Energy storage scheduling method and system based on time series load prediction

Abstract

This invention belongs to the field of power system dispatching and resource management technology, specifically relating to a method and system for energy storage dispatching based on time-series load forecasting. The method includes: acquiring power system operating status information; predicting the operating load of the power system based on the acquired status information and a preset time-series load forecasting model, obtaining a predicted operating load value; and performing energy storage dispatching based on the obtained predicted operating load value, thus completing energy storage dispatching based on time-series load forecasting. The preset time-series load forecasting model employs an improved... Prophet The algorithm determines timestamps based on the time sequence in the acquired state information, identifies time-series load influencing factors based on the timestamps, and determines a time-series load prediction model based on the identified time-series load influencing factors.

Description

A method and system for energy storage dispatch based on time series load forecasting Technical Field

[0001] This invention belongs to the field of power system dispatching and resource management technology, specifically relating to an energy storage dispatching method and system based on time series load forecasting. Background Technology

[0002] The statements in this section are merely background information related to the present invention and do not necessarily constitute prior art.

[0003] The importance of energy storage dispatch in power systems is becoming increasingly prominent, especially given the rising proportion of renewable energy sources. Energy storage systems, as flexible devices capable of storing and releasing energy, play a crucial role. First, with the continuous increase in the proportion of renewable energy, power systems face daily and seasonal load fluctuations. Energy storage systems can effectively store excess energy for release during peak demand periods, thereby balancing supply and demand on both the supply and demand sides and improving the stability of the power system. Second, traditional power generation equipment struggles to achieve instantaneous start-up and shutdown, while the rapid response characteristics of energy storage systems enable instantaneous charging and discharging, adapting to the transient demands of the power system.

[0004] In existing technologies, traditional rule-based and empirically driven dispatching methods struggle to cope with the increasing complexity of power systems. Based on static rules, they cannot adapt to dynamic and nonlinear load and energy fluctuations within power systems. Furthermore, they often fail to effectively integrate user-side load forecasting and dispatch optimization, resulting in an inability to maximize overall system benefits.

[0005] However, current energy storage dispatch methods are still constrained by a series of limitations: traditional rule-based and empirically driven dispatch methods struggle to cope with the increasing complexity of power systems, and their static rules fail to adapt to dynamic and nonlinear load and energy fluctuations. Secondly, while mathematical optimization models improve dispatch accuracy to some extent, their response to spatiotemporal dynamic changes in the power system remains insufficient, resulting in poor performance of traditional dispatch methods in dealing with renewable energy fluctuations, user load changes, and market price uncertainties. Furthermore, existing methods often fail to effectively integrate user-side load forecasting and dispatch optimization, leading to a failure to maximize overall system benefits. Therefore, the limitations of traditional energy storage dispatch methods urgently need to be addressed through innovative approaches to drive power system dispatch towards a more flexible, intelligent, and sustainable direction. Summary of the Invention

[0006] To address the aforementioned issues, this invention proposes an energy storage dispatching method and system based on time-series load forecasting. By introducing load forecasting combined with time series data, it achieves accurate, real-time, and adaptive optimization of energy storage dispatching, thereby improving the dispatching efficiency of the energy storage system while reducing the operating costs of the power system.

[0007] According to some embodiments, the first aspect of the present invention provides an energy storage scheduling method based on time series load forecasting, employing the following technical solution:

[0008] An energy storage dispatch method based on time-series load forecasting includes:

[0009] Obtain power system operating status information;

[0010] Based on the acquired state information and the preset time series load forecasting model, the operating load of the power system is predicted, and the predicted operating load value is obtained.

[0011] Energy storage scheduling is carried out based on the obtained operating load forecast values ​​to complete energy storage scheduling based on time series load forecasts;

[0012] The preset time series load forecasting model uses an improved Prophet algorithm to determine timestamps based on the time sequence in the acquired state information, determine the time series load influencing factors based on the timestamps, and determine the time series load forecasting model based on the determined time series load influencing factors.

