Energy storage charging station management method and system based on dynamic load monitoring, and storage medium

Abstract

The invention relates to the technical field of energy storage charging station management, in particular to an energy storage charging station management method and system based on dynamic load monitoring and a storage medium. Charging power demand data, the state of charge (SOC) of an energy storage system and power grid feeder load characteristic data are collected in real time through a multi-source sensor array deployed at a power grid access point and an energy storage unit of an energy storage charging station, and a dynamically updated load characteristic database is constructed. By introducing the dynamic impedance matching algorithm, the equivalent impedance model of the power grid-energy storage interface is established, and the dynamic matching degree function is defined, so that the impedance matching degree between the power grid side and the energy storage system can be monitored and adjusted in real time.

Description

Technical Field

[0001] This invention relates to the field of energy storage charging station management technology, specifically to a dynamic load monitoring method, system, and storage medium for energy storage charging station management. Background Technology

[0002] As a key node in energy storage and distribution, the management and optimization of energy storage charging stations are of paramount importance. Traditional management methods for energy storage charging stations typically employ fixed control strategies, which are ineffective in addressing the challenges posed by dynamic load changes and grid fluctuations. Especially under high load and complex grid environments, fixed control strategies often fail to achieve efficient energy management and system stability. Furthermore, existing methods lack adaptive adjustment mechanisms for the intensity of data fluctuations when processing real-time data, resulting in insufficient accuracy of time warping algorithms. This not only affects the operating efficiency of the energy storage system but may also lead to premature battery degradation. Therefore, the challenge is: how to improve the accuracy of time warping algorithms for energy storage charging stations under dynamic load and complex grid environments to achieve more efficient and stable energy management. Summary of the Invention

[0003] This disclosure proposes a dynamic load monitoring management method, system, and storage medium for energy storage charging stations, aiming to overcome at least one of the deficiencies in the prior art.

[0004] To achieve the above objectives, the technical solution disclosed in this invention is as follows: According to one aspect of this disclosure, a management method for energy storage charging stations based on dynamic load monitoring is provided, the steps of which include: By using a multi-source sensor array of energy storage units deployed at grid access points and energy storage charging stations, charging power demand data, energy storage system state of charge (SOC) and grid feeder load characteristic data are collected in real time to build a dynamically updated load characteristic database. Based on the load characteristic data, a load prediction model is constructed. A sliding time window mechanism is used to perform spatiotemporal alignment processing on the load characteristic data. A dynamic weight allocation algorithm is used to generate a real-time load characteristic matrix of the grid-energy storage joint system. A dynamic scheduling model based on two-layer optimization is established. The upper-layer model uses a genetic algorithm to optimize the charging and discharging timing of the energy storage system, while the lower-layer model uses a dynamic priority evaluation function to perform hierarchical scheduling of charging requests. When the peak load of the power grid is detected to exceed the preset threshold, the adaptive peak shaving control module is activated, and the output power of the energy storage system is adjusted through the dynamic impedance matching algorithm, and a dynamic charging power allocation command is generated. A closed-loop feedback control loop is constructed, and the load forecasting model parameters are updated using a sliding time window mechanism. At the same time, the constraints of the dynamic scheduling model are corrected based on real-time operating data. The closed-loop feedback control loop includes a grid frequency fluctuation compensation module and an energy storage lifetime decay compensation module.

[0005] Furthermore, the step of generating the real-time load feature matrix includes: The data from multiple sensors are timestamp aligned. A dynamic time warping algorithm is used to eliminate sampling frequency differences. Interpolation is used to unify the temporal resolution of each data stream. The dynamic warping function is applied to calculate the time series similarity. The aligned data is stored as a three-dimensional tensor structure. A composite monitoring index set including voltage harmonic distortion rate, power fluctuation slope, and SOC change rate is constructed. The voltage harmonic distortion rate is calculated by fast Fourier transform to determine the proportion of each harmonic component, and the power fluctuation slope is the second derivative of the power change per unit time. Define a dynamic weight allocation function to calculate the real-time weight coefficients of the composite monitoring index set. The calculation formula is as follows: Where i and j represent the indices of different indicators, n represents the total number of indicators, and σ i (t) represents the standard deviation of the i-th indicator within the sliding time window, ΔP i (t) represents the rate of change of power at the current moment, ΔP j (t) represents the power change rate of the j-th index at the current time, and λ is the dynamic adjustment factor, with a value range of 0.5-2.0. The load fluctuation index LFI(t) is calculated using a sliding time window mechanism: , among which, T w μ is the length of the time window. k and σ k Let be the mean and standard deviation of the k-th indicator within the time window, respectively; K represents the total number of indicators; and ||·|| represents the second norm.

[0006] Furthermore, the genetic algorithm includes: The design incorporates a composite fitness function that includes charge / discharge timing, SOC retention rate, and grid interaction cost, expressed as: , where a, b, and c are dynamic adjustment coefficients; A dynamic mutation probability mechanism is introduced, where the mutation probability is adjusted according to a non-linear decay curve with the number of iterations. The specific formula is as follows: Where η is the decay rate factor, p m p is the mutation probability. m0 p is the initial mutation probability. minMinimum protection probability; A co-evolutionary framework is adopted, which combines an elite retention strategy with a population diversity maintenance mechanism, in which elite individuals directly enter the next generation, while the remaining individuals are generated through a tournament selection mechanism. A constraint handling mechanism based on KKT conditions is established, which dynamically adjusts the penalty coefficient of infeasible solutions through Lagrange multipliers and applies an exponentially increasing penalty term to individuals that violate the SOC safety boundary.

[0007] Furthermore, the dynamic impedance matching algorithm includes: Establish an equivalent impedance model for the grid-energy storage interface and define a dynamic matching degree function: Among them, Z grid (t) represents the equivalent impedance on the grid side, Z ess (t) represents the output impedance of the energy storage system, both of which are measured in real time using the online frequency sweep method; Construct an objective optimization function that includes harmonic suppression and transient stability terms: Where h is the harmonic frequency, I h Let be the h-th harmonic current component, α and β be weighting coefficients, and dM(t) / dt represent the rate of change of the matching degree. Multi-step rolling optimization is performed using a model predictive control (MPC) framework, within each control cycle: Predict the trend of power grid impedance changes over multiple future time steps; Establish a quadratic programming problem to solve for the optimal impedance adjustment; The control signal for the energy storage converter is generated by a pulse width modulator.

