Energy storage dynamic adjustment and frequency modulation system and method based on peak-valley price difference

Abstract

The invention discloses an energy storage dynamic adjustment and frequency modulation system and method based on peak-valley price difference, and relates to the technical field of power systems. The system comprises a data acquisition and preprocessing module, a load analysis and priority evaluation module, a dynamic parameter calculation module and a frequency modulation task processing module. The data acquisition and preprocessing module acquires and preprocesses power grid operation data and energy storage system state data; the load analysis and priority evaluation module divides a power grid load time period, evaluates the available capacity of an energy storage system and divides the priority of a battery compartment; the dynamic parameter calculation module calculates the discharge depth and the charge-discharge power threshold value of each battery compartment; and after receiving the frequency modulation task signal, the frequency modulation task processing module decomposes the frequency modulation task signal into sub-tasks, distributes tasks according to the state of the battery compartment and calculation parameters, monitors and executes the tasks in real time, outputs prompt information, and realizes efficient dynamic adjustment and frequency modulation of the energy storage system.

Description

Technical Field

[0001] The present invention relates to the technical field of power systems, and in particular to a system and method for dynamic energy storage regulation and frequency modulation based on peak-valley price differences. Background Art

[0002] With the rapid development of renewable energy and the increasing demand for electricity, energy storage technology is playing an increasingly important role in power systems. Energy storage systems not only effectively store electricity but also regulate grid load through charging and discharging processes, participating in ancillary grid services, particularly playing a key role in frequency and peak regulation. However, the operating model of traditional energy storage systems in power networks primarily relies on fixed charging and discharging strategies and a single control method. These systems fail to fully account for load fluctuations and demand changes in the power system, resulting in suboptimal system efficiency.

[0003] Current energy storage systems typically rely on preset charging and discharging plans and simple timing management methods. Although this approach simplifies system control, it ignores the dynamic changes in power load and system demand. Specifically, existing energy storage systems fail to adjust according to real-time information such as grid load changes and frequency fluctuations, resulting in a failure to provide energy storage support in a timely manner when load demand is high, and a failure to effectively protect the health of the battery when the load is low. In addition, when participating in grid frequency regulation, traditional energy storage systems often require balanced charging and discharging operations on multiple battery compartments. Although this can complete the frequency regulation task, this method easily leads to uniform loss of battery cells and cannot effectively delay battery performance degradation, thereby affecting the long-term operating efficiency of the system. Summary of the Invention

[0004] The purpose of the present invention is to provide a system and method for dynamic regulation and frequency modulation of energy storage based on peak-valley price difference, so as to solve the problems raised in the above background technology.

[0005] In order to solve the above technical problems, the present invention provides the following technical solutions: A method for dynamic regulation and frequency modulation of energy storage based on peak-valley price difference, comprising the following steps: Step S100: Collecting grid operation data and energy storage system status data, and pre-processing the grid operation data and energy storage system status data; the grid operation data includes grid load data and frequency fluctuation data at different time periods; the energy storage system status data includes the remaining power of each battery compartment, charge and discharge voltage, operating temperature, and battery health; Step S200: Perform time series analysis on the pre-processed grid load data and, combined with historical grid load patterns in the database, divide the grid load into time periods. Based on the pre-processed energy storage system status data, evaluate the overall available capacity of the energy storage system and prioritize each battery cell based on its health status, remaining charge, and overall available capacity evaluation results. Step S300. Calculate the depth of discharge and charge / discharge power thresholds of each battery compartment based on the grid load period, the division results, and the priority of each battery compartment; Step S400. When the frequency modulation task signal is received, the frequency modulation task is decomposed into multiple subtasks in combination with the status of each battery compartment; and the subtasks are assigned to the matching battery compartments for execution according to the discharge depth and charge and discharge power threshold of each battery compartment, while the execution status is monitored in real time and corresponding prompt information is output.

[0006] Furthermore, step S100 includes: S101. Collect grid operation data to obtain grid load value sequence L and frequency fluctuation data F, where L={l1,l2,...,ln}, where li represents the load value of the i-th period, and i ranges from 1 to n; F={f1,f2,...,fn}, similarly, fi represents the frequency fluctuation data of the i-th period, and fi=|fi 实际 -fi 额定 |,fi 实际 Indicates the actual frequency of the i-th period, fi 额定 Represents the rated frequency of the i-th period; for the energy storage system, the battery management system is used to collect the status data of each battery compartment of the energy storage system, including the remaining power, charge and discharge voltage, operating temperature and battery health; the remaining power is represented by SOC j , where j represents the battery compartment number; similarly, the corresponding charge and discharge voltage is expressed as Vj, and the operating temperature is expressed as Tj; the battery health is expressed as SOHj, and the corresponding calculation formula is: SOH j =SOH j,0 -α·∑mk=1(ΔQk / Q)-β·∑mk=1[(Ik / I) 2 ·Δtk]; Among them, SOH j,0represents the initial health of the j-th battery compartment, ΔQk represents the change in battery capacity during the k-th charge and discharge process, Q represents the rated capacity of the battery, Ik represents the current value during the k-th charge and discharge process, I represents the rated current of the battery, Δtk represents the duration of the k-th charge and discharge process, m represents the total number of charge and discharge cycles, α represents the capacity decay aging coefficient, which is used to quantify the impact of charge and discharge capacity changes on SOH, and β represents the current stress aging coefficient, which is used to quantify the impact of charge and discharge current on SOH; S102. Perform corresponding preprocessing on the collected grid operation data and energy storage system status data. The preprocessing includes outlier processing and data smoothing. A local outlier factor algorithm is used to identify outliers, and an adaptive weighted moving average is used for data smoothing. The preprocessed grid operation data and energy storage system status data are stored.