[0013] As a further technical limitation, the time series load forecasting model is as follows: ;in, For time series load forecasting models; This is a unit used to capture the trend of nonlinear growth of factors influencing time-series loads. , , , and All of these are hyperparameters. For the nonlinear trend of the load, This is a linear trend term used to capture changing load demand; As a seasonal unit used to capture the periodicity of factors influencing time-series loads, , These are the serial numbers of the seasonal components. The total amount of seasonal components, For the contribution of the i-th seasonal component at time t, The serial number of the periodic component related to external factors. The total number of periodic components of external factors. It is a periodic function related to external factors. The influence weights of the periodic function; This is a holiday unit. , This represents the total number of holidays. The effect of the i-th holiday at time t; To capture the error term of random variations in the factors influencing time-series loads, , This is the error term from the previous time step. These are the autoregressive parameters of the error term. This is white noise error.

[0014] Furthermore, the parameters of the time series load forecasting model are optimized by maximizing the log-likelihood estimate. During the parameter optimization process, the parameters of the trend term, seasonal unit, and holiday unit are adjusted by gradient descent. This process is repeated iteratively until the prediction error of the time series load forecasting model is minimized.

[0015] As a further technical limitation, in the process of energy storage scheduling based on the obtained operating load forecast, an improved particle swarm optimization algorithm is used to simulate the group behavior of birds in a flock, continuously adjust the particle positions to find the optimal solution, obtain the optimal operating load forecast, and complete the optimized scheduling of energy storage based on the obtained optimal operating load forecast.

[0016] Furthermore, in the process of energy storage optimization scheduling, the velocity and position of the initial particle swarm are generated by chaotic mapping. The inertial weight is adjusted based on the adaptive weight and dynamic parameter adjustment mechanism. The velocity and position of the particles are updated iteratively. Through continuous iteration, fuzzy logic control is used to prevent getting trapped in local optima, so as to obtain the best energy storage scheduling of particle swarm search, that is, the globally optimal energy storage scheduling strategy, and determine the charging and discharging operation of the energy storage system in each time period.

[0017] As a further technical limitation, the acquired power system operation status information shall include at least historical electricity consumption data, market electricity price information, meteorological data, holiday electricity consumption data, and electricity consumption data for major events.

[0018] According to some embodiments, a second aspect of the present invention provides an energy storage dispatching system based on time-series load forecasting, employing the following technical solution:

[0019] An energy storage dispatch system based on time-series load forecasting includes:

[0020] The acquisition module is configured to acquire power system operating status information;

[0021] The prediction module is configured to predict the operating load of the power system based on the acquired state information and a preset time series load prediction model, and obtain the predicted value of the operating load.

[0022] The scheduling module is configured to perform energy storage scheduling based on the obtained operating load forecast values, thereby completing energy storage scheduling based on time series load forecasts.

[0023] The preset time series load forecasting model uses an improved Prophet algorithm to determine timestamps based on the time sequence in the acquired state information, determine the time series load influencing factors based on the timestamps, and determine the time series load forecasting model based on the determined time series load influencing factors.

[0024] According to some embodiments, a third aspect of the present invention provides a computer-readable storage medium, employing the following technical solution:

[0025] A computer-readable storage medium having a program stored thereon, which, when executed by a processor, implements the steps of an energy storage scheduling method based on time-series load forecasting as described in the first aspect of the present invention.

[0026] According to some embodiments, the fourth aspect of the present invention provides an electronic device, which adopts the following technical solution:

[0027] An electronic device includes a memory, a processor, and a program stored in the memory and running on the processor. When the processor executes the program, it implements the steps in the energy storage scheduling method based on time series load forecasting as described in the first aspect of the present invention.