[0008] Furthermore, the energy storage lifetime degradation compensation module performs the following operations: Establish a battery life prediction model and calculate the real-time life degradation coefficient: Where A is a material constant, and E a Let R be the activation energy, T be the gas constant, T be the battery temperature, and SOC(t) be the state of charge at the current moment. ref For reference state of charge, I rms The current is the effective value, and the exponent γ is an empirical parameter used to adjust the effect of the state of charge on lifetime decay. Construct a dynamic compensation strategy: When the lifetime degradation coefficient D(t) exceeds the threshold, the maximum charge and discharge power limit is automatically reduced. Adjust the SOC operating range according to the real-time attenuation rate, and give priority to using the low attenuation rate region; Generate lifetime balancing scheduling instructions to dynamically distribute current among parallel battery clusters; Implement a multi-timescale compensation mechanism: Second-level time scale: Eliminate instantaneous power surges through dynamic internal resistance compensation algorithm; Hourly timescale: Adjusting depth of charge / discharge (DOD) and cycle count allocation; Monthly timescale: Optimize battery pack rotation strategy and capacity reconfiguration.

[0009] Furthermore, the improvements to the dynamic time warping algorithm include: An adaptive regularization window constraint is introduced, where the window width is dynamically adjusted based on the intensity of data fluctuations. The expression is: Among them, W base Where σ is the base window width, σ is the standard deviation of the current data segment, and k is the sensitivity coefficient; Multi-objective regularization path optimization is adopted to minimize time offset error and feature morphology difference; Cubic spline interpolation is applied to non-matching data segments to reconstruct the data while preserving the original statistical characteristics.

[0010] Furthermore, the calculation of the SOC stability index in the composite fitness function includes: Calculate the standard deviation σ of the SOC trajectory SOC ; Assess the cumulative time T of the SOC crossing the safety boundary. over ; The stability index is defined as: , where ε is a minimal constant to prevent division by zero, and w is a weighting factor.

[0011] Furthermore, the steps of using the model predictive control (MPC) framework for multi-step rolling optimization include: Establish a prediction model library that includes power grid impedance prediction models, energy storage response models, and constraint sets; At the beginning of each control cycle, the latest measurement data is loaded and the predicted state variables are initialized; The optimal control sequence is solved by using a sequential quadratic programming algorithm, and the first three control variables that minimize the objective optimization function are selected. The feasibility of the solution is verified, and the slack variable compensation mechanism is activated when a constraint conflict occurs.

[0012] According to another aspect of this disclosure, a dynamic load monitoring energy storage charging station management system is provided for implementing the management method described above, the management system comprising: The data acquisition module is a multi-source sensor array deployed at the grid access point and the energy storage unit of the energy storage charging station. It is used to collect charging power demand data, energy storage system state of charge (SOC) and grid feeder load characteristic data in real time, and to build a dynamically updated load characteristic database. The load forecasting module generates a real-time load feature matrix of the grid-energy storage joint system based on the load feature data, using a sliding time window mechanism and a dynamic weight allocation algorithm. The dynamic scheduling module establishes a dynamic scheduling model based on two-layer optimization. The upper-layer model uses a genetic algorithm to optimize the charging and discharging timing of the energy storage system, while the lower-layer model uses a dynamic priority evaluation function to perform hierarchical scheduling of charging requests. The adaptive peak shaving control module adjusts the output power of the energy storage system and generates a dynamic charging power allocation command when it detects that the peak load of the power grid exceeds a preset threshold. The closed-loop feedback control module includes a power grid frequency fluctuation compensation subunit and an energy storage lifetime decay compensation subunit. It uses a sliding time window mechanism to update the load forecasting model parameters and corrects the constraints of the dynamic scheduling model based on real-time operating data.

[0013] According to another aspect of this disclosure, a storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the energy storage charging station management method for dynamic load monitoring as described above.

[0014] The beneficial effects of this invention are: This invention introduces a dynamic impedance matching algorithm to establish an equivalent impedance model for the grid-energy storage interface and defines a dynamic matching degree function, enabling real-time monitoring and adjustment of the impedance matching degree between the grid side and the energy storage system. This method not only improves the transient stability of the system but also significantly reduces the impact of harmonic currents, thereby enhancing power quality. By constructing an objective optimization function that includes harmonic suppression and transient stability terms, the overall performance of the system is further optimized, ensuring efficient operation in complex grid environments.

[0015] Furthermore, the adaptive warping window constraint mechanism proposed in this invention dynamically adjusts the window width based on the intensity of data fluctuations, ensuring the robustness and accuracy of the time warping algorithm under different operating conditions. This adaptive adjustment mechanism enables the algorithm to better cope with complex load changes and grid fluctuations, improving the response speed and operating efficiency of the energy storage system. In this way, this invention solves the problem that traditional fixed control strategies cannot effectively cope with dynamic loads and complex grid environments, achieving more precise and reliable energy management.

[0016] Furthermore, this invention improves the accuracy of the time warping algorithm for energy storage charging stations under dynamic loads and complex power grid environments through a series of innovative technical means, thereby achieving more efficient and stable energy management. These improvements not only enhance the overall performance of the system but also extend battery life, reduce maintenance costs, and provide strong support for the development of future smart grids.