[0007] Furthermore, step S200 includes: S201. The pre-processed power grid load data is decomposed and reconstructed using wavelet to extract the trend component and detail component of the load, and the approximate component and detail component of the load data are calculated. The corresponding calculation formula is: L(t)=∑G g=1Dg(t)+AG(t), where Dg(t) represents the detail component of the g-th layer, AG(t) represents the approximate component, G represents the set number of decomposition levels, and the number of decomposition levels G is determined according to the following rule: G=[log2(T / t_min)], where T is the total data duration and t_min represents the minimum fluctuation time scale of interest; the load fluctuation entropy HL is calculated based on the detail component, and the corresponding calculation formula is: HL=-∑ni=1hi·log(hi), hi=|∑G g=1Dg(ti)| / ∑ni=1|∑G g=1Dg(ti)|, where hi represents the standardized fluctuation amplitude probability, and the larger HL is, the more severe the load fluctuation; using [AG(t),∑G g=1Dg(t),HL(t)] as the feature vector, the load pattern is classified by FCM clustering, and the corresponding optimization objective is: min∑ni=1∑C c=1ue ic·||xi-vc|| 2 , where C is the number of preset load mode categories, u icrepresents the membership of the i-th data point to the c-th category, and e is the fuzzy index; xi = [AG(ti), ∑G g = 1Dg(ti), HL(ti)] represents the i-th eigenvector, and vc represents the cluster center vector of the c-th category; by iteratively solving the optimization objective, the membership distribution of each data point to different load patterns is obtained; after clustering is completed, the cluster center threshold interval is set based on the rules of historical load data; for each dimension of each cluster center vc, the value range of the corresponding dimension of each pattern in the historical data is statistically analyzed to determine the threshold interval [LBc, UBc], where LBc represents the lower limit of the c-th category and UBc represents the upper limit of the c-th category; when LBc ≤ xt ≤ UBc is satisfied, it is divided into the C-th category load period; S202. Evaluate the overall available capacity of the energy storage system based on the pre-processed energy storage system status data. The corresponding calculation formula is: C avail =∑M j=1SOC j ·SOH j γ j , where γ j Indicates the importance coefficient of the jth battery compartment, which is determined by its position and connection structure; based on the impact of charge and discharge voltage and operating temperature on battery performance, it affects the battery health SOH. j When the charge and discharge voltage Vj of battery compartment j exceeds the preset normal range, the voltage correction coefficient kv is introduced, i.e. SOH j =SOH j ×kv, and kv < 1; when the operating temperature of battery compartment j exceeds the preset suitable range, the temperature correction coefficient kt is introduced, that is, SOH j =SOH j ×kt, and kt < 1; summarize the health status, remaining power, and overall available capacity evaluation results of each battery compartment, and calculate the priority score S of each battery compartment. The corresponding calculation formula is: S=w1·(SOH j / SOHmax)+w2·(SOC j / SOCmax)+w3·(C avail,j / C avail ), Among them, w1, w2 and w3 represent weight coefficients, and w1+w2+w3=1; SOHmax and SOCmax represent the maximum value of the remaining power and battery health of each battery compartment respectively, C avail,j Represents the available capacity of the jth battery compartment; according to the priority score S of each battery compartment, arrange them in ascending order to obtain the corresponding priority sorting list, and the larger the priority score S, the higher the priority.

[0008] Furthermore, step S300 includes: S301. Based on the grid load period division results, obtain the baseline load L_base. For each period i, calculate the corresponding load change ΔLi = L_i - L_i - 1. Based on the corresponding period interval Δt, calculate the corresponding load response coefficient fh_i, where fh_i = (ΔLi / Δt)·(1 / L_base). Obtain the basic depth of discharge DOD_base and calculate the average battery compartment health SOH_avg. For each battery compartment j, calculate the corresponding depth of discharge DOD_j, and the corresponding calculation formula is: DOD_j=DOD_base·[r1·fh_i·[(SOH j / SOH_avg)·wj]·[1-r2·(Nj / Nmax)]+r3·fi; Where Nj represents the number of cycles for battery compartment j, and Nmax is the maximum number of cycles designed for the battery. r1, r2, and r3 represent adjustment coefficients, which are determined through historical data fitting or expert experience. wj represents the priority weight correction factor, and the corresponding calculation formula is: wj=Sj / S_μ, where Sj represents the priority score S of battery compartment j, and S_μ represents the average priority score of all battery compartments. S302. Construct an LSTM model with the historical load sequence [L_{tn}, ..., L_{t-1}, L_t] as input. The model is trained to predict the load at the next moment L'_{t+1}=LSTM(L_t, L_{t-1}, ..., L_{tn}). The prediction window T1 is set and the load change rate |L'_{t+1}-L_t| / T is calculated. For each battery compartment j, its charge and discharge voltage Vj and operating temperature Tj are detected to obtain the corresponding voltage correction coefficient kv and temperature correction coefficient kt. The corresponding calculation process is obtained by the following formula: k=1-a·(|Y-Ymin| / (Ymax-Ymin)), where a is a constant representing the degree of influence caused by data deviation. Y represents the charge and discharge voltage Vj or the operating temperature Tj, and Ymax and Ymin represent the maximum and minimum values of the data Y, respectively. Obtain the maximum allowable power P_max_j of battery compartment j and calculate the corresponding power threshold P_eff_j. The corresponding calculation formula is: P_eff_j=min(P_max_j,u·(|L'_{t+1}-L_t| / T)·C_act_j·kv·kt·wj), Among them, u represents the safety factor, which is usually between 0.8 and 0.95, in order to reserve a certain safety buffer for the calculation results; C_act_j represents the actual available capacity of battery compartment j.

