Method and system for optimized dispatching of operating grid for mobile storage and charging system

Abstract

This invention discloses a method and system for optimizing the scheduling of mobile energy storage and charging systems within an operational grid, relating to the field of energy internet scheduling technology. The method includes: dividing the target operational area into a target operational grid set and monitoring multi-source information in real time; performing predictive analysis to obtain the target load curve and carbon emission analysis to obtain the target carbon emission curve; performing scheduling optimization analysis on the target grid to form a target scheduling decision; and executing operational scheduling for the target mobile energy storage and charging devices within the target grid based on the target scheduling decision. This invention solves the technical problems of extensive energy storage scheduling and insufficient precision in regional load control in existing technologies, leading to difficulties in coordinating low-carbon development and scheduling optimization. It achieves refined scheduling and low-carbon coordinated management of the target operational area grid, improving the scheduling adaptability and load balancing efficiency of mobile energy storage and charging systems.

Description

Technical Field

[0001] This invention relates to the field of energy internet dispatching technology, specifically to an operational grid optimization dispatching method and system for mobile energy storage and charging systems. Background Technology

[0002] Current grid-based operation and dispatching of mobile energy storage and charging systems primarily focuses on single load balancing or equipment efficiency as control objectives, lacking a holistic consideration of multi-source information fusion monitoring and coordinated carbon emission constraints. Regional grid division is coarse, and load forecasting and carbon emission analysis are independent, making dynamic matching difficult. Furthermore, existing dispatching strategies do not adequately incorporate adaptability assessments based on differences in renewable energy structures, resulting in highly subjective and inaccurate dispatching decisions. This leads to problems such as insufficient load fluctuation suppression, low renewable energy absorption rates, and poor cross-grid dispatching coordination. Moreover, it fails to simultaneously ensure stable power supply while meeting low-carbon operation requirements, hindering the full realization of the dispatching flexibility and comprehensive regional energy operation benefits of mobile energy storage and charging systems.

[0003] Existing technologies for energy storage dispatch are crude and lack precision in regional load regulation, leading to technical problems that make it difficult to coordinate and optimize low-carbon development and dispatch. Summary of the Invention

[0004] This application provides an operational grid optimization scheduling method and system for mobile energy storage and charging systems, which addresses the technical problems of extensive energy storage scheduling and insufficient regional load control accuracy in the prior art, making it difficult to coordinate and optimize low carbon emissions and scheduling.

[0005] In view of the above problems, this application provides an operational grid optimization scheduling method and system for mobile energy storage and charging systems.

[0006] A first aspect of this application provides an operational grid optimization scheduling method for mobile storage and charging systems, the method comprising: The target operating area is divided into target operating grid sets, and the target multi-source information of the target grids in the target operating grid sets is monitored in real time. The target multi-source information is used for predictive analysis to obtain the target load curve, and a predetermined carbon emission mechanism is invoked to perform carbon emission analysis on the target load curve to obtain the target carbon emission curve. The target load curve and the target carbon emission curve are used as scheduling constraints, and a predetermined scheduling fitness evaluation function is used as the scheduling evaluation method to perform scheduling optimization analysis on the target grid to form a target scheduling decision. The target mobile storage and charging equipment in the target grid is operated and scheduled according to the target scheduling decision.

[0007] A second aspect of this application provides an operational grid optimization scheduling system for mobile storage and charging systems, the system comprising: The system includes a target multi-source information acquisition module, which divides the target operating area into a target operating grid set and monitors the target multi-source information of the target grids in the target operating grid set in real time; a target carbon emission curve acquisition module, which predictively analyzes the target multi-source information to obtain a target load curve and retrieves a predetermined carbon emission mechanism to perform carbon emission analysis on the target load curve to obtain a target carbon emission curve; a target scheduling decision generation module, which uses the target load curve and the target carbon emission curve as scheduling constraints and a predetermined scheduling fitness evaluation function as a scheduling evaluation method to perform scheduling optimization analysis on the target grid and form a target scheduling decision; and an operation scheduling execution module, which performs operation scheduling execution on the target mobile storage and charging equipment in the target grid according to the target scheduling decision.

[0008] One or more technical solutions provided in this application have at least the following technical effects or advantages: The target operating area is divided into target operating grid sets, and multi-source information of target grids within these grids is monitored in real time. Predictive analysis is performed to obtain target load curves, and a predetermined carbon emission mechanism is invoked to analyze these load curves, resulting in target carbon emission curves. A predetermined scheduling fitness evaluation function is used as the scheduling evaluation method to perform scheduling optimization analysis on the target grids, forming target scheduling decisions. Operational scheduling is then executed for the target mobile storage and charging equipment within the target grids. This achieves refined scheduling and low-carbon collaborative management of the target operating area grids, improving the scheduling adaptability and load balancing efficiency of the mobile storage and charging system. Attached Figure Description

[0009] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0010] Figure 1 A schematic diagram of the operation grid optimization scheduling method for mobile energy storage and charging systems provided in this application embodiment; Figure 2 This is a schematic diagram of the structure of an operational grid optimization scheduling system for mobile energy storage and charging systems provided in an embodiment of this application.

