial intelligence algorithm, and the scheduling effect is evaluated. If the scheduling effect does not meet the preset requirements, the next priority scheduling unit is scheduled, and the charging and discharging scheme is adjusted and the scheduling effect is evaluated, until a priority scheduling unit that meets the scheduling effect is obtained for scheduling.

[0077] Specifically, the charging and discharging scheme of the priority scheduling unit is adjusted using artificial intelligence algorithms, including:

[0078] Step S3041: Based on the operational scenario data, the charging and discharging schemes of the priority scheduling units are adjusted using a deep reinforcement learning algorithm and a preset objective function. Specifically, in the optimization scheduling using artificial intelligence algorithms, an objective function is set according to grid demand and the operational scenario of the energy storage power station. Intelligent scheduling is performed based on the real-time operational data of the energy storage power station to obtain the charging and discharging scheme. The current charging and discharging scheme is evaluated and adjusted by collecting real-time operational data from the energy storage power station. During grid operation, online learning and updates are performed on the intelligent scheduling based on the real-time charging and discharging scheme and the corresponding operational scenario of the energy storage power station.

[0079] The objective functions set by the artificial intelligence algorithm in the optimized scheduling include minimizing charging and discharging costs, maximizing energy storage efficiency, and ensuring grid supply and demand balance. Real-time data collection involves the operating parameters of the energy storage power station, such as charging and discharging status, power levels, and environmental conditions. This data can be acquired in real time through Internet of Things (IoT) devices. Online learning and updating of intelligent scheduling refers to the algorithm's ability to continuously adjust and optimize charging and discharging schemes based on the latest operating data to adapt to changes in grid demand.

[0080] In setting the objective function, multi-objective optimization techniques can be employed to balance multiple objectives simultaneously, such as economy, reliability, and environmental impact. Real-time data acquisition can be achieved by establishing a centralized data monitoring system capable of processing and analyzing data streams from various energy storage units. Online learning and updates can be accomplished by introducing incremental learning or online learning algorithms, which can update model parameters based on new data without retraining the entire model.

[0081] More specifically, adaptive control strategies can be introduced to enhance the flexibility and robustness of intelligent scheduling. For example, Adaptive Dynamic Programming (ADP) algorithms can be used to handle uncertainties and changing grid conditions. Furthermore, Deep Reinforcement Learning (DeepRL) algorithms can be utilized to learn the optimal charging and discharging strategy through interaction with the grid environment. During implementation, edge computing technology can be considered to reduce data transmission latency and improve the real-time performance of scheduling decisions. Simultaneously, a feedback mechanism can be established to evaluate and adjust the charging and discharging scheme based on grid feedback signals, ensuring that the energy storage power station's scheduling decisions always align with the actual needs of the grid.

[0082] In one optional implementation, the deep reinforcement learning algorithm uses a policy network and a value network to adjust the charging and discharging scheme; the policy network is used to obtain the charging and discharging scheme based on the operating scenario data; the value network is used to obtain the operating scenario data and the charging and discharging scheme, determine the matching degree between the energy storage power station and the grid demand and the operating efficiency of the energy storage unit as rewards, generate a reward function based on the rewards, and use the reward function to optimize the policy network.

[0083] Specifically, the policy network generates and outputs a decision d based on the input scenario t; where scenario t is the current operating scenario of the energy storage power station, and decision d is the charging and discharging scheme guiding the next operating scenario of the energy storage power station; the "scenario" here refers to the state of the energy storage power station at a certain moment, including but not limited to the power level, charging and discharging status, grid demand, energy storage unit capacity, charging and discharging efficiency, current state, and matching degree with grid demand. The "decision" is the charging and discharging plan formulated based on these scenarios. Before receiving these scenario data, normalization processing can be performed to eliminate the influence of different dimensions between parameters. Normalization can use methods such as min-max normalization or Z-score normalization.

[0084] The value network evaluates the corresponding decision d (charge and discharge scheme) based on the received energy storage power station operating scenario t, guiding the policy network to optimize. In other words, the value network evaluates the effectiveness of these decisions, providing feedback to the policy network for optimization. Deep learning techniques can be used to construct these two networks, training them with extensive historical data to learn and predict optimal charge and discharge strategies.

