Distributed energy storage regulation and control method, device and system and medium

Abstract

The embodiment of the invention provides a distributed energy storage regulation and control method, device and system and a medium, and belongs to the technical field of electric power. The method comprises the steps of performing data processing on various random factors associated with distributed energy storage regulation and control based on multi-algorithm collaboration to generate a random scene for distributed energy storage regulation and control; processing battery operation parameters associated with distributed energy storage regulation based on a neural network model to estimate battery aging cost, the neural network model taking a battery capacity fading rate as an output; and determining a distributed energy storage regulation and control strategy based on the generated random scene and the estimated battery aging cost. According to the embodiment of the invention, scientific regulation and control and resource optimization configuration of the energy storage system are effectively realized, random factors and the battery aging cost of neural network simulation are comprehensively considered, and efficient application and accurate regulation and control of distributed energy storage in a power system are promoted.

Description

Technical Field

[0001] This invention relates to the field of power technology, and more specifically to a distributed energy storage regulation method, device, system, and medium. Background Art

[0002] Against the backdrop of a global push for energy transition and "dual carbon" goals, distributed energy storage, as a key technology for achieving efficient energy utilization and enhancing power system flexibility, is experiencing rapid development. However, existing distributed energy storage control methods typically have the following drawbacks:

[0003] First, the handling of random factors is inaccurate.

[0004] When conducting distributed energy storage regulation, it is necessary to analyze various random factors affecting distributed energy storage regulation to provide data support for energy storage regulation. These random factors include, for example, renewable energy power generation, market electricity prices, and charging demand for energy storage vehicles. Their impact on energy storage regulation is specifically manifested in the following ways: Regarding renewable energy power generation, the large-scale integration of distributed power sources such as photovoltaics and wind power, while injecting new momentum into energy structure optimization, exacerbates the difficulty of balancing power supply and demand due to the intermittency and volatility of their output power; regarding market electricity prices, they fluctuate significantly due to multiple factors such as energy supply and demand, policy changes, and weather; and regarding charging demand for energy storage vehicles, it also exhibits complex and variable characteristics due to travel patterns and user habits. These random factors pose significant challenges to the operation and regulation of distributed energy storage systems. However, existing technologies use only single methods to analyze these random factors, failing to comprehensively capture the dynamic changes and interrelationships of influencing factors. This results in a large deviation between the data provided for energy storage regulation and the actual operating scenario, making it difficult for regulation strategies to adapt to complex and volatile external environments, and reducing the stability and efficiency of energy storage system operation.

[0005] Second, the cost assessment of battery aging is inaccurate.

[0006] As the core component of distributed energy storage, the aging process of batteries is not only affected by factors such as charge-discharge cycles and temperature, but also involves complex physicochemical changes. However, the traditional method only considers a few factors to establish simple empirical models, which cannot accurately reflect the complex physicochemical changes and multi-factor coupling effects during battery aging. This leads to large errors in battery aging cost assessment, which cannot provide a reliable basis for energy storage system cost accounting and affects the reasonable prediction and optimization management of the economic benefits of energy storage systems.

[0007] Therefore, with the deepening of power market reform, higher requirements have been placed on the response speed, regulation accuracy and operation economy of energy storage systems. Traditional energy storage regulation methods can no longer meet the needs of industry development, and it is necessary to develop a distributed energy storage regulation scheme that can effectively cope with random factors and accurately assess the cost of battery aging. Summary of the Invention

[0008] The purpose of this invention is to provide a distributed energy storage regulation method, device, system, and medium to at least partially solve the above-mentioned technical problems.

[0009] To achieve the above objectives, embodiments of the present invention provide a distributed energy storage regulation method, comprising: processing data on various random factors associated with distributed energy storage regulation based on multi-algorithm collaboration to generate random scenarios for distributed energy storage regulation; processing battery operating parameters associated with distributed energy storage regulation based on a neural network model to estimate battery aging costs, wherein the neural network model outputs battery capacity decay rate; and determining a distributed energy storage regulation strategy based on the generated random scenarios and the estimated battery aging costs.

[0010] Optionally, the data processing of various random factors related to distributed energy storage regulation based on multi-algorithm collaboration includes: collecting and preprocessing the various random factors; using a probability distribution estimation algorithm to estimate the probability distribution of the various random factors after preprocessing, so as to obtain the probability distribution corresponding to each random factor; and using a probability statistical algorithm to generate the random scenario for distributed energy storage regulation based on the obtained probability distribution of various random factors.