[0017] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, the following describes the preferred embodiments of the present invention in detail with reference to the accompanying drawings. Attached Figure Description

[0018] Figure 1 This is a flowchart of an energy storage charging station management method according to an embodiment of the present invention; Figure 2 This is a schematic diagram of a spatiotemporally aligned three-dimensional feature tensor in one embodiment of the present invention; Figure 3 This is a schematic diagram of the response characteristics of the dynamic weight allocation function in one embodiment of the present invention; Figure 4 This is a schematic diagram of the statistical characteristics of a sliding time window in one embodiment of the present invention; Figure 5 This is a schematic diagram of the genetic algorithm optimization process in one embodiment of the present invention; Figure 6 This is a schematic diagram illustrating the change in dynamic impedance matching degree in one embodiment of the present invention; Figure 7 This is a schematic diagram of a battery life decay model in one embodiment of the present invention. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] The present invention provides the following preferred embodiments: Example 1 To address the accuracy issue of time warping algorithms in energy storage charging station management under dynamic load and complex power grid environments, this embodiment proposes a management method for energy storage charging stations based on dynamic load monitoring. By acquiring data in real time using a multi-source sensor array, constructing a load prediction model, establishing a two-layer optimal scheduling model, and activating an adaptive peak-shaving control module, efficient energy management and system stability are achieved. Figure 1As shown, the steps of the energy storage charging station management method are as follows: S100: Through a multi-source sensor array of energy storage units deployed at grid access points and energy storage charging stations, real-time data on charging power demand, energy storage system state of charge (SOC), and grid feeder load characteristics are collected to build a dynamically updated load characteristic database.

[0021] S200. Based on load characteristic data, a load prediction model is constructed. A sliding time window mechanism is used to perform spatiotemporal alignment processing on the load characteristic data. A dynamic weight allocation algorithm is used to generate the real-time load characteristic matrix of the grid-energy storage joint system.

[0022] S300. Establish a dynamic scheduling model based on two-layer optimization. The upper-layer model uses a genetic algorithm to optimize the charging and discharging timing of the energy storage system, while the lower-layer model uses a dynamic priority evaluation function to perform hierarchical scheduling of charging requests.

[0023] S400 When the peak load of the power grid is detected to exceed the preset threshold, the adaptive peak shaving control module is activated, and the output power of the energy storage system is adjusted through the dynamic impedance matching algorithm, and a dynamic charging power allocation command is generated.

[0024] S500 constructs a closed-loop feedback control loop, uses a sliding time window mechanism to update the load forecasting model parameters, and corrects the constraints of the dynamic scheduling model based on real-time operating data. The closed-loop feedback control loop includes a grid frequency fluctuation compensation module and an energy storage lifetime decay compensation module.

[0025] Specifically, multi-source sensor arrays deployed at grid connection points and energy storage charging stations collect real-time data on charging power demand, the energy storage system's state of charge (SOC), and grid feeder load characteristics. These sensors include, but are not limited to, current sensors, voltage sensors, and temperature sensors, used to comprehensively monitor the operating status of the grid and the energy storage system. It is important to understand that multi-source sensor arrays provide high-precision data, laying a solid foundation for subsequent data processing and model building.

[0026] Furthermore, a load forecasting model is constructed based on load characteristic data. This model employs a sliding time window mechanism to perform spatiotemporal alignment processing on the load characteristic data, and generates a real-time load characteristic matrix of the grid-energy storage integrated system through a dynamic weight allocation algorithm. It is understandable that the sliding time window mechanism can effectively capture the time-series characteristics of the data, while the dynamic weight allocation algorithm weights the data according to the importance and relevance of each data source, thereby generating a more accurate load characteristic matrix. This step ensures the real-time performance and accuracy of the load forecasting model, providing a reliable basis for subsequent scheduling decisions.

[0027] Furthermore, a dynamic scheduling model based on two-layer optimization is established. The upper-layer model uses a genetic algorithm to optimize the charging and discharging timing of the energy storage system, while the lower-layer model uses a dynamic priority evaluation function to perform hierarchical scheduling of charging requests. It's important to understand that the genetic algorithm has strong global search capabilities, enabling it to find the optimal charging and discharging timing; while the dynamic priority evaluation function classifies charging requests according to their urgency and importance, ensuring the rational allocation of resources. This two-layer optimization model not only improves the system's response speed but also enhances the flexibility and reliability of scheduling decisions.

[0028] Furthermore, when the peak load of the power grid exceeds a preset threshold, the adaptive peak shaving control module is activated. This module adjusts the output power of the energy storage system using a dynamic impedance matching algorithm and generates dynamic charging power allocation commands. It is understood that the dynamic impedance matching algorithm can adjust in real time based on impedance changes on both the grid side and the energy storage side, ensuring stable system operation. The adaptive peak shaving control module can effectively reduce grid pressure during peak load periods and improve the overall stability of the system.

[0029] Furthermore, a closed-loop feedback control loop is constructed, utilizing a sliding time window mechanism to update the load forecasting model parameters, while simultaneously revising the constraints of the dynamic scheduling model based on real-time operational data. The closed-loop feedback control loop includes a grid frequency fluctuation compensation module and an energy storage lifetime degradation compensation module. It is important to understand that the closed-loop feedback control loop continuously adjusts model parameters and constraints according to actual operating conditions, ensuring the system remains in an optimal state. The grid frequency fluctuation compensation module monitors and compensates for grid frequency changes in real time, maintaining system frequency stability; the energy storage lifetime degradation compensation module extends battery life by calculating a real-time lifetime degradation coefficient.

[0030] The advantage of this embodiment lies in achieving efficient management and optimization of energy storage charging stations through real-time data acquisition from a multi-source sensor array, the construction of a load forecasting model, the establishment of a two-layer optimized scheduling model, and the activation of an adaptive peak-shaving control module. This method not only improves the accuracy of the system but also enhances its response speed and operating efficiency, providing strong support for the development of future smart grids.

[0031] Example 2 To address the challenges of time alignment and load feature matrix generation from multi-source sensor data, this embodiment further optimizes the real-time load feature matrix generation process. Specifically, this embodiment achieves high-precision load feature matrix generation through timestamp alignment processing, dynamic time warping algorithms, interpolation methods, the construction of a composite monitoring index set, and a dynamic weight allocation function.