[0028] According to some embodiments, the fifth aspect of the present invention provides a computer program product, which adopts the following technical solution:

[0029] A computer program product includes software code, wherein the program in the software code performs the steps of an energy storage scheduling method based on time series load forecasting as described in the first aspect of the present invention.

[0030] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0031] This invention introduces a time-series load forecasting method to achieve precise, real-time, and adaptive optimization of energy storage dispatch. It accurately predicts user-side load and uses this prediction as the basis for energy supply-side dispatch, enabling precise dispatch of the energy storage system to maximize electricity demand satisfaction and improve system operating efficiency. It supports the development of optimal functional strategies under different market conditions, making energy utilization more economical and sustainable, and reducing operating costs. Through real-time monitoring and dispatch adjustment, it can flexibly respond to fluctuations in electricity market prices. It has significant advantages in improving power system efficiency, reducing operating costs, and promoting intelligentization. By comprehensively considering both users and the energy supply side and introducing advanced technologies, it provides innovative solutions for the future development of the power system. Attached Figure Description

[0032] The accompanying drawings, which form part of this embodiment, are used to provide a further understanding of this embodiment. The illustrative embodiments and their descriptions are used to explain this embodiment and do not constitute an improper limitation of this embodiment.

[0033] Figure 1 is a flowchart of the energy storage scheduling method based on time series load forecasting in Embodiment 1 of the present invention;

[0034] Figure 2 is an architecture diagram of the energy storage scheduling method based on time series load forecasting in Embodiment 1 of the present invention;

[0035] Figure 3 is a structural block diagram of the energy storage scheduling system based on time series load prediction in Embodiment 2 of the present invention. Detailed Implementation

[0036] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0037] It should be noted that the following detailed descriptions are exemplary and intended to provide further explanation of this application. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains.

[0038] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of exemplary embodiments according to the invention. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0039] In this invention, terms such as "upper," "lower," "left," "right," "front," "back," "vertical," "horizontal," "side," and "bottom" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. These terms are used only to facilitate the description of the structural relationships of the various components or elements of this invention and do not specifically refer to any component or element in this invention. They should not be construed as limiting the invention.

[0040] In this invention, terms such as "fixed connection," "connected," and "linked" should be interpreted broadly, indicating a fixed connection, an integral connection, or a detachable connection; a direct connection or an indirect connection through an intermediate medium. Those skilled in the art can determine the specific meaning of these terms in this invention based on the specific circumstances, and they should not be construed as limitations on the invention.

[0041] Where there is no conflict, the embodiments and features in the embodiments of the present invention can be combined with each other.

[0042] Example 1

[0043] Embodiment 1 of this invention introduces an energy storage scheduling method based on time series load forecasting.

[0044] Figure 1 illustrates an energy storage dispatch method based on time-series load forecasting, comprising:

[0045] Obtain power system operating status information;

[0046] Based on the acquired state information and the preset time series load forecasting model, the operating load of the power system is predicted, and the predicted operating load value is obtained.

[0047] Energy storage scheduling is carried out based on the obtained operating load forecast values ​​to complete energy storage scheduling based on time series load forecasts;

[0048] The preset time series load forecasting model uses an improved Prophet algorithm to determine timestamps based on the time sequence in the acquired state information, determine the time series load influencing factors based on the timestamps, and determine the time series load forecasting model based on the determined time series load influencing factors.

[0049] This embodiment acquires power system operating status information through monitoring systems, sensors, or historical records, including:

[0050] (1) Historical electricity consumption data: Obtain users' hourly or minute-by-minute electricity consumption data through monitoring systems or sensors;

[0051] (2) Market electricity price information: Considering the impact of peak and off-peak periods on electricity load, market electricity price is used as a predictable covariate in the model input;

[0052] (3) Meteorological data: Obtain meteorological data that are significantly correlated with electricity load, including temperature, humidity, etc.;

[0053] (4) Holiday and Special Event Data: Considering the impact of holidays and major events on electricity consumption, date data is obtained as static covariates for model input. Time series data is represented as follows: ,in, This represents the user's electricity consumption data. This refers to the other covariates mentioned above.

