Multi-time-scale energy storage capacity configuration optimization method and system

Abstract

The invention discloses a multi-time-scale energy storage capacity configuration optimization method and system, and relates to the technical field of energy storage, and the method comprises the steps: carrying out the multi-time-scale load prediction according to a power system, and obtaining short-term, medium-term and long-term prediction loads; performing energy storage capacity configuration optimization on the energy storage system according to the predicted load to obtain corresponding short-term, medium-term and long-term energy storage capacity configuration strategies; and performing multi-time scale energy storage management on the energy storage system according to the short-term, medium-term and long-term energy storage capacity configuration strategies. According to the method, the technical problem that the configured energy storage capacity is difficult to accurately match the power system due to the fact that the change characteristics of the power system load under different time scales are not fully considered in a traditional energy storage capacity configuration method is solved, and accurate matching of the energy storage capacity and the power system load under multiple time scales is achieved; the utilization efficiency of the energy storage system is improved, and the technical effect of energy storage management and configuration of the power system is optimized.

Description

Technical Field

[0001] The present invention relates to the field of energy storage technology, and in particular to a multi-time-scale energy storage capacity configuration optimization method and system. Background Art

[0002] In the power system, the reasonable configuration of energy storage capacity is crucial to the stable operation of the system. Most traditional energy storage capacity configuration methods do not fully consider the changing characteristics of the power system load at different time scales. Existing technologies mainly configure energy storage capacity based on a single time scale or simple average load data, and often use fixed rules or empirical formulas for calculation. These methods have serious shortcomings when faced with the complex and changeable load fluctuations of the power system. Due to the lack of accurate analysis of the load differences at different time scales, the configured energy storage capacity is either too large, resulting in waste of resources and increased costs, or too small to effectively cope with load changes, making it difficult to ensure the stability and reliability of the power supply, and unable to meet the power system's needs for efficient management and optimized configuration of energy storage. Summary of the Invention

[0003] This application solves the technical problem that the traditional energy storage capacity configuration method does not fully consider the changing characteristics of the power system load at different time scales, resulting in the difficulty of accurately matching the configured energy storage capacity to the power system. This application obtains forecasted load data at different time scales by performing multi-time scale (short-term, medium-term, and long-term) load forecasting on the power system. Based on these data, the energy storage demand analysis, capacity allocation, and energy storage efficiency maximization of the energy storage system are performed to obtain the energy storage capacity configuration strategy for the corresponding time scale. Finally, these strategies are combined to perform multi-time scale energy storage management on the energy storage system, thereby accurately matching the energy storage demand of the power system at different time scales, optimizing the energy storage capacity configuration, effectively responding to load fluctuations, ensuring the stability and reliability of power supply, and realizing efficient management and optimized configuration of energy storage by the power system.

[0004] In response to the above technical problems, this application proposes a multi-time-scale energy storage capacity configuration optimization method and system.

[0005] In a first aspect, the present application provides a multi-time-scale energy storage capacity configuration optimization method, wherein the method includes: performing multi-time-scale load forecasting according to the power system to obtain short-term forecast load, medium-term forecast load and long-term forecast load; optimizing the energy storage capacity configuration of the energy storage system according to the short-term forecast load to obtain a short-term energy storage capacity configuration strategy; optimizing the energy storage capacity configuration of the energy storage system according to the medium-term forecast load to obtain a medium-term energy storage capacity configuration strategy; optimizing the energy storage capacity configuration of the energy storage system according to the long-term forecast load to obtain a long-term energy storage capacity configuration strategy; and performing multi-time-scale energy storage management of the energy storage system according to the short-term energy storage capacity configuration strategy, the medium-term energy storage capacity configuration strategy and the long-term energy storage capacity configuration strategy.

[0006] In a second aspect, the present application provides a multi-time-scale energy storage capacity configuration optimization system, wherein the system includes: a load forecasting module, which is used to perform multi-time-scale load forecasting based on the power system to obtain short-term forecast load, medium-term forecast load and long-term forecast load; a short-term strategy acquisition module, which is used to optimize the energy storage capacity configuration of the energy storage system according to the short-term forecast load to obtain a short-term energy storage capacity configuration strategy; a medium-term strategy acquisition module, which is used to optimize the energy storage capacity configuration of the energy storage system according to the medium-term forecast load to obtain a medium-term energy storage capacity configuration strategy; a long-term strategy acquisition module, which is used to optimize the energy storage capacity configuration of the energy storage system according to the long-term forecast load to obtain a long-term energy storage capacity configuration strategy; an energy storage management module, which is used to perform multi-time-scale energy storage management of the energy storage system according to the short-term energy storage capacity configuration strategy, the medium-term energy storage capacity configuration strategy and the long-term energy storage capacity configuration strategy.

[0007] This application proposes one or more technical solutions, which have at least the following technical effects:

[0008] This application obtains a multi-scale prediction window for the power system to perform short-term, medium-term, and long-term load forecasts and determine the predicted loads at different time scales. Then, based on the predicted loads at each time scale, the energy storage demand analysis and capacity allocation are carried out for the energy storage system to form a corresponding capacity allocation decision set. Then, the energy storage efficiency is maximized for each decision set to generate short-term, medium-term, and long-term energy storage capacity configuration strategies. Finally, these strategies are combined to perform multi-time-scale energy storage management on the energy storage system containing multiple energy storage units, to achieve accurate configuration and efficient management of energy storage capacity at different time scales of the power system, to ensure stable and reliable power supply, to achieve accurate matching of energy storage capacity and power system load at multiple time scales, to improve the utilization efficiency of the energy storage system, and to optimize the technical effect of the power system's management and configuration of energy storage.