[0085] Preferably, for example, in the policy network, a Deep Q-Network (DQN) can be used to learn optimal decisions under different scenarios. For instance, the policy network can employ a deep neural network containing multiple hidden layers to handle complex nonlinear relationships. In the value network, Monte Carlo Tree Search (MCTS) can be used to evaluate the potential value of charging and discharging schemes. Furthermore, a reward function can be set that comprehensively considers the matching degree between the energy storage power station and grid demand, as well as the operating efficiency of the energy storage unit, to guide the evaluation process of the value network. For example, the value network can employ a critique network to evaluate the decision quality of the policy network. In addition, an experience replay mechanism can be introduced to store historical scenarios and decisions for use during the training of the policy network and value network, improving learning efficiency.

[0086] More specifically, a multi-agent system can be introduced to enhance the coordination and adaptability of intelligent scheduling. In such a system, each energy storage unit can be considered as an agent, and they collaborate to optimize the charging and discharging strategy of the entire energy storage power station through communication and cooperation. Furthermore, federated learning techniques can be used to train the policy network and value network, allowing knowledge sharing while protecting the data privacy of each energy storage unit. In practical applications, the network's input parameters and reward function can be dynamically adjusted based on the real-time demand of the power grid and the operating status of the energy storage power station to achieve more precise and flexible intelligent scheduling.

[0087] In the process of evaluating charging and discharging schemes, the value network takes the matching degree between the energy storage power station and the grid demand and the operating efficiency of the energy storage unit after each round as the reward, and takes the cumulative reward from the start of the round to the current round as the reward function of the charging and discharging scheme in the current round.

[0088] Furthermore, multi-step rewards can be introduced to evaluate the long-term effects of the charging and discharging scheme, rather than just single-step rewards. This can be achieved by setting a reward discount factor, encouraging the model to consider the cumulative effects over multiple future time steps. Simultaneously, a policy gradient method can be used to optimize the policy network. This method allows the network to directly learn the gradient information of the policy, thereby updating network weights more effectively. In practical applications, the normalization method and reward function design can be dynamically adjusted according to the specific operating conditions of the energy storage power station and changes in grid demand to adapt to different scheduling scenarios and optimization objectives.

[0089] The intelligent scheduling method for energy storage power stations based on artificial intelligence provided in this invention can achieve efficient and intelligent sorting and scheduling of energy storage units by introducing artificial intelligence technology. First, the method comprehensively evaluates the capacity, efficiency, and status of energy storage units, and selects priority scheduling units through weighted sorting, ensuring the scientific and rational nature of the scheduling scheme. Second, by constructing and training machine learning models, the efficiency and status of energy storage units are accurately calculated, further improving the accuracy and reliability of scheduling. Furthermore, the method employs optimization algorithms to search and score combinations of energy storage units, enhancing the flexibility and adaptability of the scheduling scheme. Moreover, the intelligent scheduling method of this invention can respond to changes in grid demand in real time, dynamically optimizing and adjusting charging and discharging schemes through artificial intelligence algorithms. This method not only improves the operating efficiency of energy storage power stations and reduces energy loss, but also enhances the stability and economy of the power grid. Through online learning and update mechanisms, the system can continuously improve itself and adapt to more complex grid operating environments. Overall, the intelligent scheduling method of this invention can provide strong technical support for the efficient and stable operation of energy storage power stations.

[0090] The intelligent scheduling method for energy storage power stations based on artificial intelligence provided in this invention can achieve efficient management of energy storage units and precise response to grid demands. First, by using intelligent algorithms to weightedly sort and optimize the scheduling of energy storage units, the operating efficiency and response speed of the energy storage power station can be significantly improved. Second, by using machine learning models to evaluate the efficiency and status of energy storage units, the accuracy and reliability of scheduling decisions can be further improved. Furthermore, real-time data acquisition and online learning and updating mechanisms enable the system to flexibly adapt to changes in grid demand, ensuring the real-time nature and effectiveness of the scheduling scheme. Moreover, the intelligent scheduling method of this invention also has self-learning and self-optimization capabilities. Through the collaborative work of policy networks and value networks, dynamic adjustment and optimization of energy storage unit charging and discharging schemes can be achieved. This method not only improves the economic efficiency and environmental friendliness of energy storage power stations but also enhances the stability and resilience of the power grid. Through continuous technological iteration and optimization, this invention provides strong technical support for the intelligent management of energy storage power stations and the sustainable development of the power grid, possessing significant practical value and social significance.