[0054] After obtaining the power system operating status information, this embodiment needs to perform data preprocessing, which mainly includes data cleaning and resampling. Interpolation is used to handle missing and outlier values. If there are irregular time intervals, resampling is required to ensure that the time series has the same time interval.

[0055] The Prophet algorithm is a time series-based forecasting algorithm that is well adapted to the periodic changes and trends of data. Furthermore, the algorithm can effectively integrate holiday information that has a significant impact on electricity consumption, greatly improving the accuracy of forecasts.

[0056] In this embodiment, an improved Prophet algorithm is employed to capture the main patterns of load changes by fitting trends, seasonality, and holidays in historical time series data. The time series data is represented as a series of electricity consumption data points arranged chronologically, each data point including a timestamp, the corresponding electricity consumption, and other relevant covariates. Features such as daytime and seasonal variations, as well as electricity consumption fluctuations during specific holidays, can be extracted from these timestamps. The correlation between time series and load forecasting lies in the fact that by observing and modeling patterns in historical data, the improved Prophet algorithm can extrapolate and predict future loads based on the historical characteristics of the time series, enabling the model to capture the dynamic changes in electricity consumption over time. The trend component is used to simulate long-term changes, the seasonality component to capture periodic fluctuations, and the holiday component to handle the impact of specific time points.

[0057] The time series load forecasting model used in this embodiment is: Specifically:

[0058] (1) Trend Unit

[0059] This is a trend unit used to capture the nonlinear growth of factors influencing time-series loads, i.e. ;

[0060] in, , , and All of these are hyperparameters. For the nonlinear trend of the load, This is a linear trend term used to capture changing load demand;

[0061] (2) Seasonal Units

[0062] As a seasonal unit used to capture the periodicity of factors influencing time-series loads, i.e. ;

[0063] in, This is the sequence number of the seasonal component, used to iterate through all periodic components; This represents the total quantity of seasonal components. This embodiment takes into account the periodicity of years, months, and weeks. =3; This represents the contribution of the i-th seasonal component at time t. The annual cycle affects the electricity load in winter and summer, while the weekly cycle reflects the difference in electricity consumption between weekdays and weekends. This indicates the index of the periodic component related to external factors, used to traverse the periodic components of all external factors; This represents the total number of periodic components of external factors. In this embodiment, periodic changes related to temperature, weather, and electricity demand are considered. It is a periodic function related to external factors. The influence weights of the periodic function;

[0064] (3) Holiday Unit

[0065] It is a holiday unit, that is ;

[0066] in, This represents the total number of holidays, including Spring Festival and National Day, each of which has a significant impact on electricity demand. This represents the effect of the i-th holiday at time t, used to describe the impact of a specific holiday on electricity load;

[0067] (4) Error Term

[0068] This is the error term used to capture the random variations in the factors influencing time-series loads, i.e. ;

[0069] in, For the error term from the previous time step, i.e., the autocorrelation component in the autoregressive model, we introduce... The purpose is to capture the short-term fluctuation characteristics in time series data, so that the model can more accurately predict short-term load changes; These are the autoregressive parameters of the error term. This is white noise error.

[0070] In the training process of the time series load forecasting model, this embodiment optimizes the model parameters by maximizing the log-likelihood estimate to ensure a good fit to the time series data. During optimization, gradient descent is used to adjust the parameters for trend, seasonality, and holiday effects. The loss function is defined as the sum of squared prediction errors. ,in, This is the actual load value. These are the model's predicted values. For all model parameters. For the loss function. By taking the derivative, we obtain the gradient of each parameter, which represents the rate of change of the loss function with respect to each parameter: According to the learning rate Update each parameter and adjust the formula as follows: The above process is iterated continuously until the prediction error is minimized. For example, for the parameters in the trend unit... , The same applies to other parameters.