[0009] The above content summarizes a multi-time-scale energy storage capacity configuration optimization method and system. This application will describe the steps of the technical solution in detail in the following specific implementation methods to facilitate technical personnel to have a clear and complete understanding of this application. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0011] Figure 1 This is a flow chart of a multi-time-scale energy storage capacity configuration optimization method provided in an embodiment of the present application.

[0012] Figure 2 This is a structural diagram of a multi-time-scale energy storage capacity configuration optimization system provided in an embodiment of the present application.

[0013] Explanation of the reference numerals: load forecasting module 1, short-term strategy acquisition module 2, medium-term strategy acquisition module 3, long-term strategy acquisition module 4, energy storage management module 5. DETAILED DESCRIPTION

[0014] This application obtains a multi-time-scale prediction window for the power system to carry out short-term, medium-term, and long-term load forecasts. According to the predicted loads at different time scales, the energy storage demand analysis and capacity allocation of the energy storage system are carried out to form a corresponding decision set. Then, the energy storage efficiency is maximized for each decision set to obtain short-term, medium-term, and long-term energy storage capacity configuration strategies. Finally, based on these strategies, the energy storage system containing multiple energy storage units is managed at multiple time scales to achieve accurate configuration of energy storage capacity, ensure stable and reliable power supply, and improve the energy storage management and operation efficiency of the power system. It achieves a precise match between energy storage capacity and power system load at multiple time scales, improves the utilization efficiency of the energy storage system, and thus optimizes the technical effect of the power system's management and configuration of energy storage.

[0015] The following will be combined with the accompanying drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only some of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0016] It should be noted that any variations of the terms "include" and "have" are intended to cover non-exclusive inclusions. For example, a process, method, system, product or server that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or modules that are not clearly listed or are inherent to these processes, methods, products or devices.

[0017] Example 1, as Figure 1 As shown, a multi-time-scale energy storage capacity configuration optimization method, wherein the method includes:

[0018] Step A100: Perform multi-time-scale load forecasting based on the power system to obtain short-term forecast load, medium-term forecast load and long-term forecast load.

[0019] In an embodiment of the present application, multi-time-scale load forecasting refers to the process of estimating the load changes of the power system under different time spans by obtaining a multi-scale forecast window including short-term, medium-term and long-term, and generating short-term forecast load, medium-term forecast load and long-term forecast load data respectively.

[0020] Specifically, a multi-scale prediction window including a short-term prediction window, a medium-term prediction window and a long-term prediction window is first obtained, and then the load of the power system is predicted based on the short-term prediction window, the medium-term prediction window and the long-term prediction window respectively, thereby generating the short-term prediction load, the medium-term prediction load and the long-term prediction load accordingly. The specific steps are described in detail in A110-A140.

[0021] Step A200: Optimizing the energy storage capacity configuration of the energy storage system according to the short-term predicted load to obtain a short-term energy storage capacity configuration strategy.

[0022] In the embodiment of the present application, the energy storage system is composed of multiple energy storage units, which are used to store electrical energy and adjust the balance of power supply and demand through charging and discharging operations when the load of the power system changes, thereby ensuring a stable power supply.

[0023] Optionally, an energy storage demand analysis is performed on the energy storage system based on the short-term predicted load to determine the short-term energy storage demand. Then, the energy storage capacity is allocated to each energy storage unit of the energy storage system based on the short-term energy storage demand to obtain a short-term capacity allocation decision set. Finally, the energy storage efficiency is maximized based on the short-term capacity allocation decision set to generate a short-term energy storage capacity configuration strategy. The specific steps are described in detail in A210-A230.

[0024] Step A300: Optimizing the energy storage capacity configuration of the energy storage system according to the medium-term predicted load to obtain a medium-term energy storage capacity configuration strategy.

[0025] In one embodiment of the present application, an energy storage demand analysis is performed on the energy storage system based on the medium-term predicted load to determine the medium-term energy storage demand; then, energy storage capacity is allocated to the energy storage system based on the medium-term energy storage demand to obtain a medium-term capacity allocation decision set; finally, energy storage efficiency is maximized based on the medium-term capacity allocation decision set to generate a medium-term energy storage capacity configuration strategy. The specific steps are described in detail in A310-A330.

[0026] Step A400: Optimizing the energy storage capacity configuration of the energy storage system according to the long-term predicted load to obtain a long-term energy storage capacity configuration strategy.

[0027] Specifically, the energy storage demand of the energy storage system is analyzed based on the long-term predicted load to determine the long-term energy storage demand; the energy storage capacity of each energy storage unit of the energy storage system is allocated based on the long-term energy storage demand to obtain a long-term capacity allocation decision set; the energy storage efficiency is maximized based on the long-term capacity allocation decision set to generate a long-term energy storage capacity configuration strategy. The specific steps are detailed in A410-A430.

[0028] Step A500: performing multi-time-scale energy storage management on the energy storage system according to the short-term energy storage capacity configuration strategy, the medium-term energy storage capacity configuration strategy, and the long-term energy storage capacity configuration strategy.

[0029] Specifically, when performing multi-time-scale energy storage management, the above steps are used to obtain a multi-scale forecast window based on the power system, including short-term, medium-term, and long-term forecast windows. After obtaining the corresponding energy storage capacity configuration strategy based on the corresponding predicted load, multi-time-scale energy storage management is performed on the energy storage system according to the short-term, medium-term, and long-term energy storage capacity configuration strategies. In actual operation, when encountering short-term load fluctuations, the short-term energy storage capacity configuration strategy takes effect immediately, controlling the energy storage unit to respond quickly and replenish or release electric energy. If the load change trend is consistent with the medium-term forecast, the medium-term energy storage strategy will come into play and optimize the scheduling of the energy storage system in the longer term. From a more macro perspective, the long-term energy storage strategy ensures the stable operation of the energy storage system over a longer period of time and responds to seasonal and cyclical load changes.