[0091] This embodiment also provides an intelligent dispatching device for energy storage power stations based on artificial intelligence. This device is used to implement the above embodiments and preferred embodiments, and details already described will not be repeated. As used below, the term "module" can refer to a combination of software and / or hardware that performs a predetermined function. Although the device described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0092] This embodiment provides an intelligent dispatching device for energy storage power stations based on artificial intelligence, such as... Figure 3 As shown, it includes:

[0093] The screening module 31 is used to screen the energy storage units participating in the dispatch of the energy storage power station according to the operation plan of the energy storage power station and the grid demand.

[0094] The priority scheduling determination module 32 is used to sort the selected energy storage units based on their efficiency and status, and select the first preset number of energy storage units after sorting as priority scheduling units.

[0095] The scheduling and sorting module 33 is used to calculate the optimal charging and discharging scheme for each priority scheduling unit and determine the scheduling order of the priority scheduling units based on the optimal discharging scheme.

[0096] The sorting module 34 is used to sort the first priority scheduling unit based on the scheduling order, adjust the charging and discharging scheme of the priority scheduling unit using artificial intelligence algorithm, and evaluate the scheduling effect. If the scheduling effect does not meet the preset requirements, the next priority scheduling unit is scheduled, the charging and discharging scheme is adjusted and the scheduling effect is evaluated, until a priority scheduling unit that meets the scheduling effect is obtained for scheduling.

[0097] In one optional implementation, the energy storage units participating in the scheduling of the selected energy storage power station include multiple combinations of energy storage units that meet the operation plan and grid requirements, with each combination including multiple energy storage units; the priority scheduling determination module includes: a first score matrix determination module, used to weight and score the importance of each energy storage unit in each combination to obtain a first score matrix reflecting the importance of the energy storage units; an efficiency status index determination module, used to obtain an efficiency status index reflecting the energy storage units based on the first parameter characteristics of the energy storage units and a pre-built efficiency status index prediction model; a second score matrix determination module, used to construct a second score matrix reflecting the efficiency and status of the energy storage units based on the scores of the efficiency status index; and a priority scheduling determination submodule, used to sort the selected energy storage units according to a third score matrix obtained by weighted merging of the first and second score matrices, and select the first preset number of energy storage units after sorting as priority scheduling units.

[0098] In one optional implementation, the combination of multiple energy storage units is determined using an optimization algorithm, and the efficiency status indicators include charge / discharge efficiency and response time; the second fractional matrix determination module is specifically used to: construct an indicator matrix by arranging the efficiency status indicators in descending order; construct a fractional matrix of efficiency status based on cluster analysis; and multiply the indicator matrix and the fractional matrix, combined with the energy storage unit number, to obtain the second fractional matrix.

[0099] In one optional implementation, the scheduling and sorting module is specifically used to: train a machine learning model based on the second parameter features of the energy storage unit and the optimal charging and discharging scheme to obtain a charging and discharging prediction model; and predict the optimal charging and discharging scheme of each priority scheduling unit based on the current second parameter features of each priority scheduling unit and the charging and discharging prediction model.

[0100] In one alternative implementation, the second parameter features include the location, capacity, efficiency, current status, and matching degree with grid demand of the energy storage unit, and the optimal charging and discharging scheme includes the charging and discharging time point, power magnitude, and duration.

[0101] In one optional implementation, the sorting module is specifically used to: adjust the charging and discharging scheme of the priority scheduling unit based on the running context data, a preset objective function, and a deep reinforcement learning algorithm.

[0102] In one optional implementation, the deep reinforcement learning algorithm uses a policy network and a value network to adjust the charging and discharging scheme; the policy network is used to obtain the charging and discharging scheme based on the operating scenario data; the value network is used to obtain the operating scenario data and the charging and discharging scheme, determine the matching degree between the energy storage power station and the grid demand and the operating efficiency of the energy storage unit as rewards, generate a reward function based on the rewards, and use the reward function to optimize the policy network.

[0103] Further functional descriptions of the above modules are the same as those in the corresponding embodiments described above, and will not be repeated here.

[0104] This invention also provides a computer device having the above-described features. Figure 3 The image shows an AI-based intelligent dispatching device for energy storage power stations.