[0071] This embodiment uses cross-validation to evaluate the model and verify its performance on unseen data. 80% of the data is used for training, and the remaining 20% ​​is used for validation to ensure the model has good generalization ability.

[0072] In this embodiment, during the prediction phase, preprocessed future influencing factors (such as meteorological data, holidays, etc.) are input into the trained model. The improved Prophet algorithm combines these influencing factors and predicts future load values ​​by extrapolating trends and seasonality. During the prediction process, the characteristics of the time series are reflected by capturing the time dependence of historical data and the correlation of load changes. The improved Prophet algorithm utilizes timestamps in the time series to identify change patterns at different time scales such as daytime, weektime, and monthtime, thereby inferring future load changes. For example, by learning load patterns from different seasons and times in past electricity consumption data, the model can extrapolate similar patterns in predictions, thus improving prediction accuracy. Time dependence ensures that the model can predict the future based on past load change patterns, while the correlation between load and time allows the model to capture the dynamic changes in electricity consumption over time.

[0073] It should be noted that the load forecast values ​​for future time periods obtained in this embodiment include the load values ​​for each future time step and provide confidence intervals to help decision-makers assess the reliability of the forecasts and serve as the basis for subsequent scheduling strategies.

[0074] In this embodiment, the optimization problem during the scheduling process is defined as minimizing system operating costs while simultaneously satisfying electricity market demand and energy inventory constraints. This optimization problem can be expressed as:

[0075]

[0076]

[0077]

[0078]

[0079]

[0080]

[0081]

[0082] in, Power purchased for the power grid To improve the charging and discharging efficiency of energy storage systems, To meet load demand, and The unit cost of purchasing electricity from the grid and charging / discharging energy storage systems. and The maximum power limit for grid power purchase and energy storage system capacity. For the SOC of the energy storage system, To improve the charging efficiency of energy storage systems. For time step, and These are the maximum and minimum limits for the State of Charge (SOC) of an energy storage system.

[0083] This embodiment uses an improved particle swarm optimization algorithm as the scheduling optimization algorithm. The particle swarm optimization algorithm simulates the group behavior of birds in a flock of birds. By continuously adjusting the position of particles (i.e. potential solutions), it seeks the optimal solution and can more effectively adapt to the complex scenario of energy storage scheduling optimization.

[0084] The improved algorithm is as follows:

[0085] (1) Chaotic mapping initialization

[0086] In the particle swarm initialization phase, a logistic mapping is used to generate initial positions and velocities. Leveraging the ergodic and random characteristics of the logistic mapping, the diversity of the particle swarm can be effectively enhanced, preventing it from getting trapped in local optima; that is... ;in, To control the parameters, after multiple rounds of experimental verification, the hyperparameter value of 3.82 was found to be optimal for the system's chaotic behavior.

[0087] The volatility of electricity demand and market prices is uncertain, therefore the diversity of initial solutions in particle swarm optimization directly affects the ability to search for the globally optimal energy storage scheduling strategy. Generating initial particle positions through chaotic mapping can more effectively cover various possible scheduling strategies, thereby increasing the probability of finding the globally optimal solution.

[0088] For example, in an energy storage dispatch system, the initial position represents the charging and discharging strategy of the energy storage device at different time periods in the future. Diverse initial schemes generated through chaotic mapping ensure that the system can explore multiple different dispatch strategies in the early stages, thereby adapting to fluctuations in electricity demand and uncertainties in market prices, and increasing the effectiveness and reliability of the dispatch strategy.

[0089] (2) Add adaptive weights and dynamic parameter adjustment

[0090] In energy storage dispatching systems, the volatility of power load and the dynamic changes in the market environment necessitate flexible dispatching strategies. In the original particle swarm optimization algorithm, inertia weights... and learning factors Typically, fixed weights are difficult to adapt to the complexity of the scheduling environment. The improved algorithm, through adaptive weights and dynamic parameter adjustment mechanisms, enables effective adaptive adjustments at different search stages, thereby enhancing the efficiency and flexibility of scheduling optimization.