[0030] Through the above multi-time-scale load forecasting, energy storage capacity configuration optimization and comprehensive energy storage management steps, the technical effect of accurately matching the energy storage needs of the power system at different time scales is achieved, improving the utilization efficiency of the energy storage system, and ensuring the stability and reliability of power supply is achieved.

[0031] Furthermore, step A100 in the method provided in the embodiment of the present application includes:

[0032] A110: Obtain a multi-scale prediction window, where the multi-scale prediction window includes a short-term prediction window, a mid-term prediction window, and a long-term prediction window.

[0033] A120: Perform load forecasting on the power system according to the short-term forecast window to generate the short-term forecast load.

[0034] A130: Perform load forecasting on the power system according to the medium-term forecast window to generate the medium-term forecast load.

[0035] A140: Perform load forecasting on the power system according to the long-term forecast window to generate the long-term forecast load.

[0036] Specifically, a multi-scale forecast window is first obtained. The short-term forecast window generally covers a period of hours to days, the medium-term forecast window typically spans weeks to months, and the long-term forecast window can range from months to years. These window divisions are determined by those skilled in the art based on analysis of extensive historical power load data. Research has found that power load is affected by factors such as weather changes and the difference between workdays and weekends in the short term; seasonal changes and industrial production cycles in the medium term; and macroeconomic factors such as regional economic development and population growth in the long term.

[0037] Next, we enter the specific load forecasting phase. Taking short-term forecasting as an example, after obtaining real-time load data from the power system, we construct a short-term load forecasting model based on an LSTM neural network. This model is more capable of capturing short-term load fluctuations. Relevant data within the short-term forecast window, along with real-time load data, is input into the model. After complex internal calculations and training, the model ultimately outputs a short-term forecasted load. The specific steps are detailed in A121-A123.

[0038] When conducting medium-term load forecasting, the combined use of time series analysis and regression analysis methods can generate medium-term forecast loads more accurately:

[0039] Step a: First, conduct time series analysis, a process based on historical load data of the power system. The collected historical load data is arranged in chronological order to form a time series. The autocorrelation function (ACF) and partial autocorrelation function (PACF) are used to identify data characteristics, such as determining whether the time series is stationary. If it is not stationary, it is usually stabilized by differential operation. The ARIMA model (autoregressive integrated moving average model) is used to build the data after the stationary processing. The ARIMA model consists of three parts: autoregression (AR), differencing (I), and moving average (MA). The order of AR and MA is determined based on the ACF and PACF, and the model parameters are estimated using methods such as maximum likelihood estimation, thereby exploring the inherent laws of load changes over time.

[0040] Step b: Next, perform a regression analysis, taking into account external factors such as season and temperature. Seasonal factors are coded, for example, using 0-3 to represent the four seasons, while temperature data is collected directly. A multivariate linear regression model is constructed, using load as the dependent variable and seasonal code, temperature, and other independent variables. The regression coefficients are solved using the least squares method to accurately describe the impact of external factors on load. Finally, the load trend obtained from the time series analysis and the external factor impact results from the regression analysis are combined. The data within the medium-term forecast window is processed using methods such as weighted averaging to generate a medium-term forecast load.

[0041] Long-term load forecasting uses a grey forecasting model and trend extrapolation method because it needs to consider more macroeconomic and social development factors:

[0042] Step c: When applying the grey prediction model, using the GM(1,1) model as an example, the historical load data collected within the long-term prediction window is first accumulated and processed to produce a certain regularity. A grey differential equation is then constructed. Based on the accumulated data, the parameters in the equation are estimated using the least squares method to obtain a prediction model. This model is used to predict the accumulated data, and then a cumulative reduction is performed to obtain the predicted load value. Even with limited data, the grey prediction model can fully tap into the data's potential and effectively capture long-term load trends.

[0043] Step d: The trend extrapolation method first performs curve fitting on the historical data based on the changing trends of historical load data. Common fitting curves include linear, exponential, and logarithmic forms. The appropriate curve type is selected based on the characteristics of the data, and the curve parameters are determined using the least squares method. Then, combined with assumptions such as future macroeconomic growth trends and social development (such as increased electricity demand due to accelerated urbanization), the fitted curve is extrapolated to predict long-term load. The results obtained by the gray prediction model and the trend extrapolation method are compared and comprehensively analyzed. Different weights are assigned to the two methods based on their prediction accuracy and reliability. Finally, the data from the long-term prediction window is used to obtain the long-term predicted load.

[0044] By obtaining multi-scale prediction windows and using different prediction models and methods to conduct short-term, medium-term and long-term load forecasts for the power system, we can comprehensively and accurately grasp the load change trends of the power system at different time scales, providing a reliable data basis for more reasonable subsequent energy storage capacity configuration, and achieving the effect of improving the rationality of the power system energy storage capacity configuration and the stability of power supply.

[0045] Furthermore, step A100 in the method provided in the embodiment of the present application includes:

[0046] A121: Obtain real-time load data of the power system.

[0047] A122: Build a short-term load forecasting model based on the LSTM neural network.

[0048] A123: Input the short-term forecast window and the real-time load data into the short-term load forecast model to obtain the short-term forecast load.

[0049] In the embodiments of the present application, real-time load data is collected and transmitted instantly at specific intervals by sensors and monitoring equipment distributed at key nodes. LSTM neural networks are a special type of recurrent neural network, also known as long short-term memory networks.

[0050] Optionally, when performing short-term load forecasting, first obtain real-time load data for the power system. This data typically comes from sensors (power, current, and voltage sensors) at various monitoring points in the power system. These sensors collect real-time load information for the power system, including power consumption in different regions and time periods. The collected power, current, voltage, and other data are aggregated and arranged in chronological order to form a time series dataset for subsequent analysis and processing.