[0032] Furthermore, timestamp alignment is performed on the multi-source sensor data, and a dynamic time warping algorithm is used to eliminate sampling frequency differences. It's important to understand that the dynamic time warping algorithm effectively handles the problem of inconsistent sampling frequencies between different sensors, ensuring data consistency along the time axis. Interpolation is used to unify the temporal resolution of each data stream, and a dynamic warping function is applied to calculate time series similarity. The aligned data is stored as a three-dimensional tensor structure, such as... Figure 2 As shown. It is understandable that interpolation can smooth data and improve its resolution, while dynamic warping functions can accurately calculate the similarity of time series, thereby constructing a high-precision three-dimensional tensor structure.

[0033] Furthermore, a composite monitoring index set is constructed, including voltage harmonic distortion rate, power fluctuation slope, and state of charge (SOC) rate of change. The voltage harmonic distortion rate is calculated using Fast Fourier Transform (FFT) to determine the proportion of each harmonic component, and the power fluctuation slope is the second derivative of the power change per unit time. It's important to understand that the voltage harmonic distortion rate reflects the degree of harmonic pollution in the power grid, the power fluctuation slope indicates the drastic nature of power changes, and the SOC rate of change reflects the state of charge (SOC) changes in the energy storage system. These indicators together constitute a comprehensive monitoring system, providing a foundation for subsequent weight allocation.

[0034] Furthermore, a dynamic weight allocation function is defined to calculate the real-time weight coefficients of the composite monitoring index set, such as... Figure 3 As shown. The calculation formula is: Where i and j represent the indices of different indicators, n represents the total number of indicators, and σ i (t) represents the standard deviation of the i-th indicator within the sliding time window, ΔP i (t) represents the rate of change of power at the current moment, ΔP j (t) represents the power change rate of the j-th indicator at the current time, and λ is a dynamic adjustment factor, with a value ranging from 0.5 to 2.0. It's important to understand that this weight allocation function can dynamically adjust the weights based on the standard deviation and power change rate of each indicator, thus more accurately reflecting the importance of each indicator.

[0035] Furthermore, the load fluctuation index LFI(t) is calculated using a sliding time window mechanism: , among which, T w μ is the length of the time window. k and σ k Let be the mean and standard deviation of the k-th indicator within the time window, respectively, where K represents the total number of indicators, and ||·|| represents the second norm. For example... Figure 4As shown, the sliding time window mechanism can update the load fluctuation index in real time, thereby reflecting changes in the system promptly. In this way, this embodiment can achieve accurate capture and dynamic adjustment of load characteristics.

[0036] The advantage of this embodiment lies in achieving high-precision load characteristic matrix generation through timestamp alignment, dynamic time warping algorithms, interpolation, construction of a composite monitoring index set, and a dynamic weight allocation function. This method not only improves the accuracy of data processing but also enhances the system's adaptability and response speed, providing a reliable basis for subsequent scheduling decisions.

[0037] Example 3 To address the application challenges of genetic algorithms in energy storage charging station management, this embodiment further refines the design and implementation of the genetic algorithm. Specifically, this embodiment achieves efficient optimization of the charging and discharging sequence of the energy storage system by designing a composite fitness function, introducing a dynamic mutation probability mechanism, adopting a co-evolutionary framework that combines an elite preservation strategy with a population diversity maintenance mechanism, and establishing a constraint processing mechanism based on KKT conditions.

[0038] Furthermore, a composite fitness function is designed, incorporating charge / discharge timing, SOC retention rate, and grid interaction cost. The expression is: Where a, b, and c are dynamic adjustment coefficients; it is important to understand that this fitness function comprehensively considers grid cost, SOC stability, and charge / discharge efficiency, enabling a more comprehensive evaluation of individual performance. By using dynamic adjustment coefficients, the weights of each factor can be flexibly adjusted according to actual conditions, thereby achieving a more reasonable optimization objective.

[0039] Furthermore, a dynamic mutation probability mechanism is introduced, where the mutation probability is adjusted according to a non-linear decay curve with the number of iterations. For example... Figure 5 As shown, the specific formula is: Where η is the decay rate factor, p m p is the mutation probability. m0 p is the initial mutation probability. min This represents the minimum protection probability. It's important to understand that the dynamic mutation probability mechanism maintains a high mutation probability in the early stages of the search to explore a larger solution space; however, it gradually decreases the mutation probability in the later stages to enhance local search capabilities. This mechanism helps balance global and local search, improving the algorithm's convergence speed and optimization performance.

[0040] Furthermore, a co-evolutionary framework combining an elite preservation strategy and a population diversity maintenance mechanism is adopted. In this framework, elite individuals directly enter the next generation, while the remaining individuals are selected through a tournament selection mechanism. It is understood that the elite preservation strategy ensures that the best individuals in each generation are retained, preventing the loss of excellent solutions; while the tournament selection mechanism selects superior individuals through competition, maintaining population diversity. This co-evolutionary framework effectively improves the algorithm's search efficiency and solution quality.

[0041] Furthermore, a constraint handling mechanism based on KKT conditions is established. This mechanism dynamically adjusts the penalty coefficients for infeasible solutions using Lagrange multipliers, imposing exponentially increasing penalty terms on individuals that violate the SOC safety boundary. It's important to understand that KKT conditions are an effective tool for solving constrained optimization problems. By introducing Lagrange multipliers, the constraints can be transformed into part of the objective function. Dynamically adjusting the penalty coefficients effectively guides the algorithm towards the feasible solution region, thus ensuring the feasibility of the optimization results.

[0042] This embodiment enables efficient optimization of the charging and discharging timing of energy storage systems. This method not only improves the system's response speed and operating efficiency but also enhances its stability and reliability, providing strong support for the development of smart grids.

[0043] Example 4 To address the application challenges of dynamic impedance matching algorithms in energy storage charging station management, this embodiment further refines the design and implementation of the dynamic impedance matching algorithm. Specifically, this embodiment achieves efficient impedance matching and system stability optimization by establishing an equivalent impedance model of the grid-energy storage interface, defining a dynamic matching degree function, constructing an objective optimization function, and employing a model predictive control (MPC) framework.