[0091] Inertia weight Dynamic adjustments are made at different iteration stages to balance the weights between global and local searches:

[0092] ;

[0093] Use a larger inertia weight in the early stages of the search. To enhance overall search capabilities; gradually reduce weight in the later stages of the search. This enhances the local development capability, helping particles to converge more precisely to the optimal solution.

[0094] Learning factor The learning factor controls how closely the particle follows its historical best position and global best position. In the improved algorithm, the learning factor is adaptively adjusted according to the current state of the search to enhance the algorithm's adaptability at different stages.

[0095] Initial Exploration Phase: Increasing Individual Learning Factors This enhances the particle's ability to follow its own historical best position. This encourages particles to independently explore their own optimal solutions, increases the coverage of the search space, and prevents the particle swarm from getting trapped in local optima early on.

[0096] The formula for adjusting individual learning factors:

[0097] ;

[0098] in, and These represent the maximum and minimum values ​​of the individual learning factor, respectively.

[0099] Later convergence phase: Increase the group learning factor To enhance the particle's tracking of the global optimal position, it can help the particle converge to the global optimal solution quickly, thereby improving the convergence speed of the algorithm.

[0100] The formula for adjusting the group learning factor is:

[0101] ;

[0102] in, and These are the minimum and maximum values ​​of the group learning factor, respectively.

[0103] Through such adaptive adjustments, the behavior of particles can be more flexible in different search stages, achieving a balance between exploration and development. This ensures that in energy storage scheduling scenarios, strategies that maximize economic benefits can be formulated while improving the stability and accuracy of the search.

[0104] (3) Introduce a fuzzy logic controller

[0105] A fuzzy logic controller is introduced to supplement the aforementioned adaptive weights and learning factors. Unlike the adaptive weights and dynamic learning factors, which primarily depend on the number of iterations and particle states, the introduced fuzzy logic controller adjusts parameters based on the global trend of the algorithm's operation, monitoring the dynamic characteristics of the particle swarm (mainly including the degree of particle aggregation and convergence speed). This adjustment is based on real-time state feedback. For example, when excessive particle concentration is detected, the inertia weight is increased or the learning factor is decreased to enhance the exploration capability and prevent getting trapped in local optima.

[0106] When the particle cluster concentration is high, increase the inertia weight. To enhance exploratory capabilities; if the particle swarm convergence rate is low, increase the global learning factor. This ensures that the particle swarm converges to the global optimum as quickly as possible; in each iteration, the particle velocity and position updates can be represented as: ; ;

[0107] in, This is the inertia weight, used to balance the effects of current speed and historical speed. and The learning factor controls the degree to which a particle follows its own historical best position and its global best position, respectively. , , These values ​​are all further corrected by a fuzzy logic controller based on adaptive adjustment. The particle velocity reflects the magnitude of changes in the scheduling strategy. For particles In time The position indicates the charging and discharging strategy of the energy storage system, namely the charging and discharging power and time period. and The random number is between 0 and 1. The purpose is to introduce randomness and increase the particle swarm's ability to explore unknown spaces, that is, to increase the range of strategies to choose from. The historical best position represents the optimal scheduling scheme found by the particle. The globally optimal position represents the optimal energy storage scheduling strategy found by the entire particle swarm.

[0108] This embodiment utilizes chaotic mapping to generate the initial particle swarm's velocity and position, setting the chaotic parameter r=3.82, and letting the initial position... Randomly generated within the range [0,1]; the initial particle positions and velocities are generated using the Logistic mapping formula, representing possible scheduling schemes, including the charging and discharging time periods and power levels. The maximum and minimum values ​​of the inertial weights are set as follows: , Iteratively update the particle velocity and position; through continuous iteration, the particle swarm searches for the optimal energy storage scheduling strategy to minimize system operating costs and maximize economic benefits; output the globally optimal position. This refers to the globally optimal energy storage scheduling strategy, which clarifies the charging and discharging operations of the energy storage system in each time period.