[0011] Optionally, the multiple random factors include any number of market electricity price fluctuation data, renewable energy power generation data, energy storage vehicle charging demand data, and charging / swapping load data. Furthermore, the probability distribution estimation algorithm used for the market electricity price fluctuation data is the Autoregressive Integrated Moving Average (ARIMA) model from time series analysis algorithms; the probability distribution estimation algorithm used for the renewable energy power generation data is a kernel density estimation algorithm; and the probability distribution estimation algorithms used for the energy storage vehicle charging demand data and the charging / swapping load data are clustering algorithms. Additionally, the probability statistics algorithm employs a Monte Carlo simulation algorithm.

[0012] Optionally, the process of processing battery operating parameters related to distributed energy storage regulation based on a neural network model includes: acquiring the battery operating parameters; training the neural network model using historical battery operating data, wherein the acquired battery operating parameters are used as input to the neural network model; and estimating the battery aging cost based on the battery capacity decay rate output by the neural network model.

[0013] Optionally, the neural network model is an LSTM model.

[0014] Optionally, while processing the battery operating parameters related to distributed energy storage regulation based on the neural network model, the distributed energy storage regulation method further includes: processing the battery operating parameters based on an empirical model, and combining the processing result with the processing result of the neural network model to estimate the battery aging cost, wherein the empirical model is constructed based on battery operating experience data and outputs the battery capacity decay rate.

[0015] Optionally, determining the distributed energy storage control strategy includes: setting multiple optimization objectives and multiple constraints for distributed energy storage control based on the random scenario and the battery aging cost; and using an optimization algorithm to determine the distributed energy storage control strategy based on the set multiple optimization objectives and multiple constraints, wherein the distributed energy storage control strategy includes an optimal charging and discharging strategy and a power allocation strategy.

[0016] Optionally, the plurality of optimization objectives include minimizing operating costs, maximizing the reliability of the distributed energy storage system, and maximizing the renewable energy absorption rate; and / or the plurality of constraints include any number of the following for the energy storage system: power balance constraints, voltage constraints, frequency constraints, charge / discharge power constraints, battery capacity constraints, charge / discharge cycle limits, charge / discharge efficiency constraints, and battery life constraints.

[0017] On the other hand, embodiments of the present invention also provide a distributed energy storage regulation device, comprising: a memory storing a program capable of running on a processor; and the processor configured to implement any of the above-described distributed energy storage regulation methods when executing the program.

[0018] On the other hand, embodiments of the present invention also provide a distributed energy storage system, including any of the above-mentioned distributed energy storage control devices.

[0019] On the other hand, embodiments of the present invention also provide a machine-readable storage medium storing instructions that cause a machine to execute any of the above-described distributed energy storage control methods.

[0020] Through the above technical solutions, the embodiments of the present invention effectively realize the scientific regulation and resource optimization of energy storage systems, comprehensively consider random factors and battery aging costs simulated by neural networks, and promote the efficient application and precise regulation of distributed energy storage in power systems.

[0021] Other features and advantages of the embodiments of the present invention will be described in detail in the following detailed description section. Attached Figure Description

[0022] The accompanying drawings are provided to further illustrate embodiments of the present invention and form part of the specification. They are used together with the following detailed description to explain the embodiments of the present invention, but do not constitute a limitation thereof. In the drawings:

[0023] Figure 1 This is a flowchart illustrating the distributed energy storage regulation method according to Embodiment 1 of the present invention;

[0024] Figure 2 This is a schematic diagram of the data processing procedure for random factors in a preferred embodiment of the present invention;

[0025] Figure 3 This is a flowchart illustrating the processing of battery operating parameters based on a neural network model in a preferred embodiment of the present invention.

[0026] Figure 4 This is a flowchart illustrating the process of determining a distributed energy storage control strategy in a preferred embodiment of the present invention; and

[0027] Figure 5 This is a schematic diagram of the distributed energy storage and regulation device according to Embodiment 3 of the present invention. Detailed Implementation

[0028] The specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are for illustration and explanation only and are not intended to limit the scope of the present invention.

[0029] It should be noted that the acquisition, transmission, storage, use, and processing of data in this application comply with relevant laws and regulations. In the embodiments of this invention, certain existing industry solutions such as software, components, and models may be mentioned. These should be considered exemplary, intended only to illustrate the feasibility of implementing the technical solution of this application, and do not imply that the applicant has already used or necessarily used such solutions.

[0030] Before introducing the specific embodiments of the present invention, some terms involved in the embodiments of the present invention will be introduced first, so as to better understand the embodiments of the present invention.