[0051] Next, we build a short-term load forecasting model based on an LSTM neural network. When building the model, we first determine the model's input layer. The number of neurons in the input layer depends on the number of features in the input data. Time information within the short-term forecast window and real-time load data are used as input data. Assuming the short-term forecast window is load data from the past 24 hours, the number of neurons in the input layer is related to the load data dimensions at these 24 time points. We also need to consider factors that affect load, such as whether it is a weekday or a weekend, and the real-time weather temperature. These factors are also encoded and used as input data.

[0052] Next comes the hidden layer, which consists of multiple LSTM units. Each LSTM unit contains an input gate, a forget gate, an output gate, and a memory unit. The input gate controls the entry of new information into the memory unit, the forget gate determines whether to retain or discard old information in the memory unit, and the output gate determines the output value. These gates are calculated and controlled by the sigmoid function and the tanh function. For example, the input gate calculation formula is i t =σ(W ii x t +W hi h t-1 +b i ), where i t is the value of the input gate at time t, σ is the sigmoid function, W ii and W hi is the weight matrix, xt is the input at time t, h t-1 is the hidden state of the previous moment, b i is the bias. The forget gate and output gate have similar calculation formulas. Multiple LSTM units are connected sequentially, with the output of the previous unit serving as the input of the next unit. This allows for the extraction of features and long-term dependencies in the data layer by layer.

[0053] Finally, the output layer is typically a fully connected layer, with the number of neurons determined by the prediction objective. For example, to predict load values ​​for the next hour, the output layer would have only one neuron. The output layer applies a linear transformation to the hidden layer's output using a weight matrix, then applies an activation function (such as a linear activation function) to obtain the final predicted value.

[0054] Once the model is built, the short-term forecast window and real-time load data are fed into the model for training. During the training process, a large amount of historical data (power system load data and data on external factors affecting load, such as meteorological data, temperature, humidity, wind speed, and sunshine duration) is used as the training set. The weights and biases in the model are continuously adjusted through a backpropagation algorithm to minimize the error (e.g., mean square error) between the model's predicted values ​​and actual values. Once the model training is complete and meets the accuracy requirements required by technicians in this field, the new short-term forecast window and real-time load data are fed into the trained model to obtain the short-term predicted load.

[0055] By acquiring real-time load data of the power system, building and training a short-term load forecasting model based on the LSTM neural network, and then inputting relevant data into the model for prediction, a more accurate prediction of the short-term load of the power system is achieved, providing reliable data support for subsequent short-term energy storage capacity configuration, and achieving the effect of improving the accuracy of short-term energy storage capacity configuration.

[0056] Furthermore, step A200 in the method provided in the embodiment of the present application includes:

[0057] A210: Analyze the energy storage demand of the energy storage system based on the short-term predicted load to determine the short-term energy storage demand.

[0058] A220: Allocate energy storage capacity to each energy storage unit of the energy storage system according to the short-term energy storage demand to obtain a short-term capacity allocation decision set.

[0059] A230: Maximizing energy storage efficiency according to the short-term capacity allocation decision set to generate the short-term energy storage capacity configuration strategy.

[0060] Specifically, after obtaining the short-term forecast load, the energy storage system is first analyzed based on this data to determine short-term energy storage requirements. This short-term forecast load data includes load variations in the power system over a relatively short period of time, such as forecasted power consumption values ​​at different times within the next 24 hours. By analyzing this data, peaks and valleys in load can be identified. Suppose, for example, within the predicted 24-hour period, the power load increases significantly during a certain period, exceeding the stable power supply capacity of the current power generation equipment, while the load is lower during other periods. This creates a need for energy storage.

[0061] Furthermore, when specifically determining the short-term energy storage demand, multiple factors need to be considered. On the one hand, the power difference between the peak load period and the off-peak load period needs to be calculated, that is, the additional power that the energy storage system needs to provide during the peak period. For example, if the load during the peak period is 500 megawatts and the load during the off-peak period is 200 megawatts, then the power difference is 300 megawatts. On the other hand, the duration of the peak period must also be considered, because the energy storage system must not only provide additional power, but also ensure continuous power supply during the peak period. If the peak period lasts for 3 hours, according to the energy formula (energy = power × time), this means that the energy storage system needs to provide at least 900 megawatt-hours of electricity to meet the additional demand during this period, which is the preliminarily determined short-term energy storage demand.

[0062] Next, based on the determined short-term energy storage demand, energy storage capacity is allocated to each energy storage unit in the energy storage system, thereby obtaining a short-term capacity allocation decision set. Assuming that the energy storage system consists of multiple energy storage units of different types (such as lithium battery energy storage units and lead-acid battery energy storage units) and different capacities, the characteristics of each energy storage unit, such as charge and discharge efficiency, lifespan, and cost, need to be considered when allocating capacity. Lithium battery energy storage units have high charge and discharge efficiency but relatively high cost; lead-acid battery energy storage units have low cost but slightly lower charge and discharge efficiency. Based on these characteristics of different energy storage units and combined with short-term energy storage demand, a specific allocation algorithm can be used to allocate capacity. For example, a linear programming algorithm can be used, with meeting the short-term energy storage demand as the constraint and the lowest cost or best overall performance as the objective function, to calculate the capacity to be allocated to each energy storage unit. In this way, a short-term capacity allocation decision set containing the allocated capacity of each energy storage unit can be obtained.

[0063] Finally, based on the short-term capacity allocation decision set, the energy storage efficiency is maximized to generate a short-term energy storage capacity allocation strategy. This process considers factors such as energy loss during the charge and discharge process, based on the allocated capacity of each energy storage unit. Because different energy storage units have different efficiencies at different charge and discharge powers, a storage efficiency model is necessary. For example, lithium batteries are more efficient at lower charge and discharge powers, while lead-acid batteries perform better at medium charge and discharge powers.