[0009] Furthermore, step S400 includes: S401. Obtain the FM task signal and process the FM task signal based on the wavelet packet decomposition algorithm, thereby decomposing the FM task signal into different frequency bands, thereby splitting the FM task into several subtasks, each subtask corresponding to the regulation requirements of a specific frequency band; calculate the matching degree Mj of each battery compartment j for each subtask, and the corresponding calculation formula is: Mj=(P_eff_j / P_task)+(SOH j / SOH_avg)+(SOC j / SOC_target), Among them, P_task represents the power requirement of the subtask, SOC_task represents the target remaining power, and the maximum matching degree Mj is selected as the task allocation result; S402. Monitor the charge and discharge voltage Vj and operating temperature Tj of each battery compartment in real time. When the voltage or temperature exceeds the preset range, immediately output prompt information for adjusting the charge and discharge power of the battery compartment and participating in frequency modulation to relevant personnel for further processing.

[0010] A dynamic energy storage regulation and frequency modulation system based on peak-valley price differences, comprising: a data acquisition and preprocessing module, a load analysis and priority assessment module, a dynamic parameter calculation module, and a frequency modulation task processing module; The data acquisition and preprocessing module collects and preprocesses grid operation data and energy storage system status data. Grid operation data includes grid load data and frequency fluctuation data at different time periods. Energy storage system status data includes the remaining power of each battery compartment, charge and discharge voltage, operating temperature, and battery health. The load analysis and priority assessment module performs time series analysis on pre-processed grid load data and, combined with historical grid load patterns in the database, divides grid load periods. It also assesses the overall available capacity of the energy storage system based on pre-processed energy storage system status data and prioritizes each battery cell based on its health status, remaining charge, and overall available capacity assessment results. The dynamic parameter calculation module calculates the depth of discharge and charge and discharge power thresholds of each battery compartment based on the grid load period, the division results, and the priority of each battery compartment; When the frequency modulation task processing module receives the frequency modulation task signal, it decomposes the frequency modulation task into multiple subtasks based on the status of each battery compartment; and assigns the subtasks to the matching battery compartment for execution based on the discharge depth and charge and discharge power threshold of each battery compartment. At the same time, it monitors the execution status in real time and outputs corresponding prompt information.

[0011] Furthermore, the data acquisition and preprocessing module includes a data acquisition unit and a data preprocessing unit; The data acquisition unit collects grid operation data and energy storage system status data. The grid operation data includes grid load data and frequency fluctuation data in different time periods; the energy storage system status data includes the remaining power of each battery compartment, charge and discharge voltage, operating temperature and battery health; the data preprocessing unit preprocesses the grid operation data and energy storage system status data.

[0012] Furthermore, the load analysis and priority assessment module includes a load period division unit and an energy storage priority assessment unit; The load period division unit performs time series analysis on the pre-processed grid load data and divides the grid load period into periods based on the historical grid load patterns in the database. The energy storage priority assessment unit evaluates the overall available capacity of the energy storage system based on the pre-processed energy storage system status data and assigns priorities based on the health status, remaining power and overall available capacity assessment results of each battery compartment.

[0013] Furthermore, the dynamic parameter calculation module includes a discharge depth calculation unit and a power threshold calculation unit; The discharge depth calculation unit calculates the discharge depth of each battery compartment according to the grid load period, the division results and the priority of each battery compartment; the power threshold calculation unit calculates the charge and discharge power threshold of each battery compartment according to the grid load period, the division results and the priority of each battery compartment.

[0014] Furthermore, the frequency modulation task processing module includes a task decomposition and allocation unit and a real-time monitoring and adjustment unit; When the task decomposition and allocation unit receives the frequency modulation task signal, it decomposes the frequency modulation task into multiple subtasks based on the status of each battery compartment, and allocates the subtasks to the matching battery compartment for execution according to the discharge depth and charge and discharge power threshold of each battery compartment; the real-time monitoring and adjustment unit monitors the execution status in real time and outputs corresponding prompt information for further processing by relevant personnel.

[0015] Compared with existing technologies, the present invention offers the following advantages: By collecting real-time data on grid load, frequency fluctuations, and energy storage status, it utilizes algorithms such as wavelet decomposition and FCM clustering to precisely divide load periods. Furthermore, it dynamically adjusts the depth of discharge and charge / discharge power thresholds based on assessment priorities such as the health status and remaining capacity of the energy storage battery compartments. Furthermore, the present invention splits frequency modulation tasks based on wavelet packet decomposition and constructs a matching evaluation model that comprehensively considers factors such as battery compartment power thresholds, health, remaining capacity, and frequency fluctuations to assign subtasks to the most appropriate battery compartments. Furthermore, it incorporates model predictive control (MPC) for real-time feedback regulation, enabling precise frequency modulation, slowing battery performance degradation, and extending the life of the energy storage system. When assessing battery health, the present invention not only considers charge / discharge capacity and current stress, but also makes corrections based on charge / discharge voltage and operating temperature. When voltage or temperature are abnormal, the present invention automatically adjusts the charge / discharge power and frequency modulation participation, and adjusts battery health in real time, providing comprehensive battery protection and improving energy storage system reliability. The present invention uses a variety of data processing and analysis algorithms, such as the local outlier factor algorithm to handle outliers, the LSTM to predict loads, and the hierarchical analysis method to determine weights, to achieve data-driven intelligent decision-making. Compared with traditional simple timing management methods, it can more accurately grasp the operating laws of the power system, provide a scientific basis for energy storage regulation and frequency regulation task allocation, and enhance the system's intelligence level and adaptability. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings: Figure 1 This is a module schematic diagram of a dynamic energy storage regulation and frequency modulation system based on peak-valley price difference according to the present invention. DETAILED DESCRIPTION