[0031] 1. Distributed Energy Storage Regulation: This refers to the centralized or coordinated management of distributed energy storage systems (such as batteries and supercapacitors). By optimizing charging and discharging strategies, it aims to improve the balance, economy, and reliability of the power system. The core of distributed energy storage regulation lies in coordinating multiple distributed energy storage units to address issues such as renewable energy fluctuations, changes in load demand, or grid failures.

[0032] 2. ARIMA (AutoRegressive Integrated Moving Average) model: also known as the autoregressive summation moving average model, it transforms a non-stationary time series into a stationary time series, and then regresses the dependent variable only on its lagged values ​​(autoregressive model) and the present value and lagged values ​​of the random error term (moving average model).

[0033] 3. K-means (K-means Clustering) Clustering Algorithm: A commonly used unsupervised learning algorithm used to divide a dataset into K non-overlapping clusters, such that data points within the same cluster are as similar as possible, while data points between different clusters are as different as possible. K-means is widely used in tasks such as data clustering, feature engineering, image segmentation, and anomaly detection.

[0034] 4. Adam Optimizer (Adaptive Moment Estimation): An optimization algorithm with an adaptive learning rate that combines the advantages of momentum optimization and RMSProp. The Adam optimizer is widely used in deep learning training, especially suitable for large-scale datasets and high-dimensional parameter optimization problems.

[0035] 5. ReLU (Rectified Linear Unit) is a non-linear activation function used in neural networks to enable neurons to learn complex patterns. The mathematical definition of ReLU is: f(x) = max(0,x).

[0036] 6. LSTM (Long Short-Term Memory) is a special type of recurrent neural network (RNN) specifically designed to solve the gradient vanishing or exploding problems that traditional RNNs encounter when processing long-sequence data. LSTM models perform exceptionally well in processing time-series data, effectively capturing long-term dependencies during battery aging.

[0037] 7. MSE (Mean Squared Error): A common regression loss function used to measure the difference between predicted and true values. It calculates the mean of the squared errors of all samples.

[0038] 8. Energy utilization rate: This refers to the proportion of energy that is effectively utilized during production and use, and is usually used to measure the utilization rate of renewable energy sources (such as wind power and solar power). Specifically, the energy utilization rate is the ratio between the actual amount of energy used and the total amount of electricity generated.

[0039] The embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0040] Example 1

[0041] Figure 1 This is a flowchart illustrating the distributed energy storage control method according to Embodiment 1 of the present invention. Figure 1 As shown, the distributed energy storage control method may include the following steps S100-S300.

[0042] Step S100: Based on multi-algorithm collaboration, data processing is performed on various random factors related to distributed energy storage regulation to generate random scenarios for distributed energy storage regulation.

[0043] Step S200: Based on a neural network model, the battery operating parameters associated with distributed energy storage regulation are processed to estimate the battery aging cost, wherein the neural network model outputs the battery capacity decay rate.

[0044] Step S300: Based on the generated random scenario and the estimated battery aging cost, determine the distributed energy storage control strategy.

[0045] According to the above steps S100-S300, the embodiments of the present invention first accurately characterize random factors based on multi-algorithm collaboration, providing a random scenario with small deviation from the actual operation scenario for distributed energy storage regulation; secondly, accurately evaluate the battery aging cost using a neural network model, providing support for energy storage cost accounting; finally, comprehensively consider the random scenario and battery aging cost to determine the distributed energy storage regulation strategy, which can achieve precise regulation of the distributed energy storage system.

[0046] Regarding step S100, Figure 2 A schematic flowchart illustrating data processing for random factors in a preferred embodiment is shown. Figure 2 As shown, the following steps S110-S130 may be included:

[0047] Step S110: Collect and preprocess the various random factors.

[0048] The various random factors include any combination of market electricity price fluctuation data, renewable energy power generation data, energy storage vehicle charging demand data, and charging / swapping load data. For example, various sensors and data acquisition devices are used to comprehensively collect market electricity price fluctuation data, real-time power data of local distributed renewable energy generation, energy storage vehicle charging demand data, and charging / swapping load data. Then, data cleaning algorithms are used to remove noise and outliers from the data, ensuring the accuracy and reliability of the data and laying a solid foundation for subsequent analysis.

[0049] Step S120: The probability distribution estimation algorithm is used to estimate the probability distribution of the various random factors after the preprocessing, so as to obtain the probability distribution corresponding to each random factor.