[0064] To build the energy storage efficiency model, we first collected historical operating data for each energy storage unit under various conditions, including charge and discharge power, ambient temperature, and state of charge. This data included charge and discharge current, voltage, power, and corresponding energy input and output values. Based on this data, we analyzed the correlation between various factors and energy storage efficiency. Using multiple regression analysis, we constructed a mathematical expression that reflects the relationship between charge and discharge power, ambient temperature, state of charge, and other factors and energy storage efficiency. Finally, we used new operating data to verify the model and adjust its parameters to ensure that it accurately reflects the energy storage efficiency of each energy storage unit under different operating conditions.

[0065] The charging and discharging strategies (e.g., charging and discharging time and power) of each energy storage unit are continuously adjusted. Using optimization algorithms (e.g., genetic algorithms and particle swarm optimization), the system seeks a configuration solution that maximizes the efficiency of the entire energy storage system while meeting short-term energy storage requirements and other constraints (e.g., maximum charge and discharge power limits and state of charge limits for energy storage units). When the optimal solution is found, a short-term energy storage capacity configuration strategy is generated.

[0066] By analyzing the energy storage demand for short-term predicted loads, rationally allocating energy storage unit capacity, and optimizing energy storage efficiency, we can optimize the energy storage capacity configuration on a short-term time scale, improve the utilization efficiency of the energy storage system, and better meet the short-term load fluctuation needs of the power system.

[0067] Furthermore, step A300 in the method provided in the embodiment of the present application includes:

[0068] A310: Analyze the energy storage demand of the energy storage system according to the medium-term forecast load to determine the medium-term energy storage demand.

[0069] A320: Allocate energy storage capacity to the energy storage system according to the medium-term energy storage demand to obtain a medium-term capacity allocation decision set.

[0070] A330: Optimizing energy storage efficiency maximization based on the medium-term capacity allocation decision set to generate the medium-term energy storage capacity configuration strategy.

[0071] Specifically, after obtaining a medium-term load forecast, the energy storage system's energy storage needs are first analyzed based on this data to determine the medium-term energy storage requirements. Medium-term load forecasts typically cover a timeframe of several weeks to several months. For example, they predict average weekly electricity load data for the next month, including load variations between weekdays and weekends, and between weeks due to factors such as industrial production activities and changes in residents' lifestyles. When conducting this analysis, those skilled in the art should not only focus on peak and valley load values ​​but also examine the cyclical nature of load fluctuations. For example, during the peak production season of an industrial park, the average weekly load increases significantly compared to the off-season. Assuming the average off-season load is 800 megawatts and rises to 1,200 megawatts during the peak season, which lasts three weeks, the additional energy storage demand arising solely from changes in production activities during these three weeks needs to be considered. Furthermore, the volatility of renewable energy generation (such as wind power and photovoltaic power generation) during this period can be taken into account to further calculate the power gap that the medium-term energy storage system needs to fill. For example, photovoltaic power generation in the region fluctuates greatly during certain periods, with an average of 100 megawatt-hours of electricity per week unable to be supplied stably. After taking these factors into comprehensive consideration, the medium-term energy storage needs are ultimately determined through methods such as energy balance calculations.

[0072] After clarifying the medium-term energy storage demand, the next step is to allocate storage capacity to the energy storage system based on this demand, thereby obtaining a medium-term capacity allocation decision set. An energy storage system may consist of a variety of energy storage units with different functions and characteristics, such as pumped-storage power plants and large-scale lithium-ion battery energy storage stations. When allocating capacity, it is necessary to fully consider the characteristics of each energy storage unit. For example, pumped-storage power plants are suitable for large-scale, long-term energy storage and release, and have long charge and discharge cycles; lithium-ion battery energy storage stations have fast response times and are suitable for frequent charge and discharge operations. The Analytic Hierarchy Process (AHP) can be used to evaluate each energy storage unit based on multiple dimensions, such as cost, capacity, efficiency, and lifespan, and determine their weights. For example, in terms of cost, pumped-storage power plants have high construction costs but low operating costs, while lithium-ion battery energy storage stations have relatively low construction costs but high battery replacement costs. In terms of capacity, pumped-storage power plants generally have larger capacities, while lithium-ion battery energy storage stations have relatively smaller capacities. Based on the weights of each dimension and combined with medium-term energy storage needs, a reasonable capacity allocation plan is formulated to determine the energy storage capacity that each energy storage unit should bear in the medium term, thereby forming a medium-term capacity allocation decision set.

[0073] Finally, based on the medium-term capacity allocation decision set, the optimization strategy for maximizing energy storage efficiency is generated. Due to the long medium-term time span, the operation of the energy storage system is affected by various factors, such as the impact of seasonal ambient temperature on energy storage unit efficiency and performance degradation due to equipment aging. A multi-factor energy storage efficiency model is constructed (using the same construction process as the previous step), incorporating factors such as ambient temperature, operating time, and charge and discharge cycles. For example, lithium batteries have the highest charge and discharge efficiency at around 25°C; efficiency decreases when the temperature is too high or too low. Pumped-storage power plants experience increased energy losses when they are frequently started and stopped. Using optimization algorithms such as simulated annealing, the operating parameters of each energy storage unit (such as charge and discharge time and power) are continuously adjusted to find a configuration that maximizes the efficiency of the entire energy storage system during medium-term operation, while meeting medium-term energy storage requirements and various constraints (such as the charge and discharge cycle limit of the energy storage unit and equipment maintenance cycles). Once the optimal solution is determined, a medium-term energy storage capacity allocation strategy is generated.

[0074] By conducting a comprehensive energy storage demand analysis of medium-term forecast loads, scientifically and rationally allocating energy storage system capacity, and optimizing energy storage efficiency based on a multi-factor model, we have achieved the effect of optimizing energy storage capacity configuration on a medium-term time scale, improving the long-term operating efficiency of the energy storage system, effectively responding to medium-term load changes in the power system, and ensuring a stable power supply.