[0044] Furthermore, an equivalent impedance model for the grid-energy storage interface is established, and a dynamic matching degree function is defined: Among them, Z grid (t) represents the equivalent impedance on the grid side, Z ess (t) represents the output impedance of the energy storage system, both of which are measured in real time using an online frequency sweep method. For example... Figure 6 As shown, this matching degree function can quantify the impedance matching degree between the grid side and the energy storage side, thus providing a basis for subsequent optimization. Through online frequency sweeping, impedance data can be acquired in real time, ensuring the real-time performance and accuracy of the model.

[0045] Furthermore, a target optimization function is constructed that includes harmonic suppression terms and transient stability terms: Where h is the harmonic frequency, I hLet be the h-th harmonic current component, α and β be weighting coefficients, and dM(t) / dt represent the rate of change of the matching degree. It can be understood that this objective function comprehensively considers harmonic suppression and transient stability, and by optimizing the harmonic current and the rate of change of the matching degree, it can effectively improve the power quality and stability of the system.

[0046] Furthermore, a model predictive control (MPC) framework is employed for multi-step rolling optimization. Within each control cycle, the grid impedance change trend over multiple future time steps is predicted, a quadratic programming problem is established to solve for the optimal impedance adjustment, and a control signal for the energy storage converter is generated via a pulse width modulator. It's important to understand that the MPC framework can make optimization decisions in advance by predicting future system states, thereby achieving more precise control. By establishing a quadratic programming problem, the optimal impedance adjustment can be solved efficiently, ensuring stable system operation.

[0047] This embodiment enables efficient impedance matching and system stability optimization. This method not only improves the system's power quality but also enhances its stability and response speed.

[0048] Example 5 To address the application challenges of energy storage lifespan degradation compensation modules in energy storage charging station management, this embodiment further refines the design and implementation of such modules. Specifically, this embodiment achieves efficient battery lifespan management by establishing a battery lifespan prediction model, calculating real-time lifespan degradation coefficients, constructing dynamic compensation strategies, and implementing multi-timescale compensation mechanisms.

[0049] Furthermore, such as Figure 7 As shown, a battery life prediction model is established to calculate the real-time lifespan degradation coefficient: Where A is a material constant, and E a Let R be the activation energy, T be the gas constant, T be the battery temperature, and SOC(t) be the state of charge at the current moment. ref For reference state of charge, I rms The current is the effective value, and the exponential term γ is an empirical parameter used to adjust the impact of state of charge on battery life degradation. It's important to understand that this life prediction model can accurately calculate the real-time battery life degradation coefficient, thus providing a basis for subsequent compensation strategies.

[0050] Furthermore, a dynamic compensation strategy is constructed. When the lifetime degradation coefficient D(t) exceeds a threshold, the maximum charge / discharge power limit is automatically reduced. The SOC operating range is adjusted based on the real-time degradation rate, prioritizing the use of regions with low degradation rates. A lifetime balancing scheduling command is generated to dynamically distribute current among parallel battery clusters. It is understandable that by dynamically adjusting the charge / discharge power limit and SOC operating range, the battery life can be effectively extended. Simultaneously, dynamic current distribution enables balanced use of the battery pack, improving the overall system reliability and stability.

[0051] Furthermore, a multi-timescale compensation mechanism is implemented. At the second-level timescale: a dynamic internal resistance compensation algorithm eliminates instantaneous power surges; at the hour-level timescale: the depth of charge / discharge (DOD) and cycle count allocation are adjusted; at the monthly-level timescale: battery pack rotation strategies and capacity recombination are optimized. It's important to understand that this multi-timescale compensation mechanism manages and optimizes battery life from different time dimensions. The second-level dynamic internal resistance compensation algorithm can promptly eliminate instantaneous power surges, protecting the battery from damage; the hour-level DOD and cycle count adjustments can rationally allocate battery usage intensity; and the monthly-level battery pack rotation and capacity recombination enable long-term battery life management.

[0052] This embodiment enables efficient battery life management. This method not only extends battery life but also improves system stability and reliability.

[0053] Example 6 To address the accuracy and robustness issues of the dynamic time warping algorithm in multi-source sensor data alignment, this embodiment further optimizes the algorithm. Specifically, this embodiment achieves high-precision data alignment by introducing adaptive warping window constraints, employing multi-objective warping path optimization, and applying cubic spline interpolation to reconstruct non-matching data segments.

[0054] Furthermore, an adaptive regularization window constraint is introduced, where the window width is dynamically adjusted based on the intensity of data fluctuations. The expression is: Among them, W base The base window width is given by σ, the standard deviation of the current data segment is given by σ, and k is the sensitivity coefficient. It's important to understand that this adaptive warping window automatically adjusts its width based on the intensity of data fluctuations, thus more accurately capturing the temporal characteristics of the data. When data fluctuations are significant, the window width increases accordingly to ensure data smoothness and consistency.

[0055] Furthermore, a multi-objective regularization path optimization is employed to simultaneously minimize temporal offset error and feature morphology differences. It is understandable that multi-objective optimization can find a balance among multiple evaluation metrics, thereby achieving more comprehensive data alignment. Temporal offset error reflects the degree of alignment of data along the time axis, while feature morphology differences reflect the similarity of data in shape. By simultaneously optimizing these two metrics, the accuracy and robustness of data alignment can be improved.

[0056] Furthermore, cubic spline interpolation is applied to reconstruct the data from the mismatched data segments. It's important to understand that cubic spline interpolation is a commonly used interpolation method that can smooth the data and improve its resolution while preserving the original statistical characteristics. This method is particularly suitable for processing mismatched data segments, ensuring data consistency and integrity.

[0057] Furthermore, to further improve the accuracy of data alignment, this embodiment also introduces a data preprocessing step. For example, a low-pass filter is used to remove high-frequency noise, and a median filter is used to remove outliers. These preprocessing steps can effectively improve the quality of the data, thereby providing a better foundation for subsequent time warping.