[0050] In the example, the probability distribution estimation algorithm used for the market electricity price fluctuation data is the Autoregressive Integrated Moving Average (ARIMA) model in time series analysis; the probability distribution estimation algorithm used for the renewable energy power generation data is the kernel density estimation algorithm; and the probability distribution estimation algorithm used for the energy storage vehicle charging demand data and the charging / swapping load data is a clustering algorithm. The specific descriptions based on the type of random factor are as follows:

[0051] 1. Electricity price probability distribution.

[0052] In existing technologies, market electricity price forecasting often relies on simple time series analysis models, analyzing only the time trends of historical electricity price data. This makes it difficult to account for the impact of sudden changes in energy supply and demand, temporary policy adjustments, and other unforeseen factors. Therefore, this invention employs the ARIMA model from time series analysis algorithms to conduct an in-depth analysis of the changing patterns of electricity prices over time. Let the time series Y... t Representing electricity price, the ARIMA(p,d,q) model can be expressed as:

[0053] Φ(B) d Y t =Θ(B)∈ t (1)

[0054] In the formula, It is an autoregressive operator. B is the moving average operator, and ∈ t It is a noise sequence, Y t This represents the electricity price over a time series t. The model parameters p, d, q, and autoregressive coefficients are determined by fitting historical electricity price data. and moving average coefficient θ i This allows us to determine the probability distribution function of electricity prices.

[0055] 2. Probability distribution of distributed renewable energy generation power.

[0056] Existing technologies often rely on historical statistical data and fixed probability distribution assumptions, such as simply setting it as a normal distribution, to deal with the randomness of renewable energy power generation, without combining the real-time characteristics of the power generation equipment for dynamic analysis.

[0057] To address this, this invention combines historical data and power generation equipment characteristics, employing a kernel density estimation method to calculate the probability distribution of distributed renewable energy power generation. Specifically, for a set of observation data x1, x2, ..., x...n Its probability density function is estimated as follows:

[0058]

[0059] In the formula, K(·) is the kernel function (such as the commonly used Gaussian kernel function). ), where h is the bandwidth parameter. By adjusting the bandwidth parameter h, the probability distribution of power generation of distributed renewable energy is determined based on historical power data.

[0060] 3. Probability distribution of charging demand and charging / swapping load for energy storage vehicles.

[0061] In existing technologies, the charging demand for energy storage vehicles is generally estimated based on rough experience or fixed demand patterns in local areas, which cannot accurately match the ever-changing travel patterns and user habits.

[0062] This embodiment of the invention utilizes methods such as cluster analysis, including the K-means clustering algorithm, to calculate the corresponding probability distribution. Specifically, for n data points x1, x2, ..., x... n First, randomly select k initial cluster centers c1, c2, ..., c3. k Next, the Euclidean distance from each data point to each cluster center is calculated. (Where m is the dimension of the data), the data points are assigned to the cluster containing the nearest cluster center. Then, the cluster center of each cluster is updated to the mean of all data points within that cluster, and this process is repeated until the cluster centers no longer change. Through cluster analysis, different demand patterns are identified, and the probability of each pattern occurring is determined.

[0063] Step S130: Based on the probability distribution of the obtained random factors, a probabilistic statistical algorithm is used to generate the random scenario for distributed energy storage regulation.

[0064] The probability and statistical algorithm used here is the Monte Carlo simulation algorithm. For example, using the Monte Carlo simulation algorithm, a large number of random scenarios are generated based on the determined probability distribution functions of each random factor. Assume there are N random factors, and the probability distribution function of each random factor i is f. i (x i By randomly sampling within the range of values ​​for each random factor, a set of random scenarios S is obtained. j =(x 1j ,x 2j ,…,x Nj ), where j = 1, 2, ..., M (M is the number of generated random scenarios). Each scenario contains different combinations of random factors at a specific moment, thereby simulating various complex situations that may occur in actual operation.

[0065] Regarding step S200, Figure 3 A flowchart illustrating the processing of battery operating parameters based on a neural network model in a preferred embodiment is shown. Figure 3 As shown, the following steps S210-S230 may be included:

[0066] Step S210: Obtain the battery operating parameters.

[0067] For example, in a distributed energy storage system, parameters such as battery charging and discharging current (I), voltage (V), and temperature (T) are monitored in real time. The battery management system collects this data and records information such as the number of charge-discharge cycles (n) and usage time (t), providing data support for the subsequent construction of a battery aging cost estimation model.

[0068] Step S220: Train the neural network model using the battery's historical operating data.