[0105] Please see Figure 4 , Figure 4 This is a schematic diagram of the structure of a computer device provided in an optional embodiment of the present invention, such as... Figure 4 As shown, the computer device includes one or more processors 10, memory 20, and interfaces for connecting the components, including high-speed interfaces and low-speed interfaces. The components communicate with each other via different buses and can be mounted on a common motherboard or otherwise installed as needed. The processors can process instructions executed within the computer device, including instructions stored in or on memory to display graphical information of a GUI on external input / output devices (such as display devices coupled to the interfaces). In some alternative implementations, multiple processors and / or multiple buses can be used with multiple memories and multiple memory modules, if desired. Similarly, multiple computer devices can be connected, each providing some of the necessary operations (e.g., as a server array, a group of blade servers, or a multiprocessor system). Figure 4 Take a processor 10 as an example.

[0106] Processor 10 may be a central processing unit, a network processor, or a combination thereof. Processor 10 may further include a hardware chip. The hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof. The programmable logic device may be a complex programmable logic device (CAMP), a field-programmable gate array (FPGA), a general-purpose array logic (GDA), or any combination thereof.

[0107] The memory 20 stores instructions executable by at least one processor 10 to cause at least one processor 10 to perform the method shown in the above embodiments.

[0108] The memory 20 may include a program storage area and a data storage area. The program storage area may store the operating system and applications required for at least one function; the data storage area may store data created based on the use of the computer device as shown by a landing page for an app. Furthermore, the memory 20 may include high-speed random access memory and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some alternative embodiments, the memory 20 may optionally include memory remotely located relative to the processor 10, which can be connected to the computer device via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0109] The memory 20 may include volatile memory, such as random access memory; the memory may also include non-volatile memory, such as flash memory, hard disk or solid-state drive; the memory 20 may also include a combination of the above types of memory.

[0110] The computer device also includes a communication interface 30 for communicating with other devices or communication networks.

[0111] This invention also provides a computer-readable storage medium. The methods described above according to embodiments of the invention can be implemented in hardware or firmware, or implemented as computer code that can be recorded on a storage medium, or implemented as computer code downloaded via a network and originally stored on a remote storage medium or a non-transitory machine-readable storage medium and then stored on a local storage medium. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code, which, when accessed and executed by the computer, processor, or hardware, implements the methods shown in the above embodiments.

[0112] A portion of this invention can be applied as a computer program product, such as computer program instructions, which, when executed by a computer, can invoke or provide the methods and / or technical solutions according to the invention through the operation of the computer. Those skilled in the art will understand that the forms in which computer program instructions exist in a computer-readable medium include, but are not limited to, source files, executable files, installation package files, etc. Correspondingly, the ways in which computer program instructions are executed by a computer include, but are not limited to: the computer directly executing the instructions, or the computer compiling the instructions and then executing the corresponding compiled program, or the computer reading and executing the instructions, or the computer reading and installing the instructions and then executing the corresponding installed program. Here, the computer-readable medium can be any available computer-readable storage medium or communication medium accessible to a computer.

[0113] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. An intelligent scheduling method for energy storage power stations based on artificial intelligence, characterized in that, The method includes: Based on the operation plan of the energy storage power station and the grid demand, select the energy storage units in the energy storage power station that will participate in the dispatch. The selected energy storage units are sorted based on their efficiency and status, and the first preset number of sorted energy storage units are selected as priority scheduling units. Calculate the optimal charging and discharging scheme for each priority scheduling unit, and determine the scheduling order of the priority scheduling units based on the optimal charging and discharging scheme; Based on the scheduling order, the highest priority scheduling unit is scheduled. The charging and discharging scheme of the priority scheduling unit is adjusted by artificial intelligence algorithm, and the scheduling effect is evaluated. If the scheduling effect does not meet the preset requirements, the next priority scheduling unit is scheduled, the charging and discharging scheme is adjusted and the scheduling effect is evaluated, until a priority scheduling unit that meets the scheduling effect is obtained for scheduling. The energy storage units participating in the dispatch of the selected energy storage power station include a variety of energy storage unit combinations that meet the operation plan and grid demand, and each energy storage unit combination includes multiple energy storage units; The selected energy storage units are sorted based on their efficiency and status. A predetermined number of these sorted units are then selected as priority scheduling units, including: The importance of each energy storage unit in each energy storage unit combination is weighted and scored to obtain the first score matrix reflecting the importance of the energy storage units; Based on the first parameter characteristics of the energy storage unit and the pre-built efficiency status index prediction model, an efficiency status index reflecting the energy storage unit is obtained. A second score matrix reflecting the efficiency and state of the energy storage unit is constructed based on the scores of the efficiency status indicators. The selected energy storage units are sorted according to the third fraction matrix obtained by weighted merging of the first and second fraction matrices, and the first preset number of energy storage units after sorting are selected as priority scheduling units.