[0109] This embodiment introduces a time-series load forecasting method to achieve precise, real-time, and adaptive optimization of energy storage dispatch. It accurately predicts user-side load and uses this as the basis for energy supply-side dispatch, enabling precise dispatch of the energy storage system to maximize electricity demand satisfaction and improve system operating efficiency. It supports the development of optimal functional strategies under different market conditions, making energy utilization more economical and sustainable, and reducing operating costs. Through real-time monitoring and dispatch adjustment, it can flexibly respond to fluctuations in electricity market prices. It has significant advantages in improving power system efficiency, reducing operating costs, and promoting intelligentization. By comprehensively considering both users and the energy supply side and introducing advanced technologies, it provides innovative solutions for the future development of the power system.

[0110] Example 2

[0111] Embodiment 2 of the present invention introduces an energy storage scheduling system based on time series load forecasting.

[0112] Figure 3 shows an energy storage dispatch system based on time-series load forecasting, which includes:

[0113] The acquisition module is configured to acquire power system operating status information;

[0114] The prediction module is configured to predict the operating load of the power system based on the acquired state information and a preset time series load prediction model, and obtain the predicted value of the operating load.

[0115] The scheduling module is configured to perform energy storage scheduling based on the obtained operating load forecast values, thereby completing energy storage scheduling based on time series load forecasts.

[0116] The preset time series load forecasting model uses an improved Prophet algorithm to determine timestamps based on the time sequence in the acquired state information, determine the time series load influencing factors based on the timestamps, and determine the time series load forecasting model based on the determined time series load influencing factors.

[0117] The detailed steps are the same as those of the energy storage scheduling method based on time series load forecasting provided in Example 1, and will not be repeated here.

[0118] Example 3

[0119] Embodiment 3 of the present invention provides a computer-readable storage medium.

[0120] A computer-readable storage medium having a program stored thereon, which, when executed by a processor, implements the steps of an energy storage scheduling method based on time-series load forecasting as described in Embodiment 1 of the present invention.

[0121] The detailed steps are the same as those of the energy storage scheduling method based on time series load forecasting provided in Example 1, and will not be repeated here.

[0122] Example 4

[0123] Embodiment 4 of the present invention provides an electronic device.

[0124] An electronic device includes a memory, a processor, and a program stored in the memory and running on the processor. When the processor executes the program, it implements the steps in the energy storage scheduling method based on time series load forecasting as described in Embodiment 1 of the present invention.

[0125] The detailed steps are the same as those of the energy storage scheduling method based on time series load forecasting provided in Example 1, and will not be repeated here.

[0126] Example 5

[0127] Embodiment 5 of the present invention provides a computer program product.

[0128] A computer program product includes software code, wherein the program in the software code performs the steps of an energy storage scheduling method based on time series load forecasting as described in Embodiment 1 of the present invention.

[0129] The detailed steps are the same as those of the energy storage scheduling method based on time series load forecasting provided in Example 1, and will not be repeated here.

[0130] The above description is merely a preferred embodiment of this practice and is not intended to limit the scope of this practice. Various modifications and variations can be made to this practice by those skilled in the art. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of this practice should be included within the protection scope of this practice.