[0075] Furthermore, step A400 in the method provided in the embodiment of the present application includes:

[0076] A410: Analyze the energy storage demand of the energy storage system based on the long-term predicted load to determine the long-term energy storage demand.

[0077] A420: Allocate energy storage capacity to each energy storage unit of the energy storage system according to the long-term energy storage demand to obtain a long-term capacity allocation decision set.

[0078] A430: Maximizing energy storage efficiency according to the long-term capacity allocation decision set to generate the long-term energy storage capacity configuration strategy.

[0079] Specifically, after obtaining long-term load forecasts, the energy storage system's energy storage needs are first analyzed based on these forecasts to determine long-term energy storage requirements. Long-term load forecasts typically cover a timeframe of several months to several years. For example, a forecast might include a three-year forecast of electricity load trends in a region. This forecast must consider factors such as the region's GDP growth rate and the layout of emerging industries. If a large industrial park is expected to be built and operational in the region over the next three years, industrial electricity demand is projected to increase by 30%, significantly impacting long-term energy storage requirements. Energy structure adjustments are also a significant factor. If a region plans to significantly increase installed capacity of renewable energy sources such as solar and wind power over the next three years, the intermittent and volatile nature of renewable energy generation necessitates supporting energy storage systems to ensure stable power supply. Assuming 500 MW of new renewable energy capacity, historical data suggests that 10%-20% of this capacity is typically required for energy storage systems to balance power generation fluctuations. This provides a preliminary estimate of the energy storage capacity required for the integration of these new energy sources. By comprehensively considering these factors, simulations and calculations using tools such as system dynamics models can ultimately determine long-term energy storage requirements.

[0080] After clarifying the long-term energy storage demand, the next step is to allocate storage capacity to each energy storage unit in the energy storage system based on this demand, thereby obtaining a long-term capacity allocation decision set. Each energy storage unit in an energy storage system has different characteristics and applicable scenarios. For example, pumped hydro storage technology is mature and has a long lifespan, but it also has a long construction period and limited site selection. Compressed air storage is suitable for large-scale energy storage, but its efficiency is relatively low. Electrochemical energy storage (such as lithium batteries and sodium-sulfur batteries) offers fast response and flexible construction, but it is also costly and has a limited lifespan. When allocating capacity, a mixed integer programming approach is employed, minimizing total cost as the objective function while considering various constraints. These constraints include technical parameter limitations of each energy storage unit, such as upper and lower reservoir capacity limits for pumped hydro storage and charge and discharge cycles for electrochemical storage; geographical constraints, such as certain regions being unsuitable for pumped storage power plants; and time constraints, such as the potential for adjustments to the allocation ratio of each energy storage unit over time as technology advances and costs fluctuate. By solving this optimization problem, those skilled in the art can determine the optimal allocation capacity of each energy storage unit over the long term, thereby forming a long-term capacity allocation decision set.

[0081] Finally, based on the long-term capacity allocation decision set, the system optimizes energy storage efficiency to generate a long-term energy storage capacity allocation strategy. During long-term operation, the efficiency of energy storage systems is affected by various factors, such as equipment performance improvements due to technological advancements and efficiency declines due to equipment aging. A dynamic efficiency model that accounts for both technological advancements and equipment aging is constructed. This model can predict the efficiency changes of each energy storage unit under different operating conditions over many years. For example, with the continuous advancement of lithium battery technology, its energy density is expected to increase by 20% over the next three years, and its charge and discharge efficiency will rise from the current 90% to 95%. However, after 10 years of operation, the efficiency of a pumped-storage power station may decrease by 5% due to equipment wear. Using optimization algorithms such as genetic algorithms, the system searches for a configuration solution that maximizes the long-term efficiency of the entire energy storage system while meeting long-term energy storage requirements and various constraints (such as investment budget limits and environmental requirements). This may include determining the optimal construction and decommissioning time for each energy storage unit, as well as the operating strategy. Once the optimal solution is found, a long-term energy storage capacity allocation strategy is generated.

[0082] By conducting a comprehensive energy storage demand analysis of long-term forecast loads, using mixed integer programming methods to rationally allocate the capacity of each energy storage unit, and optimizing energy storage efficiency based on a dynamic efficiency model, we achieve the effect of optimizing energy storage capacity configuration on a long-term time scale, adapting to the macro-economy and energy structure, and improving the overall efficiency and economic benefits of the energy storage system.

[0083] Furthermore, step A600 in the method provided in the embodiment of the present application includes:

[0084] A610: The energy storage system includes multiple energy storage units.

[0085] In one embodiment, in practical applications, the energy storage requirements of power systems are diverse and complex, and a single energy storage unit often cannot meet all requirements. Therefore, energy storage systems are typically composed of multiple energy storage units. These energy storage units vary in type, such as the common lithium battery energy storage unit, pumped hydro energy storage unit, compressed air energy storage unit, and lead-acid battery energy storage unit, each with unique technical parameters and performance characteristics. For example, lithium battery energy storage units have a fast response speed, capable of completing charge and discharge switching within milliseconds, making them suitable for scenarios requiring rapid power regulation, such as smoothing short-term fluctuations in renewable energy generation. However, lithium batteries have relatively limited capacity and are relatively expensive, with construction costs ranging from approximately 1,500 to 2,000 yuan per kilowatt-hour. Pumped hydro energy storage units, on the other hand, have large capacity, with a single power station capable of storing hundreds of megawatt-hours or even more, making them suitable for large-scale energy storage and long-term power supply. However, their construction is limited by geographical conditions and has a long construction period, typically taking 5-10 years.