[0058] The advantage of this embodiment lies in achieving high-precision data alignment by introducing adaptive regularization window constraints, employing multi-objective regularization path optimization, and applying cubic spline interpolation to reconstruct non-matching data segments. This not only improves the accuracy of data processing but also enhances the system's response speed, providing a reliable basis for subsequent load forecasting and scheduling decisions.

[0059] Example 7 To address the problem of calculating the SOC stability index in the composite fitness function, this embodiment further refines the calculation method for the SOC stability index. Specifically, this embodiment achieves a more accurate SOC stability assessment by calculating the standard deviation of the SOC trajectory, evaluating the cumulative time for the SOC to cross the safety boundary, and defining the stability index.

[0060] Furthermore, the standard deviation σ of the SOC trajectory is calculated. SOC It's important to understand that the standard deviation of the SOC trajectory reflects the degree of fluctuation in SOC changes. A smaller standard deviation indicates more stable SOC changes and higher system stability. Calculating the standard deviation quantifies the volatility of SOC changes, providing a basis for subsequent stability assessments.

[0061] Furthermore, assess the cumulative time T of the SOC crossing the safety boundary. overUnderstandably, the time it takes for a System-on-Chip (SOC) to cross the safety boundary reflects the system's operational security. A longer cumulative time indicates a longer period of unsafe operation and lower stability. By assessing the cumulative time, potential security vulnerabilities in system operation can be identified promptly, allowing for appropriate corrective measures.

[0062] Furthermore, the stability index is defined as: , where ε is a minimal constant to prevent division by zero, and w is a weighting factor. It's important to understand that this stability index comprehensively considers both the standard deviation of the State of Charge (SOC) and the cumulative time to cross the safety boundary. A smaller standard deviation results in a higher stability index; a shorter cumulative time also results in a higher stability index. This approach allows for a comprehensive evaluation of the system's stability and provides a reliable fitness function for genetic algorithms.

[0063] Furthermore, to further improve the accuracy of the stability index, this embodiment also introduces data smoothing processing. For example, the SOC data is smoothed using a moving average method or a Kalman filter to reduce the impact of noise. These smoothing methods can effectively improve the quality of the data, thereby providing a better foundation for the calculation of the stability index.

[0064] Example 8 To address the application challenges of the Model Predictive Control (MPC) framework in energy storage charging station management, this embodiment further refines the specific implementation steps of the MPC framework. Specifically, this embodiment achieves efficient multi-step rolling optimization by establishing a predictive model library, loading the latest measurement data and initializing the predictive state variables, solving for the optimal control sequence using a sequential quadratic programming algorithm, and verifying the feasibility of the solution results.

[0065] Furthermore, a prediction model library is established, comprising a grid impedance prediction model, an energy storage response model, and a constraint set. It's important to understand that this prediction model library forms the foundation of the MPC framework, containing all necessary models and constraints. The grid impedance prediction model predicts future grid impedance trends, the energy storage response model describes the dynamic response characteristics of the energy storage system, and the constraint set includes various restrictions on system operation. Establishing a comprehensive prediction model library provides a reliable foundation for subsequent optimization.

[0066] Furthermore, at the beginning of each control cycle, the latest measurement data is loaded and the predicted state variables are initialized. It is understood that loading the latest measurement data in real time ensures the model's real-time performance and accuracy. Initializing the predicted state variables sets initial conditions for subsequent optimization calculations. In this way, it is ensured that the optimization of each control cycle is based on the latest data and state.

[0067] Furthermore, the optimal control sequence is solved using a sequential quadratic programming algorithm, selecting the top three control variables that minimize the objective function. It's important to understand that sequential quadratic programming is a highly efficient optimization algorithm capable of quickly finding the optimal control sequence while satisfying constraints. Selecting the top three control variables ensures the continuity and stability of control, avoiding system instability caused by frequent switching of control strategies.

[0068] Furthermore, the feasibility of the solution is verified, and a slack variable compensation mechanism is activated when constraint conflicts occur. It is understandable that feasibility verification is a crucial step in ensuring the feasibility of the optimization results. When constraint conflicts arise, activating the slack variable compensation mechanism can relax the constraints to a certain extent, thereby finding a feasible control scheme. This mechanism can effectively handle complex constraints and ensure the stable operation of the system.

[0069] Example 9 To address the implementation challenges of energy storage charging station management methods based on dynamic load monitoring, this embodiment proposes an energy storage charging station management system based on dynamic load monitoring. Specifically, this embodiment achieves efficient energy storage charging station management through a data acquisition module, a load forecasting module, a dynamic scheduling module, an adaptive peak shaving control module, and a closed-loop feedback control module.

[0070] Furthermore, the data acquisition module is deployed in multi-source sensor arrays at grid connection points and energy storage units in energy storage charging stations. This arrays are used to collect real-time data on charging power demand, the energy storage system's state of charge (SOC), and grid feeder load characteristics, and to construct a dynamically updated load characteristic database. It is important to understand that the multi-source sensor arrays provide comprehensive data, laying a solid foundation for subsequent data processing and model building. The dynamically updated load characteristic database reflects the system's operating status in real time, providing a reliable basis for subsequent forecasting and scheduling.

[0071] Furthermore, the load forecasting module generates a real-time load characteristic matrix for the grid-energy storage integrated system based on load characteristic data, employing a sliding time window mechanism and a dynamic weight allocation algorithm. Understandably, the sliding time window mechanism effectively captures the time-series characteristics of the data, while the dynamic weight allocation algorithm weights data according to the importance and relevance of each data source, thereby generating a more accurate load characteristic matrix. This step ensures the real-time performance and accuracy of the load forecasting model, providing a reliable basis for subsequent scheduling decisions.

[0072] Furthermore, the dynamic scheduling module establishes a dynamic scheduling model based on two-layer optimization. The upper-layer model uses a genetic algorithm to optimize the charging and discharging timing of the energy storage system, while the lower-layer model uses a dynamic priority evaluation function to perform hierarchical scheduling of charging requests. It's important to understand that the genetic algorithm has strong global search capabilities, enabling it to find the optimal charging and discharging timing; while the dynamic priority evaluation function classifies charging requests according to their urgency and importance, ensuring the rational allocation of resources. This two-layer optimization model not only improves the system's response speed but also enhances the flexibility and reliability of scheduling decisions.