[0069] The neural network model is preferably an LSTM model, and the acquired battery operating parameters are used as inputs to the neural network model. For example, an LSTM neural network model is constructed, which takes the battery's charge / discharge current I, voltage V, temperature T, and number of charge / discharge cycles n as inputs, and the battery's capacity decay rate as input. As the output, the network structure is designed as follows:

[0070] 1. Input layer: The input dimension is 4, corresponding to the four input features mentioned above.

[0071] 2. LSTM layer: Set up 2 layers, each containing 128 neurons, so as to capture long-term dependencies in time series.

[0072] 3. Fully connected layer: There are 2 layers, which use the ReLU activation function and the linear activation function respectively.

[0073] 4. Loss function: MSE is used, and the Adam optimizer is used to update the model parameters.

[0074] The neural network is trained using historical data, and 5-fold cross-validation is employed during training to prevent overfitting. By comparing the loss curves of the training and validation sets, the model with the minimum loss on the validation set is selected as the optimal model.

[0075] Existing technologies often rely on simple empirical formulas, considering only a few factors such as the number of battery charge-discharge cycles and current, to construct linear or relatively simple nonlinear models. For example, models are built based on the linear relationship between battery capacity decay and the number of charge-discharge cycles, neglecting the combined effects of other key factors such as temperature and voltage on battery aging, and failing to account for the complex physicochemical changes during battery aging. This invention, however, employs a trained neural network model to construct an accurate battery aging cost assessment model.

[0076] Step S230: Estimate the battery aging cost based on the battery capacity decay rate output by the neural network model.

[0077] That is, a battery aging model was constructed through steps S210-S230. For example, the remaining battery life L predicted by the battery aging model... remaining And the degree of aging (e.g., characterized by capacity decay rate ΔC / C0), combined with the initial cost C of the battery. initial and replacement cost C replace Calculate the aging cost of the battery at different operating stages. Assume the economic life of the battery is L. economic Then the aging cost C of the battery at time t is... aging (t) can be converted to:

[0078]

[0079] Battery aging cost converted to C per unit time aging-per-time Or unit of electricity C aging-per-energy So that it can be comprehensively considered in subsequent regulatory strategies, such as (Δt is the time interval), (ΔE represents the change in electrical charge).

[0080] Regarding step S300, Figure 4 A flowchart illustrating the process of determining a distributed energy storage control strategy in a preferred embodiment is shown. Figure 4 As shown, the following steps S310-S320 may be included:

[0081] Step S310: Based on the random scenario and the battery aging cost, set multiple optimization objectives and multiple constraints for distributed energy storage regulation.

[0082] For example, the multiple optimization objectives include minimizing operating costs, maximizing the reliability of distributed energy storage systems, and maximizing the renewable energy absorption rate, which are combined to form a comprehensive optimization objective.

[0083] Among them, operating cost C total Covering electricity procurement costs C purchase Battery aging cost Caging And equipment maintenance costs C maintenance ,Right now:

[0084] C total =C purchase +C aging +C maintenance (4)

[0085] Electricity procurement costs Where P purchase p(t) is the power purchased from the grid at time t, and p(t) is the electricity price at time t. Equipment maintenance cost C maintenance It can be calculated based on the equipment's maintenance plan and maintenance cost rate.

[0086] System reliability is measured by several indicators, such as voltage deviation ΔV and frequency deviation Δf, with the goal of minimizing these indicators.

[0087] Renewable energy consumption rate The goal is to maximize the absorption rate, where P renewable-consumed (t) represents the renewable energy power consumed at time t, P renewable-generated (t) represents the power generation of renewable energy at time t.

[0088] To further illustrate, the multiple constraints include any combination of the following for the energy storage system: power balance constraints, voltage constraints, frequency constraints, charge / discharge power constraints, battery capacity constraints, charge / discharge cycle limits, charge / discharge efficiency constraints, and battery life constraints. These are detailed below:

[0089] 1. Power balance constraints in power systems:

[0090] P generation (t)+P storage-discharge (t)-P storage-charge (t)=P load (t)+P loss (t) (5)

[0091] In the formula, P generation P(t) is the total power generation of the power generation equipment at time t. storage-discharge (t) and P storage-charge (t) represent the discharge power and charging power of the energy storage system at time t, respectively. load (t) is the load power at time t, P loss (t) represents the power loss of the power system at time t.

[0092] 2. Voltage constraint:

[0093] V min ≤V(t)≤V max(6)

[0094] Where V(t) is the voltage at a node in the power system at time t, V min and V max These are the lower and upper limits of the node voltage, respectively.

[0095] 3. Frequency constraints:

[0096] f min ≤f(t)≤f max (7)

[0097] Where f(t) is the frequency of the power system at time t, f min and f max These are the lower and upper limits of the frequency, respectively.