[0017] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0018] See also Figure 1 , the present invention provides a technical solution: A dynamic energy storage regulation and frequency modulation system based on peak-valley price differences, comprising: a data acquisition and preprocessing module, a load analysis and priority assessment module, a dynamic parameter calculation module, and a frequency modulation task processing module; The data acquisition and preprocessing module collects and preprocesses grid operation data and energy storage system status data. Grid operation data includes grid load data and frequency fluctuation data at different time periods. Energy storage system status data includes the remaining power of each battery compartment, charge and discharge voltage, operating temperature, and battery health. The load analysis and priority assessment module performs time series analysis on pre-processed grid load data and, combined with historical grid load patterns in the database, divides grid load periods. It also assesses the overall available capacity of the energy storage system based on pre-processed energy storage system status data and prioritizes each battery cell based on its health status, remaining charge, and overall available capacity assessment results. The dynamic parameter calculation module calculates the depth of discharge and charge and discharge power thresholds of each battery compartment based on the grid load period, the division results, and the priority of each battery compartment; When the frequency modulation task processing module receives the frequency modulation task signal, it decomposes the frequency modulation task into multiple subtasks based on the status of each battery compartment; and assigns the subtasks to the matching battery compartment for execution based on the discharge depth and charge and discharge power threshold of each battery compartment. At the same time, it monitors the execution status in real time and outputs corresponding prompt information.

[0019] The data acquisition and preprocessing module includes a data acquisition unit and a data preprocessing unit; The data acquisition unit collects grid operation data and energy storage system status data. The grid operation data includes grid load data and frequency fluctuation data in different time periods; the energy storage system status data includes the remaining power of each battery compartment, charge and discharge voltage, operating temperature and battery health; the data preprocessing unit preprocesses the grid operation data and energy storage system status data.

[0020] The load analysis and priority assessment module includes a load period division unit and an energy storage priority assessment unit; The load period division unit performs time series analysis on the pre-processed grid load data and divides the grid load period into periods based on the historical grid load patterns in the database. The energy storage priority assessment unit evaluates the overall available capacity of the energy storage system based on the pre-processed energy storage system status data and assigns priorities based on the health status, remaining power and overall available capacity assessment results of each battery compartment.

[0021] The dynamic parameter calculation module includes a discharge depth calculation unit and a power threshold calculation unit; The discharge depth calculation unit calculates the discharge depth of each battery compartment according to the grid load period, the division results and the priority of each battery compartment; the power threshold calculation unit calculates the charge and discharge power threshold of each battery compartment according to the grid load period, the division results and the priority of each battery compartment.

[0022] The frequency modulation task processing module includes a task decomposition and allocation unit and a real-time monitoring and adjustment unit; When the task decomposition and allocation unit receives the frequency modulation task signal, it decomposes the frequency modulation task into multiple subtasks based on the status of each battery compartment, and allocates the subtasks to the matching battery compartment for execution according to the discharge depth and charge and discharge power threshold of each battery compartment; the real-time monitoring and adjustment unit monitors the execution status in real time and outputs corresponding prompt information for further processing by relevant personnel.

[0023] A method for dynamic regulation and frequency modulation of energy storage based on peak-valley price difference, comprising the following steps: Step S100: Collect and pre-process grid operation data and energy storage system status data; the grid operation data includes grid load data and frequency fluctuation data for different time periods; the energy storage system status data includes the remaining power of each battery compartment, charge and discharge voltage, operating temperature, and battery health; Step S200: Perform time series analysis on the pre-processed grid load data and, combined with historical grid load patterns in the database, divide the grid load into time periods. Based on the pre-processed energy storage system status data, evaluate the overall available capacity of the energy storage system and prioritize each battery cell based on its health status, remaining charge, and overall available capacity evaluation results. Step S300. Calculate the depth of discharge and charge / discharge power thresholds of each battery compartment based on the grid load period, the division results, and the priority of each battery compartment; Step S400. When the frequency modulation task signal is received, the frequency modulation task is decomposed into multiple subtasks in combination with the status of each battery compartment; and the subtasks are assigned to the matching battery compartments for execution according to the discharge depth and charge and discharge power threshold of each battery compartment, while the execution status is monitored in real time and corresponding prompt information is output.