[0073] Furthermore, when the adaptive peak shaving control module detects that the peak load of the power grid exceeds a preset threshold, it adjusts the output power of the energy storage system through a dynamic impedance matching algorithm and generates a dynamic charging power allocation command. It is understood that the dynamic impedance matching algorithm can adjust in real time according to the impedance changes on both the grid side and the energy storage side, ensuring the stable operation of the system. The adaptive peak shaving control module can effectively reduce the pressure on the power grid during peak load periods and improve the overall stability of the system.

[0074] Furthermore, the closed-loop feedback control module includes a grid frequency fluctuation compensation subunit and an energy storage lifetime degradation compensation subunit. It utilizes a sliding time window mechanism to update the load forecasting model parameters and simultaneously corrects the constraints of the dynamic scheduling model based on real-time operating data. It's important to understand that the closed-loop feedback control loop can continuously adjust model parameters and constraints according to actual operating conditions, ensuring the system is always in an optimal state. The grid frequency fluctuation compensation subunit can monitor and compensate for changes in grid frequency in real time, maintaining system frequency stability; the energy storage lifetime degradation compensation subunit extends battery life by calculating a real-time lifetime degradation coefficient.

[0075] Example 10 To address the application problem of computer programs in implementing a dynamic load monitoring-based energy storage charging station management method, this embodiment proposes a storage medium design. Specifically, this embodiment stores a computer program on the storage medium, and when the computer program is executed by a processor, it implements the aforementioned dynamic load monitoring-based energy storage charging station management method.

[0076] Furthermore, the storage medium can be any form of computer-readable storage medium, including but not limited to hard drives, solid-state drives, optical discs, USB flash drives, SD cards, etc. It's important to understand that the choice of storage medium depends on the specific application scenario and requirements. For example, for a fixed-installation system, a hard drive or solid-state drive can be chosen; for portable devices, a USB flash drive or SD card can be selected.

[0077] Furthermore, the computer program is designed with a modular structure, including a data acquisition module, a load forecasting module, a dynamic scheduling module, an adaptive peak shaving control module, and a closed-loop feedback control module. Each module is responsible for a specific function, and the entire management method is implemented through the collaborative work between the modules. Understandably, the modular structure not only improves the maintainability and scalability of the program but also facilitates the independent development and testing of different functions.

[0078] Furthermore, the computer program also includes data interfaces and communication protocols for exchanging data with multi-source sensor arrays, energy storage systems, power grids, and other external systems. It is important to understand that standardized data interfaces and communication protocols ensure seamless connectivity and data transmission between different devices and systems.

[0079] Furthermore, to improve the performance and reliability of the computer program, this embodiment also introduces an exception handling mechanism and a logging function. The exception handling mechanism can capture and handle various abnormal situations during program execution, ensuring the stable operation of the system. The logging function is used to record the system's operating status and events, facilitating troubleshooting and performance analysis.

[0080] Although the present invention has been specifically described above with reference to preferred embodiments, it should be understood that the present invention is not limited to the embodiments described above. Various modifications and variations can be made by those skilled in the art without departing from the spirit of the present invention, and such modifications and variations should fall within the scope defined by the appended claims and their equivalents.

Claims

1. A management method for energy storage charging stations with dynamic load monitoring, characterized in that, The steps of the management method include: By using a multi-source sensor array of energy storage units deployed at grid access points and energy storage charging stations, charging power demand data, energy storage system state of charge (SOC) and grid feeder load characteristic data are collected in real time to build a dynamically updated load characteristic database. Based on the load characteristic data, a load prediction model is constructed. A sliding time window mechanism is used to perform spatiotemporal alignment processing on the load characteristic data. A dynamic weight allocation algorithm is used to generate a real-time load characteristic matrix of the grid-energy storage joint system. A dynamic scheduling model based on two-layer optimization is established. The upper-layer model uses a genetic algorithm to optimize the charging and discharging timing of the energy storage system, while the lower-layer model uses a dynamic priority evaluation function to perform hierarchical scheduling of charging requests. When the peak load of the power grid is detected to exceed the preset threshold, the adaptive peak shaving control module is activated, and the output power of the energy storage system is adjusted through the dynamic impedance matching algorithm, and a dynamic charging power allocation command is generated. A closed-loop feedback control loop is constructed, and the load forecasting model parameters are updated using a sliding time window mechanism. At the same time, the constraints of the dynamic scheduling model are corrected based on real-time operating data. The closed-loop feedback control loop includes a grid frequency fluctuation compensation module and an energy storage lifetime decay compensation module.

2. The energy storage charging station management method with dynamic load monitoring as described in claim 1, characterized in that, The steps for generating the real-time load feature matrix include: The data from multiple sensors are timestamp aligned. A dynamic time warping algorithm is used to eliminate sampling frequency differences. Interpolation is used to unify the temporal resolution of each data stream. The dynamic warping function is applied to calculate the time series similarity. The aligned data is stored as a three-dimensional tensor structure. A composite monitoring index set including voltage harmonic distortion rate, power fluctuation slope, and SOC change rate is constructed. The voltage harmonic distortion rate is calculated by fast Fourier transform to determine the proportion of each harmonic component, and the power fluctuation slope is the second derivative of the power change per unit time. Define a dynamic weight allocation function to calculate the real-time weight coefficients of the composite monitoring index set. The calculation formula is as follows: Where i and j represent the indices of different indicators, n represents the total number of indicators, and σ i (t) represents the standard deviation of the i-th indicator within the sliding time window, ΔP i (t) represents the rate of change of power at the current moment, ΔP j (t) represents the power change rate of the j-th index at the current time, and λ is the dynamic adjustment factor, with a value range of 0.5-2.

0. The load fluctuation index LFI(t) is calculated using a sliding time window mechanism: , among which, T w μ is the length of the time window. k and σ k Let be the mean and standard deviation of the k-th indicator within the time window, respectively; K represents the total number of indicators; and ||·|| represents the second norm.