[0098] 4. Charging and discharging power limitations of distributed energy storage systems:

[0099] 0≤P storage-charge (t)≤P charge-max

[0100] 0≤P storage-discharge (t)≤P discharge-max (8)

[0101] Among them, P charge-max and P discharge-max These are the maximum charging power and maximum discharging power of the energy storage system, respectively.

[0102] 5. Battery capacity limitation:

[0103] E min ≤E(t)≤E max (9)

[0104] Where E(t) is the remaining capacity of the battery at time t, E min and E max These are the lower and upper limits of battery capacity, respectively.

[0105] 6. Charge / discharge cycle limit:

[0106] n charge (t)≤n charge-limit

[0107] n discharge (t)≤n discharge-limit (10)

[0108] Where, n charge (t) and n discharge (t) represents the number of times the battery is charged and discharged at time t, respectively, and n charge-limit and n discharge-limit These are the upper limits of the number of times the battery can be charged and discharged.

[0109] 7. Other constraints: Based on the safety requirements and policies and regulations in actual operation, other relevant constraints shall be determined, such as the charging and discharging efficiency constraints of the energy storage system and the battery life constraints.

[0110] In other words, the embodiments of the present invention use a combination of multi-objective and multi-constraint random or intelligent optimization algorithms to formulate scientific energy storage regulation strategies. In contrast, existing technologies often focus on a single objective when formulating distributed energy storage regulation strategies, such as only pursuing the minimization of operating costs or only focusing on the renewable energy consumption rate. Moreover, the constraints are not comprehensive enough, and the selection of optimization algorithms is also relatively limited, often using traditional linear programming algorithms, which are difficult to adapt to complex and ever-changing actual operating scenarios.

[0111] Step S320: Based on the set multiple optimization objectives and multiple constraints, an optimization algorithm is used to determine the distributed energy storage control strategy, wherein the distributed energy storage control strategy includes the optimal charging and discharging strategy and the power allocation strategy.

[0112] For example, the optimization algorithm may employ a stochastic optimization algorithm, such as stochastic programming. Let the random variable ξ represent a random factor (such as market electricity price, renewable energy power generation, etc.), and its probability distribution be P(ξ). The optimization problem can be expressed as:

[0113]

[0114] stg(x,ξ)≤0

[0115] Where x is the decision variable (such as the charging and discharging strategy of the energy storage system, power allocation scheme, etc.), F(x,ξ) is the objective function (such as the operating cost function), and g(x,ξ) is the constraint function.

[0116] By combining generated random scenarios with defined optimization objectives and constraints, stochastic simulation methods are used to perform multiple sampling calculations on the objective and constraint functions. Then, mathematical programming algorithms (such as linear programming and nonlinear programming) are employed to solve for the optimal charging and discharging strategies and power allocation schemes of the distributed energy storage system under different scenarios. Alternatively, stochastic dynamic programming algorithms can be used. By defining state variables, decision variables, state transition equations, and stage index functions, the optimal decision is selected at each stage based on the current state and the values ​​of random factors, progressively solving for the optimal strategy throughout the entire planning period. Furthermore, intelligent optimization algorithms, such as genetic algorithms, can be used. First, the decision variables are encoded (e.g., binary encoding) to generate an initial population. Then, the fitness value of each individual in the population is calculated (based on the optimization objective function). Through selection, crossover, and mutation operations, a new population is generated. This process is repeated continuously to improve the fitness value of the population, ultimately yielding a control strategy that meets actual needs. The particle swarm optimization algorithm is similar. It initializes a group of particles (each particle representing a possible solution), and continuously adjusts the speed and position of the particles based on their fitness values ​​(also calculated according to the optimization objective function) to eventually find the optimal solution.

[0117] Therefore, through the above examples, this embodiment of the invention proposes a complete process for the precise control of distributed energy storage systems, including stochastic factor analysis and modeling, battery aging cost assessment using neural network simulation, and the formulation of distributed energy storage control strategies. Each sub-process has the following advantages compared to existing technologies:

[0118] 1. Existing technologies rely on limited methods for analyzing stochastic factors such as market electricity prices, renewable energy generation capacity, and energy storage vehicle charging demand. This invention addresses the challenge of comprehensively utilizing multiple algorithms to accurately analyze and simulate stochastic factors, providing more realistic data support for energy storage regulation. Specifically, this invention comprehensively considers various factors to improve the stability, economy, and renewable energy absorption capacity of the power system. It further utilizes the ARIMA model, kernel density estimation method, and cluster analysis to determine the probability distribution of data such as market electricity prices and renewable energy generation capacity, and uses Monte Carlo simulation to generate stochastic scenarios, providing data support for energy storage regulation.