[0024] Step S100 includes: S101. Collect grid operation data to obtain grid load value sequence L and frequency fluctuation data F, where L={l1,l2,...,ln}, where li represents the load value of the i-th period, and i ranges from 1 to n; F={f1,f2,...,fn}, similarly, fi represents the frequency fluctuation data of the i-th period, and fi=|fi 实际 -fi 额定 |,fi 实际 Indicates the actual frequency of the i-th period, fi 额定 Represents the rated frequency of the i-th period; for the energy storage system, the battery management system is used to collect the status data of each battery compartment of the energy storage system, including the remaining power, charge and discharge voltage, operating temperature and battery health; the remaining power is represented by SOCj , where j represents the battery compartment number; similarly, the corresponding charge and discharge voltage is expressed as Vj, and the operating temperature is expressed as Tj; the battery health is expressed as SOHj, and the corresponding calculation formula is: SOH j =SOH j,0 -α·∑mk=1(ΔQk / Q)-β·∑mk=1[(Ik / I) 2 ·Δtk]; Among them, SOH j,0 represents the initial health of the j-th battery compartment, ΔQk represents the change in battery capacity during the k-th charge and discharge process, Q represents the rated capacity of the battery, Ik represents the current value during the k-th charge and discharge process, I represents the rated current of the battery, Δtk represents the duration of the k-th charge and discharge process, m represents the total number of charge and discharge cycles, α represents the capacity decay aging coefficient, which is used to quantify the impact of charge and discharge capacity changes on SOH, and β represents the current stress aging coefficient, which is used to quantify the impact of charge and discharge current on SOH; S102. Perform corresponding preprocessing on the collected grid operation data and energy storage system status data. The preprocessing includes outlier processing and data smoothing. A local outlier factor algorithm is used to identify outliers, and an adaptive weighted moving average is used for data smoothing. The preprocessed grid operation data and energy storage system status data are stored.

[0025] In this embodiment, a local outlier factor algorithm is used to identify outliers. The corresponding calculation process is: For each sample point p of the grid operation data and the energy storage system status data, the corresponding k-neighborhood is calculated, that is, the k_p neighbors closest to the point are found. For point p and its corresponding neighbor q, the reachable distance Rd is calculated. The corresponding calculation formula is: Rd(p,q)=max(kdist(q),d(p,q)), where kdist(q) is the k-th nearest neighbor distance of point q, and d(p,q) is the distance between points p and q; For each point p, calculate its local reachability density LRD, the corresponding calculation formula is: LRD(p)=1 / [∑ q∈Nk(p) Rd(p,q) / k], where Nk(p) represents the k-neighborhood of point p; for each sample point, the corresponding LOF value is calculated, and the corresponding calculation formula is: LOF(p)=[∑ q∈Nk(p) LRD(q) / LRD(p)] / k, if the LOF value is greater than a set threshold, it means that point p is an outlier; according to this algorithm, outliers are identified as points with larger LOF values, usually values greater than 1, and then these outliers can be replaced by cubic spline interpolation.

[0026] Step S200 includes: S201. The pre-processed power grid load data is decomposed and reconstructed using wavelet to extract the trend component and detail component of the load, and the approximate component and detail component of the load data are calculated. The corresponding calculation formula is: L(t)=∑G g=1Dg(t)+AG(t), where Dg(t) represents the detail component of the g-th layer, AG(t) represents the approximate component, G represents the set number of decomposition levels, and the number of decomposition levels G is determined according to the following rule: G=[log2(T / t_min)], where T is the total data duration and t_min represents the minimum fluctuation time scale of interest; the load fluctuation entropy HL is calculated based on the detail component, and the corresponding calculation formula is: HL=-∑ni=1hi·log(hi), hi=|∑G g=1Dg(ti)| / ∑ni=1|∑G g=1Dg(ti)|, where hi represents the standardized fluctuation amplitude probability, and the larger HL is, the more severe the load fluctuation; using [AG(t),∑G g=1Dg(t),HL(t)] as the feature vector, the load pattern is classified by FCM clustering, and the corresponding optimization objective is: min∑ni=1∑C c=1ue ic·||xi-vc|| 2 , where C is the number of preset load mode categories, u ic represents the membership of the i-th data point to the c-th category, and e is the fuzzy index; xi = [AG(ti), ∑G g = 1Dg(ti), HL(ti)] represents the i-th eigenvector, and vc represents the cluster center vector of the c-th category; by iteratively solving the optimization objective, the membership distribution of each data point to different load patterns is obtained; after clustering is completed, the cluster center threshold interval is set based on the rules of historical load data; for each dimension of each cluster center vc, the value range of the corresponding dimension of each pattern in the historical data is statistically analyzed to determine the threshold interval [LBc, UBc], where LBc represents the lower limit of the c-th category and UBc represents the upper limit of the c-th category; when LBc ≤ xt ≤ UBc is satisfied, it is divided into the C-th category load period; S202. Evaluate the overall available capacity of the energy storage system based on the pre-processed energy storage system status data. The corresponding calculation formula is: C avail =∑M j=1SOC j ·SOH j γ j , where γ j Indicates the importance coefficient of the jth battery compartment, which is determined by its position and connection structure; based on the impact of charge and discharge voltage and operating temperature on battery performance, it affects the battery health SOH. j When the charge and discharge voltage Vj of battery compartment j exceeds the preset normal range, the voltage correction coefficient kv is introduced, that is, SOHj =SOH j ×kv, and kv < 1; when the operating temperature of battery compartment j exceeds the preset suitable range, the temperature correction coefficient kt is introduced, that is, SOH j =SOH j ×kt, and kt < 1; summarize the health status, remaining power, and overall available capacity evaluation results of each battery compartment, and calculate the priority score S of each battery compartment. The corresponding calculation formula is: S=w1·(SOH j / SOHmax)+w2·(SOC j / SOCmax)+w3·(C avail,j / C avail ), Among them, w1, w2 and w3 represent weight coefficients, and w1+w2+w3=1; SOHmax and SOCmax represent the maximum value of the remaining power and battery health of each battery compartment respectively, C avail,j Represents the available capacity of the jth battery compartment; according to the priority score S of each battery compartment, arrange them in ascending order to obtain the corresponding priority sorting list, and the larger the priority score S, the higher the priority.