[0086] Multiple energy storage units working together can fully leverage their respective strengths and compensate for each other's shortcomings. In short-term load fluctuation scenarios, lithium battery energy storage units, with their fast response characteristics, can quickly absorb or release electricity and stabilize the grid frequency. Assuming that during a certain period of time, due to a sudden drop in photovoltaic power generation, a 50-megawatt gap in grid power occurs, the lithium battery energy storage unit can release electricity to fill the gap in a very short time. In the mid-term seasonal energy demand changes, such as the significant increase in electricity demand during the winter heating season, pumped storage units and large-capacity lithium battery energy storage units can work together. The pumped storage unit uses its large capacity advantage to store electricity during power troughs and release it during peaks; the lithium battery energy storage unit performs fine power regulation.

[0087] In long-term energy storage system planning, the combined configuration of multiple energy storage units is even more important. With the adjustment of energy structure and technological development, new energy storage technologies such as compressed air energy storage can also be incorporated into the system. Different energy storage units have different service lives. Lithium batteries generally have a service life of 8-10 years, while pumped storage units can reach 30-50 years. This requires considering the replacement and coordinated operation of each unit in long-term planning. By rationally configuring multiple energy storage units and using optimization algorithms for capacity allocation and operation strategy formulation based on parameters such as their cost, capacity, charge and discharge efficiency, and lifespan, the energy storage system can operate efficiently at different time scales.

[0088] By analyzing the technical parameters, performance characteristics, and power demand of each energy storage unit at different time scales, and using optimization algorithms for combined configuration and operation strategy planning, we can meet the diverse energy storage needs of the power system and improve the overall performance, reliability, and economy of the energy storage system.

[0089] In summary, the multi-time-scale energy storage capacity configuration optimization method provided by the embodiments of the present application has the following technical effects:

[0090] This application obtains power load forecast data at multiple time scales, conducts demand analysis, capacity allocation, and efficiency optimization for the energy storage system for short-term, medium-term, and long-term forecast loads, and generates corresponding energy storage capacity configuration strategies. On this basis, the configuration strategies at different time scales are integrated, taking into account the characteristics of each energy storage unit and factors such as operating costs and lifespan, and the working mode and capacity allocation of each unit at different time periods are determined through optimization algorithms. At the same time, combined with the real-time operating status of the power system, the energy storage system is dynamically managed and adjusted, ultimately achieving accurate matching and efficient configuration of energy storage capacity and power load at multiple time scales, achieving accurate matching of energy storage capacity and power system load at multiple time scales, improving the utilization efficiency of the energy storage system, and thus optimizing the technical effects of the power system's management and configuration of energy storage.

[0091] Example 2, as Figure 2 As shown, based on the same inventive concept as the aforementioned embodiment 1, this embodiment of the present application provides a multi-time-scale energy storage capacity configuration optimization system, the system comprising:

[0092] The load forecasting module 1 is used to perform multi-time scale load forecasting according to the power system to obtain short-term forecast load, medium-term forecast load and long-term forecast load.

[0093] The short-term strategy acquisition module 2 is used to optimize the energy storage capacity configuration of the energy storage system according to the short-term predicted load and obtain a short-term energy storage capacity configuration strategy.

[0094] The medium-term strategy acquisition module 3 is used to optimize the energy storage capacity configuration of the energy storage system according to the medium-term predicted load and obtain a medium-term energy storage capacity configuration strategy.

[0095] The long-term strategy acquisition module 4 is used to optimize the energy storage capacity configuration of the energy storage system according to the long-term predicted load and obtain a long-term energy storage capacity configuration strategy.

[0096] The energy storage management module 5 is used to perform multi-time-scale energy storage management on the energy storage system according to the short-term energy storage capacity configuration strategy, the medium-term energy storage capacity configuration strategy and the long-term energy storage capacity configuration strategy.

[0097] Furthermore, the load forecasting module 1 is configured to perform the following steps:

[0098] Obtain a multi-scale prediction window, the multi-scale prediction window including a short-term prediction window, a medium-term prediction window, and a long-term prediction window; perform load forecasting on the power system according to the short-term prediction window to generate the short-term predicted load; perform load forecasting on the power system according to the medium-term prediction window to generate the medium-term predicted load; perform load forecasting on the power system according to the long-term prediction window to generate the long-term predicted load.

[0099] Furthermore, the load forecasting module 1 is configured to perform the following steps:

[0100] Obtaining real-time load data of the power system; constructing a short-term load forecasting model based on an LSTM neural network; inputting the short-term forecast window and the real-time load data into the short-term load forecasting model to obtain the short-term forecast load.

[0101] Furthermore, the short-term strategy acquisition module 2 is configured to perform the following steps:

[0102] An energy storage demand analysis is performed on the energy storage system based on the short-term predicted load to determine the short-term energy storage demand; energy storage capacity is allocated to each energy storage unit of the energy storage system based on the short-term energy storage demand to obtain a short-term capacity allocation decision set; energy storage efficiency is maximized based on the short-term capacity allocation decision set to generate the short-term energy storage capacity configuration strategy.

[0103] Furthermore, the mid-term strategy acquisition module 3 is configured to perform the following steps:

[0104] An energy storage demand analysis is performed on the energy storage system based on the medium-term forecast load to determine the medium-term energy storage demand; energy storage capacity is allocated to the energy storage system based on the medium-term energy storage demand to obtain a medium-term capacity allocation decision set; energy storage efficiency is maximized based on the medium-term capacity allocation decision set to generate the medium-term energy storage capacity configuration strategy.

[0105] Furthermore, the long-term strategy acquisition module 4 is configured to perform the following steps:

[0106] An energy storage demand analysis is performed on the energy storage system based on the long-term predicted load to determine the long-term energy storage demand; energy storage capacity is allocated to each energy storage unit of the energy storage system based on the long-term energy storage demand to obtain a long-term capacity allocation decision set; energy storage efficiency is maximized based on the long-term capacity allocation decision set to generate the long-term energy storage capacity configuration strategy.