Claims

1. A method for energy storage dispatching based on time-series load forecasting, characterized in that, include: Obtain power system operating status information; Based on the acquired state information and a pre-set time-series load forecasting model, the operating load of the power system is predicted, yielding the predicted operating load value. Energy storage scheduling is then performed based on this predicted value, completing the energy storage scheduling based on time-series load forecasting. The pre-set time-series load forecasting model employs an improved Prophet algorithm. Timestamps are determined based on the time sequence in the acquired state information, and time-series load influencing factors are identified based on these timestamps. An improved particle swarm optimization algorithm is used for scheduling optimization. During energy storage optimization scheduling, chaotic mapping is used to generate the initial particle swarm velocity and position. Inertial weights are adjusted based on adaptive weights and dynamic parameter adjustment mechanisms. The particle velocity and position are iteratively updated, and fuzzy logic control is used to prevent getting trapped in local optima, resulting in the optimal energy storage scheduling strategy obtained through particle swarm search. This globally optimal energy storage scheduling strategy determines the charging and discharging operations of the energy storage system in each time period. The inertial weights... The formula is: In the improved particle swarm optimization algorithm, the learning factor is adaptively adjusted according to the current search state. The adjustment formula for the individual learning factor is as follows: ， and Let be the maximum and minimum values ​​of the individual learning factor, respectively, and t be time; the adjustment formula for the group learning factor is... ， and These are the minimum and maximum values ​​of the group learning factor, respectively; the time series load prediction model is... ;in, For time series load forecasting models; This is a unit used to capture the trend of nonlinear growth of factors influencing time-series loads. ， 、 、 and All of these are hyperparameters. For the nonlinear trend of the load, This is a linear trend term used to capture changing load demand; As a seasonal unit used to capture the periodicity of factors influencing time-series loads, ， These are the serial numbers of the seasonal components. The total amount of seasonal components, For the contribution of the i-th seasonal component at time t, The serial number of the periodic component related to external factors. The total number of periodic components of external factors. It is a periodic function related to external factors. The influence weights of the periodic function; This is a holiday unit. ， This represents the total number of holidays. The effect of the i-th holiday at time t; To capture the error term of random variations in the factors influencing time-series loads, ， This is the error term from the previous time step. These are the autoregressive parameters of the error term. This is white noise error.

2. The energy storage dispatching method based on time series load forecasting as described in claim 1, characterized in that, The parameters of the time series load forecasting model are optimized by maximizing the log-likelihood estimate. During the parameter optimization process, the parameters of the trend term, seasonal unit, and holiday unit are adjusted by gradient descent. This process is repeated iteratively until the prediction error of the time series load forecasting model is minimized.

3. The energy storage dispatching method based on time series load forecasting as described in claim 1, characterized in that, In the process of energy storage scheduling based on the obtained operating load forecast, an improved particle swarm optimization algorithm is used to simulate the group behavior of birds in a flock, continuously adjust the particle positions to find the optimal solution, obtain the optimal operating load forecast, and complete the optimized scheduling of energy storage based on the obtained optimal operating load forecast.

4. The energy storage dispatching method based on time series load forecasting as described in claim 1, characterized in that, The acquired power system operation status information includes at least historical electricity consumption data, market electricity price information, meteorological data, holiday electricity consumption data, and electricity consumption data for major events.

5. An energy storage dispatch system based on time-series load forecasting, comprising the steps of an energy storage dispatch method based on time-series load forecasting as described in any one of claims 1-4, characterized in that, include: The acquisition module is configured to acquire power system operating status information; The prediction module is configured to predict the operating load of the power system based on the acquired state information and a preset time-series load prediction model, and obtain the predicted operating load value. The dispatch module is configured to perform energy storage dispatch based on the obtained operating load prediction value, and complete the energy storage dispatch based on time-series load prediction. The preset time-series load prediction model adopts an improved Prophet algorithm, determines the timestamp based on the time sequence in the acquired state information, determines the time-series load influencing factors based on the timestamp, and determines the time-series load prediction model based on the determined time-series load influencing factors.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the steps of an energy storage scheduling method based on time series load forecasting as described in any one of claims 1-4.

7. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the program, it implements the steps of an energy storage scheduling method based on time series load forecasting as described in any one of claims 1-4.

8. A computer program product, comprising software code, characterized in that, The program in the software code executes the steps of an energy storage scheduling method based on time series load forecasting as described in any one of claims 1-4.

Images