[0119] 2. To address the problem that existing technologies cannot accurately reflect the complex physicochemical changes and multi-factor coupling effects of battery aging, leading to large errors in battery aging cost assessment, this invention aims to overcome the challenge of constructing an accurate battery aging cost assessment model by combining empirical knowledge with neural network simulation technology. Specifically, this invention constructs an experience-based aging model and combines it with neural network simulation methods to accurately assess aging costs, providing a basis for energy storage cost accounting.

[0120] 3. Existing technologies, employing single optimization objectives and traditional, limited optimization algorithms, struggle to balance operating costs, system reliability, and renewable energy absorption rates. This invention addresses how to formulate scientific energy storage control strategies using a combination of multi-objective, multi-constraint, stochastic or intelligent optimization algorithms. This enables scientific control and optimal resource allocation of energy storage systems, thereby improving power system stability, economy, and renewable energy absorption capacity, and promoting the efficient application of distributed energy storage in power systems. Specifically, with the objectives of minimizing operating costs, maximizing reliability, and maximizing renewable energy absorption rates, and considering various constraints, stochastic or intelligent optimization algorithms are used to solve for the optimal charging / discharging strategy and power allocation scheme.

[0121] Thus, the embodiments of the present invention effectively realize the scientific regulation and resource optimization of energy storage systems, comprehensively consider random factors and battery aging costs simulated by neural networks, promote the efficient application and precise regulation of distributed energy storage in power systems, and have important practical value and strategic significance for promoting sustainable energy development and improving the overall performance of power systems.

[0122] Example 2

[0123] In Embodiment 2, regarding step S200, while processing the battery operating parameters related to distributed energy storage regulation based on the neural network model, the distributed energy storage regulation method may further include: processing the battery operating parameters based on an empirical model, and combining the processing result with the processing result of the neural network model to estimate the battery aging cost, wherein the empirical model is constructed based on battery operating experience data and outputs the battery capacity decay rate.

[0124] That is, an empirical model is combined with a neural network model to estimate battery aging costs. The empirical model is a capacity decay model based on empirical data, and its construction process is as follows:

[0125] Assuming that battery capacity degradation is related to factors such as the number of charge-discharge cycles (n), current (I), and temperature (T), an empirical model can be constructed as follows:

[0126]

[0127] Where a, b, and c are model parameters, E a R is the activation energy, R is the gas constant, and T is the absolute temperature.

[0128] This empirical model can roughly describe the relationship between battery capacity decay and some key factors at a macroscopic level, providing a basic theoretical framework for battery aging analysis. However, the battery aging process is a complex nonlinear process, influenced by a combination of factors, and these factors may have complex coupling relationships. Empirical models are often based on simplified assumptions and are difficult to accurately capture the subtle changes and complex dynamic characteristics in the battery aging process. Therefore, in order to more accurately simulate and predict the battery aging process, Embodiment 2 of this invention further introduces the simulation and prediction based on the neural network model in Embodiment 1. Neural networks have powerful nonlinear mapping capabilities and self-learning capabilities, and can automatically mine the complex relationships between data from a large amount of historical data, thereby providing a more accurate model of the battery aging process. That is, Embodiment 2 of this application combines empirical knowledge with neural network simulation technology to construct an accurate battery aging cost assessment model, providing a more accurate basis for energy storage cost accounting.

[0129] Example 3

[0130] Figure 5 This is a schematic diagram of the structure of a distributed energy storage and control device according to another embodiment of the present invention. Figure 5 As shown, the power prediction device includes: a memory storing a program that can run on a processor; and the processor configured to implement the distributed energy storage control method described in the above embodiments when executing the program.

[0131] In a specific implementation, the method steps involved in the above embodiments can be stored in a memory in the form of program units, and the processor executes the program units stored in the memory to realize the corresponding functions.

[0132] The processor contains a kernel, from which corresponding program units are retrieved from memory. One or more kernels can be configured, and the power prediction methods of the various embodiments described above can be implemented by adjusting the kernel parameters.

[0133] The memory may include non-permanent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.

[0134] In the example, the distributed energy storage control device of this embodiment can be a server, PC, PAD, mobile phone, etc.

[0135] Other embodiments of the present invention also provide a distributed energy storage system, including the distributed energy storage control device described in the above embodiments.

[0136] Other embodiments of the present invention also provide a processor for running a program, wherein the program executes the distributed energy storage regulation method described in the above embodiments.