[0107] Furthermore, the energy storage management module 5 is configured to perform the following steps:

[0108] The energy storage system includes a plurality of energy storage units.

[0109] The multi-time-scale energy storage capacity configuration optimization system provided by an embodiment of the present invention can execute the multi-time-scale energy storage capacity configuration optimization method provided by any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0110] Although the present application makes various references to certain modules in the system according to the embodiments of the present application, any number of different modules may be used and run on the user terminal and / or server, and the various units and modules included are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be achieved; in addition, the specific names of the functional units are only for the convenience of distinguishing each other and are not used to limit the scope of protection of the present invention.

[0111] The above specific embodiments do not constitute a limitation to the scope of protection of this application. It should be understood by those skilled in the art that various modifications, combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent replacements and improvements made within the spirit and principles of this application should be included in the scope of protection of this application. In some cases, the actions or steps recorded in this application can be performed in an order different from that in the embodiments and can still achieve the desired results. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

Claims

1. A multi-time scale energy storage capacity configuration optimization method, characterized in that: The method comprises: Carry out multi-time scale load forecasting based on the power system to obtain short-term forecast load, medium-term forecast load and long-term forecast load; Optimizing the energy storage capacity configuration of the energy storage system according to the short-term predicted load to obtain a short-term energy storage capacity configuration strategy; Optimizing the energy storage capacity configuration of the energy storage system according to the medium-term predicted load to obtain a medium-term energy storage capacity configuration strategy; Optimizing the energy storage capacity configuration of the energy storage system according to the long-term predicted load to obtain a long-term energy storage capacity configuration strategy; Multi-time-scale energy storage management is performed on the energy storage system according to the short-term energy storage capacity configuration strategy, the medium-term energy storage capacity configuration strategy, and the long-term energy storage capacity configuration strategy.

2. The multi-time-scale energy storage capacity configuration optimization method according to claim 1, characterized in that: Multi-timescale load forecasting is performed based on the power system to obtain short-term, medium-term, and long-term forecast loads, including: Obtaining a multi-scale prediction window, wherein the multi-scale prediction window includes a short-term prediction window, a medium-term prediction window, and a long-term prediction window; Performing load forecasting on the power system according to the short-term forecast window to generate the short-term forecast load; Performing load forecasting on the power system according to the medium-term forecast window to generate the medium-term forecast load; The load of the power system is predicted according to the long-term prediction window to generate the long-term predicted load.

3. The multi-time-scale energy storage capacity configuration optimization method according to claim 2, characterized in that: Performing load forecasting on the power system according to the short-term forecast window to generate the short-term forecast load includes: Obtaining real-time load data of the power system; Build a short-term load forecasting model based on LSTM neural network; The short-term forecast window and the real-time load data are input into the short-term load forecast model to obtain the short-term forecast load.

4. The multi-time-scale energy storage capacity configuration optimization method according to claim 1, characterized in that: Optimizing the energy storage capacity configuration of the energy storage system according to the short-term predicted load to obtain a short-term energy storage capacity configuration strategy, including: Performing an energy storage demand analysis on the energy storage system according to the short-term predicted load to determine the short-term energy storage demand; Allocating energy storage capacity to each energy storage unit of the energy storage system according to the short-term energy storage demand to obtain a short-term capacity allocation decision set; The energy storage efficiency is maximized according to the short-term capacity allocation decision set to generate the short-term energy storage capacity configuration strategy.

5. The multi-time-scale energy storage capacity configuration optimization method according to claim 1, characterized in that: Optimizing the energy storage capacity configuration of the energy storage system according to the medium-term forecast load to obtain a medium-term energy storage capacity configuration strategy includes: performing an energy storage demand analysis on the energy storage system according to the medium-term predicted load to determine the medium-term energy storage demand; Allocating energy storage capacity to the energy storage system according to the medium-term energy storage demand to obtain a medium-term capacity allocation decision set; The energy storage efficiency is maximized according to the medium-term capacity allocation decision set to generate the medium-term energy storage capacity configuration strategy.

6. The multi-time-scale energy storage capacity configuration optimization method according to claim 1, characterized in that: Optimizing the energy storage capacity configuration of the energy storage system according to the long-term predicted load to obtain a long-term energy storage capacity configuration strategy includes: Performing an energy storage demand analysis on the energy storage system according to the long-term predicted load to determine the long-term energy storage demand; Allocating energy storage capacity to each energy storage unit of the energy storage system according to the long-term energy storage demand to obtain a long-term capacity allocation decision set; The energy storage efficiency is maximized according to the long-term capacity allocation decision set to generate the long-term energy storage capacity configuration strategy.

7. The multi-time-scale energy storage capacity configuration optimization method according to claim 1, characterized in that: The energy storage system includes a plurality of energy storage units.

8. A multi-time scale energy storage capacity configuration optimization system, characterized in that: A system for implementing a multi-time-scale energy storage capacity configuration optimization method according to any one of claims 1 to 7, comprising: The load forecasting module is used to perform multi-time scale load forecasting based on the power system to obtain short-term forecast load, medium-term forecast load and long-term forecast load; A short-term strategy acquisition module is used to optimize the energy storage capacity configuration of the energy storage system according to the short-term predicted load and obtain a short-term energy storage capacity configuration strategy; a medium-term strategy acquisition module, configured to optimize the energy storage capacity configuration of the energy storage system according to the medium-term predicted load and obtain a medium-term energy storage capacity configuration strategy; A long-term strategy acquisition module, configured to optimize the energy storage capacity configuration of the energy storage system according to the long-term predicted load and obtain a long-term energy storage capacity configuration strategy; An energy storage management module is used to perform multi-time-scale energy storage management on the energy storage system according to the short-term energy storage capacity configuration strategy, the medium-term energy storage capacity configuration strategy, and the long-term energy storage capacity configuration strategy.

Images