3. The energy storage charging station management method with dynamic load monitoring as described in claim 2, characterized in that, The genetic algorithm includes: The design incorporates a composite fitness function that includes charge / discharge timing, SOC retention rate, and grid interaction cost, expressed as: , where a, b, and c are dynamic adjustment coefficients; A dynamic mutation probability mechanism is introduced, where the mutation probability is adjusted according to a non-linear decay curve with the number of iterations. The specific formula is as follows: Where η is the decay rate factor, p m p is the mutation probability. m0 p is the initial mutation probability. min Minimum protection probability; A co-evolutionary framework is adopted, which combines an elite retention strategy with a population diversity maintenance mechanism, in which elite individuals directly enter the next generation, while the remaining individuals are generated through a tournament selection mechanism. A constraint handling mechanism based on KKT conditions is established, which dynamically adjusts the penalty coefficient of infeasible solutions through Lagrange multipliers and applies an exponentially increasing penalty term to individuals that violate the SOC safety boundary.

4. The energy storage charging station management method with dynamic load monitoring as described in claim 3, characterized in that, The dynamic impedance matching algorithm includes: Establish an equivalent impedance model for the grid-energy storage interface and define a dynamic matching degree function: Among them, Z grid (t) represents the equivalent impedance on the grid side, Z ess (t) represents the output impedance of the energy storage system, both of which are measured in real time using the online frequency sweep method; Construct an objective optimization function that includes harmonic suppression and transient stability terms: Where h is the harmonic frequency, I h Let be the h-th harmonic current component, α and β be weighting coefficients, and dM(t) / dt represent the rate of change of the matching degree. Multi-step rolling optimization is performed using a model predictive control (MPC) framework, within each control cycle: Predict the trend of power grid impedance changes over multiple future time steps; Establish a quadratic programming problem to solve for the optimal impedance adjustment; The control signal for the energy storage converter is generated by a pulse width modulator.

5. The energy storage charging station management method with dynamic load monitoring as described in claim 4, characterized in that, The energy storage lifetime attenuation compensation module performs the following operations: Establish a battery life prediction model and calculate the real-time life degradation coefficient: Where A is a material constant, and E a Let R be the activation energy, T be the gas constant, T be the battery temperature, and SOC(t) be the state of charge at the current moment. ref For reference state of charge, I rms The current is the effective value, and the exponent γ is an empirical parameter used to adjust the effect of the state of charge on lifetime decay. Construct a dynamic compensation strategy: When the lifetime degradation coefficient D(t) exceeds the threshold, the maximum charge and discharge power limit is automatically reduced. Adjust the SOC operating range according to the real-time attenuation rate, and give priority to using the low attenuation rate region; Generate lifetime balancing scheduling instructions to dynamically distribute current among parallel battery clusters; Implement a multi-timescale compensation mechanism: Second-level time scale: Eliminate instantaneous power surges through a dynamic internal resistance compensation algorithm; Hourly timescale: Adjusting depth of charge / discharge (DOD) and cycle count allocation; Monthly timescale: Optimize battery pack rotation strategy and capacity reconfiguration.

6. The energy storage charging station management method with dynamic load monitoring as described in claim 2, characterized in that, The improvements to the dynamic time warping algorithm include: An adaptive regularization window constraint is introduced, where the window width is dynamically adjusted based on the intensity of data fluctuations. The expression is: Among them, W base Where σ is the base window width, σ is the standard deviation of the current data segment, and k is the sensitivity coefficient; Multi-objective regularization path optimization is adopted to minimize time offset error and feature morphology differences; For non-matching data segments, cubic spline interpolation is applied to reconstruct the data while preserving the original statistical characteristics.

7. The energy storage charging station management method with dynamic load monitoring as described in claim 3, characterized in that, The calculation of the SOC stability index in the composite fitness function includes: Calculate the standard deviation σ of the SOC trajectory SOC ; Assess the cumulative time T of the SOC crossing the safety boundary. over ; The stability index is defined as: , where ε is a minimal constant to prevent division by zero, and w is a weighting factor.

8. The energy storage charging station management method with dynamic load monitoring as described in claim 4, characterized in that, The steps of multi-step rolling optimization using the Model Predictive Control (MPC) framework include: Establish a prediction model library that includes power grid impedance prediction models, energy storage response models, and constraint condition sets; At the beginning of each control cycle, the latest measurement data is loaded and the predicted state variables are initialized; The optimal control sequence is solved by using a sequential quadratic programming algorithm, and the first three control variables that minimize the objective optimization function are selected. The feasibility of the solution is verified, and the slack variable compensation mechanism is activated when a constraint conflict occurs.

9. A dynamic load monitoring energy storage charging station management system, used to implement the management method as described in any one of claims 1 to 8, characterized in that, The management system includes: The data acquisition module is a multi-source sensor array deployed at the grid access point and the energy storage unit of the energy storage charging station. It is used to collect charging power demand data, energy storage system state of charge (SOC) and grid feeder load characteristic data in real time, and to build a dynamically updated load characteristic database. The load forecasting module generates a real-time load feature matrix of the grid-energy storage joint system based on the load feature data, using a sliding time window mechanism and a dynamic weight allocation algorithm. The dynamic scheduling module establishes a dynamic scheduling model based on two-layer optimization. The upper-layer model uses a genetic algorithm to optimize the charging and discharging timing of the energy storage system, while the lower-layer model uses a dynamic priority evaluation function to perform hierarchical scheduling of charging requests. The adaptive peak shaving control module adjusts the output power of the energy storage system and generates a dynamic charging power allocation command when it detects that the peak load of the power grid exceeds a preset threshold. The closed-loop feedback control module includes a power grid frequency fluctuation compensation subunit and an energy storage lifetime decay compensation subunit. It uses a sliding time window mechanism to update the load forecasting model parameters and corrects the constraints of the dynamic scheduling model based on real-time operating data.

10. A storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the energy storage charging station management method based on dynamic load monitoring as described in any one of claims 1 to 8.

Images