[0027] Step S300 includes: S301. Based on the grid load period division results, obtain the baseline load L_base. For each period i, calculate the corresponding load change ΔLi = L_i - L_i - 1. Based on the corresponding period interval Δt, calculate the corresponding load response coefficient fh_i, where fh_i = (ΔLi / Δt)·(1 / L_base). Obtain the basic depth of discharge DOD_base and calculate the average battery compartment health SOH_avg. For each battery compartment j, calculate the corresponding depth of discharge DOD_j, and the corresponding calculation formula is: DOD_j=DOD_base·[r1·fh_i·[(SOH j / SOH_avg)·wj]·[1-r2·(Nj / Nmax)]+r3·fi; Where Nj represents the number of cycles for battery compartment j, and Nmax is the maximum number of cycles designed for the battery. r1, r2, and r3 represent adjustment coefficients, which are determined through historical data fitting or expert experience. wj represents the priority weight correction factor, and the corresponding calculation formula is: wj=Sj / S_μ, where Sj represents the priority score S of battery compartment j, and S_μ represents the average priority score of all battery compartments. S302. Construct an LSTM model with the historical load sequence [L_{tn}, ..., L_{t-1}, L_t] as input. The model is trained to predict the load at the next moment L'_{t+1}=LSTM(L_t, L_{t-1}, ..., L_{tn}). The prediction window T1 is set and the load change rate |L'_{t+1}-L_t| / T is calculated. For each battery compartment j, its charge and discharge voltage Vj and operating temperature Tj are detected to obtain the corresponding voltage correction coefficient kv and temperature correction coefficient kt. The corresponding calculation process is obtained by the following formula: k=1-a·(|Y-Ymin| / (Ymax-Ymin)), where a is a constant representing the degree of influence caused by data deviation. Y represents the charge and discharge voltage Vj or the operating temperature Tj, and Ymax and Ymin represent the maximum and minimum values of the data Y, respectively. Obtain the maximum allowable power P_max_j of battery compartment j and calculate the corresponding power threshold P_eff_j. The corresponding calculation formula is: P_eff_j=min(P_max_j,u·(|L'_{t+1}-L_t| / T)·C_act_j·kv·kt·wj), Among them, u represents the safety factor, which is usually between 0.8 and 0.95, in order to reserve a certain safety buffer for the calculation results; C_act_j represents the actual available capacity of battery compartment j.

[0028] Step S400 includes: S401. Obtain the FM task signal and process the FM task signal based on the wavelet packet decomposition algorithm, thereby decomposing the FM task signal into different frequency bands, thereby splitting the FM task into several subtasks, each subtask corresponding to the regulation requirements of a specific frequency band; calculate the matching degree Mj of each battery compartment j for each subtask, and the corresponding calculation formula is: Mj=(P_eff_j / P_task)+(SOH j / SOH_avg)+(SOC j / SOC_target), Among them, P_task represents the power requirement of the subtask, SOC_task represents the target remaining power, and the maximum matching degree Mj is selected as the task allocation result; S402. Monitor the charge and discharge voltage Vj and operating temperature Tj of each battery compartment in real time. When the voltage or temperature exceeds the preset range, immediately output prompt information for adjusting the charge and discharge power of the battery compartment and participating in frequency modulation to relevant personnel for further processing.

[0029] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0030] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art will be able to modify the technical solutions described in the aforementioned embodiments or substitute equivalents for some of the technical features. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.

Claims

1. A method for dynamic energy storage regulation and frequency modulation based on peak-valley price differences, characterized by: The method comprises the following steps: Step S100: Collecting grid operation data and energy storage system status data, and pre-processing the grid operation data and energy storage system status data; the grid operation data includes grid load data and frequency fluctuation data for different time periods; the energy storage system status data includes the remaining power of each battery compartment, charge and discharge voltage, operating temperature, and battery health; Step S200: Perform time series analysis on the pre-processed grid load data and, combined with historical grid load patterns in the database, divide the grid load into time periods. Based on the pre-processed energy storage system status data, evaluate the overall available capacity of the energy storage system and prioritize each battery cell based on its health status, remaining charge, and overall available capacity evaluation results. Step S300. Calculate the depth of discharge and charge / discharge power thresholds of each battery compartment based on the grid load period, the division results, and the priority of each battery compartment; Step S400. When the frequency modulation task signal is received, the frequency modulation task is decomposed into multiple subtasks in combination with the status of each battery compartment; and the subtasks are assigned to the matching battery compartments for execution according to the discharge depth and charge and discharge power threshold of each battery compartment, while the execution status is monitored in real time and corresponding prompt information is output.