2. The method according to claim 1, characterized in that, The combination of various energy storage units is determined using an optimization algorithm, and the efficiency status indicators include charge and discharge efficiency and response time. A second score matrix reflecting the efficiency and state of the energy storage unit is constructed based on the scores of efficiency status indicators, including: Construct an indicator matrix by arranging the efficiency status indicators in descending order; A score matrix of efficiency status is constructed based on cluster analysis; Multiply the index matrix and the fractional matrix, and combine them with the energy storage unit number to obtain the second fractional matrix.

3. The method according to claim 1, characterized in that, Calculate the optimal charging and discharging scheme for each priority scheduling unit, including: The machine learning model is trained based on the second parameter characteristics of the energy storage unit and the optimal charging and discharging scheme to obtain the charging and discharging prediction model; Based on the current second parameter characteristics of each priority scheduling unit and the charging and discharging prediction model, the optimal charging and discharging scheme of each priority scheduling unit is predicted.

4. The method according to claim 3, characterized in that, The second parameter features include the location, capacity, efficiency, current status, and matching degree with grid demand of the energy storage unit. The optimal charging and discharging scheme includes the charging and discharging time point, power magnitude, and duration.

5. The method according to claim 1, characterized in that, The charging and discharging scheme of the priority scheduling unit is adjusted using artificial intelligence algorithms, including: Based on operational scenario data, the charging and discharging scheme of the priority scheduling unit is adjusted using a preset objective function and a deep reinforcement learning algorithm.

6. The method according to claim 5, characterized in that, The deep reinforcement learning algorithm uses a policy network and a value network to adjust the charging and discharging scheme; The strategy network is used to obtain a charging and discharging scheme based on the operational context data; The value network is used to acquire the operating scenario data and charging / discharging scheme, determine the matching degree between the energy storage power station and the grid demand and the operating efficiency of the energy storage unit as rewards, generate a reward function based on the rewards, and use the reward function to optimize the strategy network.

7. An intelligent dispatching device for energy storage power stations based on artificial intelligence, characterized in that, The device includes: The screening module is used to screen energy storage units in the energy storage power station that participate in the dispatch based on the operation plan of the energy storage power station and the grid demand. The priority scheduling determination module is used to sort the selected energy storage units based on their efficiency and status, and select the first preset number of energy storage units after sorting as priority scheduling units. The scheduling and sorting module is used to calculate the optimal charging and discharging scheme for each priority scheduling unit and determine the scheduling order of the priority scheduling units based on the optimal charging and discharging scheme. The sorting module is used to schedule the highest priority scheduling unit based on the scheduling order, adjust the charging and discharging scheme of the priority scheduling unit using artificial intelligence algorithms, and evaluate the scheduling effect. If the scheduling effect does not meet the preset requirements, the next priority scheduling unit is scheduled, the charging and discharging scheme is adjusted and the scheduling effect is evaluated, until a priority scheduling unit that meets the scheduling effect is obtained for scheduling. The energy storage units participating in the dispatch of the selected energy storage power station include a variety of energy storage unit combinations that meet the operation plan and grid demand, and each energy storage unit combination includes multiple energy storage units; The selected energy storage units are sorted based on their efficiency and status. A predetermined number of these sorted units are then selected as priority scheduling units, including: The importance of each energy storage unit in each energy storage unit combination is weighted and scored to obtain the first score matrix reflecting the importance of the energy storage units; Based on the first parameter characteristics of the energy storage unit and the pre-built efficiency status index prediction model, an efficiency status index reflecting the energy storage unit is obtained. A second score matrix reflecting the efficiency and state of the energy storage unit is constructed based on the scoring of efficiency status indicators; The selected energy storage units are sorted according to the third fraction matrix obtained by weighted merging of the first and second fraction matrices, and the first preset number of energy storage units after sorting are selected as priority scheduling units.

8. A computer device, characterized in that, include: The system includes a memory and a processor, which are interconnected. The memory stores computer instructions, and the processor executes the computer instructions to perform the intelligent scheduling method for energy storage power stations based on artificial intelligence, as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions, which are used to cause the computer to execute the intelligent scheduling method for energy storage power stations based on artificial intelligence as described in any one of claims 1 to 6.

Images