[0137] Other embodiments of the present invention also provide a computer program product, which, when executed on a data processing device, is adapted to perform the steps of initializing the distributed energy storage control method of the above embodiments.

[0138] Other embodiments of the present invention also provide a machine-readable storage medium storing instructions that cause a machine to execute the distributed energy storage control method described in the above embodiments.

[0139] Those skilled in the art will appreciate that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0140] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0141] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0142] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0143] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0144] Memory may include non-persistent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0145] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0146] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0147] The above are merely embodiments of this application and are not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A distributed energy storage regulation method, characterized in that, include: Data processing is performed on various random factors related to distributed energy storage regulation based on multi-algorithm collaboration to generate random scenarios for distributed energy storage regulation. The battery operating parameters associated with distributed energy storage regulation are processed based on a neural network model to estimate the battery aging cost, wherein the neural network model outputs the battery capacity decay rate. as well as Based on the generated random scenarios and the estimated battery aging costs, a distributed energy storage regulation strategy is determined.

2. The distributed energy storage regulation method according to claim 1, characterized in that, The data processing based on multi-algorithm collaboration for various random factors related to distributed energy storage regulation includes: Collect and preprocess the various random factors; A probability distribution estimation algorithm is used to estimate the probability distributions of the various preprocessed random factors to obtain the probability distributions corresponding to each random factor; and Based on the probability distributions of various random factors obtained, a probabilistic statistical algorithm is used to generate the random scenarios for distributed energy storage regulation.

3. The distributed energy storage regulation method according to claim 2, characterized in that, The various random factors include any number of the following: market electricity price fluctuation data, renewable energy power generation data, energy storage vehicle charging demand data, and charging / swapping load data.

4. The distributed energy storage regulation method according to claim 3, characterized in that, The probability distribution estimation algorithm used for the market electricity price fluctuation data is the Autoregressive Integrated Moving Average (ARIMA) model in time series analysis algorithms; the probability distribution estimation algorithm used for the renewable energy power generation data is the kernel density estimation algorithm; and the probability distribution estimation algorithm used for the energy storage vehicle charging demand data and the charging and swapping load data is the clustering algorithm. Furthermore, the probability and statistics algorithm employs the Monte Carlo simulation algorithm.

5. The distributed energy storage regulation method according to claim 1, characterized in that, The battery operating parameters related to distributed energy storage regulation processed using the neural network model include: Obtain the battery operating parameters; The neural network model is trained using historical battery operating data, wherein the acquired battery operating parameters serve as input to the neural network model; and The battery aging cost is estimated based on the battery capacity degradation rate output by the neural network model.

6. The distributed energy storage regulation method according to claim 4, characterized in that, The neural network model is a Long Short-Term Memory (LSTM) network model.

7. The distributed energy storage regulation method according to claim 1, characterized in that, While processing battery operating parameters related to distributed energy storage regulation based on a neural network model, the distributed energy storage regulation method also includes: The battery operating parameters are processed based on an empirical model, and the processing results are combined with the processing results of the neural network model to estimate the battery aging cost, wherein the empirical model is constructed based on battery operating experience data and outputs the battery capacity decay rate.

8. The distributed energy storage regulation method according to claim 1, characterized in that, The determination of the distributed energy storage control strategy includes: Based on the aforementioned random scenario and battery aging cost, multiple optimization objectives and constraints are set for distributed energy storage regulation; and Based on multiple optimization objectives and constraints, an optimization algorithm is used to determine the distributed energy storage regulation strategy, which includes an optimal charging and discharging strategy and a power allocation strategy.

9. The distributed energy storage regulation method according to claim 8, characterized in that, The multiple optimization objectives include minimizing operating costs, maximizing the reliability of distributed energy storage systems, and maximizing renewable energy integration rates; and / or The multiple constraints include any number of the following for the energy storage system: power balance constraint, voltage constraint, frequency constraint, charge / discharge power constraint, battery capacity constraint, charge / discharge cycle limit, charge / discharge efficiency constraint, and battery life constraint.

10. A distributed energy storage and control device, characterized in that, include: Memory, which stores programs that can run on a processor; as well as The processor is configured to implement the distributed energy storage control method according to any one of claims 1 to 9 when executing the program.

11. A distributed energy storage system, characterized in that, Includes the distributed energy storage and regulation device as described in claim 10.

12. A machine-readable storage medium, characterized in that, The machine-readable storage medium stores instructions for causing the machine to perform the distributed energy storage control method according to any one of claims 1 to 9.

Images