2. The method for dynamic energy storage regulation and frequency modulation based on peak-valley price difference according to claim 1, characterized in that: The step S100 includes: S101. Collect grid operation data to obtain grid load value sequence L and frequency fluctuation data F, where L={l1,l2,...,ln}, where li represents the load value of the i-th period, and i ranges from 1 to n; F={f1,f2,...,fn}, similarly, fi represents the frequency fluctuation data of the i-th period, and fi=|fi 实际 -fi 额定 |,fi 实际 Indicates the actual frequency of the i-th period, fi 额定 Represents the rated frequency of the i-th period; for the energy storage system, the battery management system is used to collect the status data of each battery compartment of the energy storage system, including the remaining power, charge and discharge voltage, operating temperature and battery health; the remaining power is represented by SOC j , where j represents the battery compartment number; similarly, the corresponding charge and discharge voltage is expressed as Vj, and the operating temperature is expressed as Tj; the battery health is expressed as SOHj, and the corresponding calculation formula is: SOH j =SOH j,0 -α·∑mk=1(ΔQk / Q)-β·∑mk=1[(Ik / I) 2 ·Δtk]; Among them, SOH j,0 represents the initial health of the j-th battery compartment, ΔQk represents the change in battery capacity during the k-th charge and discharge process, Q represents the rated capacity of the battery, Ik represents the current value during the k-th charge and discharge process, I represents the rated current of the battery, Δtk represents the duration of the k-th charge and discharge process, m represents the total number of charge and discharge cycles, α represents the capacity decay aging coefficient, and β represents the current stress aging coefficient; S102. Perform corresponding preprocessing on the collected power grid operation data and energy storage system status data, wherein the preprocessing includes outlier processing and data smoothing, wherein a local outlier factor algorithm is used to identify outliers, and an adaptive weighted moving average is used for data smoothing; the preprocessed power grid operation data and energy storage system status data are stored.

3. The method for dynamic energy storage regulation and frequency modulation based on peak-valley price difference according to claim 2, characterized in that: The step S200 includes: S201. The pre-processed power grid load data is decomposed and reconstructed using wavelet to extract the trend component and detail component of the load, and the approximate component and detail component of the load data are calculated. The corresponding calculation formula is: L(t)=∑G g=1Dg(t)+AG(t), where Dg(t) represents the detail component of the g-th layer, AG(t) represents the approximate component, and G represents the set number of decomposition levels. The load fluctuation entropy HL is calculated based on the detail component, and the corresponding calculation formula is: HL=-∑ni=1hi·log(hi), hi=|∑G g=1Dg(ti)| / ∑ni=1|∑G g=1Dg(ti)|, where hi represents the standardized fluctuation amplitude. With [AG(t), ∑G g=1Dg(t), HL(t)] as the feature vector, the load pattern is classified by FCM clustering, and the corresponding optimization objective is: min∑ni=1∑C c=1ue ic·||xi-vc|| 2 , where C is the number of preset load mode categories, u ic represents the membership of the i-th data point to the c-th category, and e is the fuzzy index; xi = [AG(ti), ∑G g = 1Dg(ti), HL(ti)] represents the i-th eigenvector, and vc represents the cluster center vector of the c-th category; by iteratively solving the optimization objective, the membership distribution of each data point to different load patterns is obtained; after clustering is completed, the cluster center threshold interval is set based on the rules of historical load data; for each dimension of each cluster center vc, the value range of the corresponding dimension of each pattern in the historical data is statistically analyzed to determine the threshold interval [LBc, UBc], where LBc represents the lower limit of the c-th category and UBc represents the upper limit of the c-th category; when LBc ≤ xt ≤ UBc is satisfied, it is divided into the C-th category load period; S202. Evaluate the overall available capacity of the energy storage system based on the pre-processed energy storage system status data. The corresponding calculation formula is: C avail =∑M j=1SOC j ·SOH j γ j , where γ j Indicates the importance coefficient of the jth battery compartment; based on the impact of charge and discharge voltage and operating temperature on battery performance, the battery health SOH j When the charge and discharge voltage Vj of battery compartment j exceeds the preset normal range, the voltage correction coefficient kv is introduced, that is, SOH j =SOH j ×kv, and kv < 1; when the operating temperature of battery compartment j exceeds the preset suitable range, the temperature correction coefficient kt is introduced, that is, SOH j =SOH j ×kt, and kt < 1; summarize the health status, remaining power, and overall available capacity evaluation results of each battery compartment, and calculate the priority score S of each battery compartment. The corresponding calculation formula is: S=w1·(SOH j / SOHmax)+w2·(SOC j / SOCmax)+w3·(C avail,j / C avail ), Among them, w1, w2 and w3 represent weight coefficients, and w1+w2+w3=1; SOHmax and SOCmax represent the maximum value of the remaining power and battery health of each battery compartment respectively, C avail,j Represents the available capacity of the jth battery compartment; according to the priority score S of each battery compartment, arrange them in ascending order to obtain the corresponding priority sorting list, and the larger the priority score S, the higher the priority.

4. The method for dynamic energy storage regulation and frequency modulation based on peak-valley price difference according to claim 3, characterized in that: The step S300 includes: S301. Based on the grid load period division results, obtain the baseline load L_base. For each period i, calculate the corresponding load change ΔLi = L_i - L_i - 1. Based on the corresponding period interval Δt, calculate the corresponding load response coefficient fh_i, where fh_i = (ΔLi / Δt)·(1 / L_base). Obtain the basic depth of discharge DOD_base and calculate the average battery compartment health SOH_avg. For each battery compartment j, calculate the corresponding depth of discharge DOD_j, and the corresponding calculation formula is: DOD_j=DOD_base·[r1·fh_i·[(SOH j / SOH_avg)·wj]·[1-r2·(Nj / Nmax)]+r3·fi; Where Nj represents the number of cycles of battery compartment j, Nmax is the maximum number of cycles designed for the battery; r1, r2, and r3 represent adjustment coefficients; wj represents the priority weight correction factor, and the corresponding calculation formula is: wj=Sj / S_μ, where Sj represents the priority score S of battery compartment j, and S_μ represents the average priority score of all battery compartments; S302. Construct an LSTM model with the historical load sequence [L_{tn}, ..., L_{t-1}, L_t] as input. The model is trained to predict the load at the next moment L'_{t+1}=LSTM(L_t, L_{t-1}, ..., L_{tn}). The prediction window T1 is set and the load change rate |L'_{t+1}-L_t| / T is calculated. For each battery compartment j, its charge and discharge voltage Vj and operating temperature Tj are detected to obtain the corresponding voltage correction coefficient kv and temperature correction coefficient kt. The corresponding calculation process is obtained by the following formula: k=1-a·(|Y-Ymin| / (Ymax-Ymin)), where a is a constant representing the degree of influence caused by data deviation. Y represents the charge and discharge voltage Vj or the operating temperature Tj, and Ymax and Ymin represent the maximum and minimum values of the data Y, respectively. Obtain the maximum allowable power P_max_j of battery compartment j and calculate the corresponding power threshold P_eff_j. The corresponding calculation formula is: P_eff_j=min(P_max_j,u·(|L'_{t+1}-L_t| / T)·C_act_j·kv·kt·wj), Where u represents the safety factor; C_act_j represents the actual available capacity of battery compartment j.

5. The method for dynamic energy storage regulation and frequency modulation based on peak-valley price difference according to claim 4, characterized in that: The step S400 includes: S401. Obtain the FM task signal and process the FM task signal based on the wavelet packet decomposition algorithm, thereby decomposing the FM task signal into different frequency bands, thereby splitting the FM task into several subtasks, each subtask corresponding to the regulation requirements of a specific frequency band; calculate the matching degree Mj of each battery compartment j for each subtask, and the corresponding calculation formula is: Mj=(P_eff_j / P_task)+(SOH j / SOH_avg)+(SOC j / SOC_target), Among them, P_task represents the power requirement of the subtask, SOC_task represents the target remaining power, and the maximum matching degree Mj is selected as the task allocation result; S402. Monitor the charge and discharge voltage Vj and operating temperature Tj of each battery compartment in real time. When the voltage or temperature exceeds the preset range, immediately output prompt information for adjusting the charge and discharge power of the battery compartment and participating in frequency modulation to relevant personnel for further processing.

6. A system for dynamic energy storage regulation and frequency modulation based on peak-valley price difference, applied to a method for dynamic energy storage regulation and frequency modulation based on peak-valley price difference according to any one of claims 1 to 5, characterized in that: The system includes: a data acquisition and preprocessing module, a load analysis and priority assessment module, a dynamic parameter calculation module and a frequency modulation task processing module; The data acquisition and preprocessing module collects grid operation data and energy storage system status data, and preprocesses the grid operation data and energy storage system status data; the grid operation data includes grid load data and frequency fluctuation data at different time periods; the energy storage system status data includes the remaining power of each battery compartment, charge and discharge voltage, operating temperature and battery health; The load analysis and priority assessment module performs time series analysis on the pre-processed grid load data and divides the grid load time periods based on the historical grid load patterns in the database. It also assesses the overall available capacity of the energy storage system based on the pre-processed energy storage system status data and prioritizes each battery compartment based on its health status, remaining power, and overall available capacity assessment results. The dynamic parameter calculation module calculates the discharge depth and charge and discharge power threshold of each battery compartment according to the grid load period, the division results and the priority of each battery compartment; When the frequency modulation task processing module receives the frequency modulation task signal, it decomposes the frequency modulation task into multiple subtasks based on the status of each battery compartment; and assigns the subtasks to the matching battery compartment for execution based on the discharge depth and charge and discharge power threshold of each battery compartment, while monitoring the execution status in real time and outputting corresponding prompt information.

7. The energy storage dynamic regulation and frequency modulation system based on peak-valley price difference according to claim 6, characterized in that: The data acquisition and preprocessing module includes a data acquisition unit and a data preprocessing unit; The data acquisition unit collects grid operation data and energy storage system status data, wherein the grid operation data includes grid load data and frequency fluctuation data in different time periods; the energy storage system status data includes the remaining power, charge and discharge voltage, operating temperature and battery health of each battery compartment; the data preprocessing unit preprocesses the grid operation data and energy storage system status data.

8. The energy storage dynamic regulation and frequency modulation system based on peak-valley price difference according to claim 6, characterized in that: The load analysis and priority assessment module includes a load period division unit and an energy storage priority assessment unit; The load period division unit performs time series analysis on the pre-processed grid load data and divides the grid load period based on the historical grid load patterns in the database; the energy storage priority evaluation unit evaluates the overall available capacity of the energy storage system based on the pre-processed energy storage system status data, and divides the priority based on the health status, remaining power and overall available capacity evaluation results of each battery compartment.

9. The energy storage dynamic regulation and frequency modulation system based on peak-valley price difference according to claim 6, characterized in that: The dynamic parameter calculation module includes a discharge depth calculation unit and a power threshold calculation unit; The discharge depth calculation unit calculates the discharge depth of each battery compartment according to the grid load period, the division results and the priority of each battery compartment; the power threshold calculation unit calculates the charge and discharge power threshold of each battery compartment according to the grid load period, the division results and the priority of each battery compartment.

10. The energy storage dynamic regulation and frequency modulation system based on peak-valley price difference according to claim 6, characterized in that: The frequency modulation task processing module includes a task decomposition and allocation unit and a real-time monitoring and adjustment unit; When the task decomposition and allocation unit receives the frequency modulation task signal, it decomposes the frequency modulation task into multiple subtasks based on the status of each battery compartment, and allocates the subtasks to the matching battery compartment for execution according to the discharge depth and charge and discharge power threshold of each battery compartment; the real-time monitoring and adjustment unit monitors the execution status in real time and outputs corresponding prompt information, which is further processed by relevant personnel.

Images