[0011] Figure labeling: Target multi-source information acquisition module 10, Target carbon emission curve acquisition module 20, Target scheduling decision generation module 30, Operation scheduling execution module 40. Detailed Implementation

[0012] This application provides an operational grid optimization scheduling method and system for mobile energy storage and charging systems, which addresses the technical problems in existing technologies such as extensive energy storage scheduling, insufficient precision in regional load regulation, and difficulty in coordinating and optimizing low-carbon practices and scheduling.

[0013] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0014] Example 1, as Figure 1 As shown, this application provides an operational grid optimization scheduling method for mobile storage and charging systems, the method comprising: Step S100: Divide the target operating area into a target operating grid set, and monitor the target multi-source information of the target grids in the target operating grid set in real time.

[0015] Specifically, based on the geographical distribution, electricity load density, distributed power supply access, and service range of mobile energy storage and charging equipment in the target operating area, the target operating area is divided and numbered into grids according to a unified spatial scale, forming a target operating grid set that covers the entire operating area without overlap or omission. Through sensing and monitoring units deployed on the grid side, user side, and mobile energy storage and charging equipment, real-time monitoring of each target grid in the target operating grid set is carried out, collecting multi-dimensional operating data including real-time load power, real-time photovoltaic output power, real-time state of charge of target mobile energy storage and charging equipment, historical electricity consumption records, user distribution dispersion, and grid operating status. After data cleaning and standardization, target multi-source information that can comprehensively characterize the current operating status of the target grid is formed.

[0016] Step S200: Perform predictive analysis on the target multi-source information to obtain the target load curve, and retrieve the predetermined carbon emission mechanism to perform carbon emission analysis on the target load curve to obtain the target carbon emission curve.

[0017] Specifically, based on the acquired target multi-source information, the target grid's historical electricity consumption records are first extracted, analyzed to obtain the historical electricity consumption per unit time window, and a target historical electricity consumption time series is formed. Through incremental transformation analysis, the target historical load time series is obtained. Then, the real-time unit load in the target multi-source information is fitted with the target historical load time series using polynomial regression to obtain the target historical load curve. Multiple prediction time points are constructed based on the unit time window, and multiple prediction loads are obtained by combining them with the target historical load curve. An initial target load curve is formed based on the correspondence between the prediction load and the prediction time points. Subsequently, based on predetermined state characteristics, including the target grid's historical total load demand, photovoltaic output, and the charging and discharging power of the target mobile energy storage and charging equipment within the unit time window, historical state analysis is performed to obtain target historical state characteristic parameters, and the target state is constructed. The transition matrix is ​​used to predict the prior state covariance by comparing the target state estimate with the target state transition matrix. A target observation matrix is ​​constructed by combining the real-time load power, real-time photovoltaic output power, and real-time state of charge of the target mobile energy storage and charging equipment from the target multi-source information. The Kalman gain matrix is ​​then calculated using the target observation matrix, and the prior estimate covariance is corrected to obtain the posterior estimate covariance. The target user distribution dispersion from the target multi-source information is extracted as a regulating factor to adjust the posterior estimate covariance, thus completing the accurate calibration of the target load curve. Simultaneously, a target carbon potential transfer model based on carbon emission flow theory is retrieved, and the carbon potential data corresponding to the unit time window is obtained in conjunction with the total target carbon emissions. Carbon emission analysis is then performed on the calibrated target load curve, ultimately forming a target carbon emission curve characterizing the temporal variation of carbon emissions in the target grid.

[0018] Step S300: Using the target load curve and the target carbon emission curve as scheduling constraints, and using a predetermined scheduling fitness evaluation function as the scheduling evaluation method, perform scheduling optimization analysis on the target grid to form a target scheduling decision.

[0019] Specifically, the target load curve and target carbon emission curve are used together as constraints for the optimized scheduling of the mobile energy storage and charging system. Each time point within a unit time window is traversed, and the target load value and target carbon emission value at the corresponding time are matched sequentially in the target load curve and target carbon emission curve. A variation-weighted calculation is performed on the load value and carbon emission value to obtain the scheduling demand index corresponding to the current time point. When the scheduling demand index reaches a predetermined threshold, a scheduling signal is triggered. After the scheduling signal is triggered, a predetermined scheduling fitness evaluation function is used as the evaluation standard for the scheduling scheme. A predetermined renewable energy benchmark containing multiple energy types with proportional identifiers is read. The target renewable energy information of the target grid is similar to the predetermined renewable energy benchmark to obtain the target similarity. The predetermined scheduling action value, including the value of scheduling actions within the grid and the value of scheduling actions across grids, is weighted using the target similarity as a coefficient to obtain the estimated value of the target scheduling action, which is then used as the scheduling fitness. By comparing and filtering the scheduling fitness of each candidate scheduling strategy, the scheduling strategy corresponding to the maximum scheduling fitness is selected as the optimal scheme, ultimately forming a target scheduling decision to guide the scheduling execution of mobile energy storage and charging equipment.

[0020] Step S400: Perform operational scheduling execution on the target mobile storage and charging equipment in the target grid according to the target scheduling decision.

[0021] Specifically, the generated target scheduling decisions are sent to the scheduling execution units corresponding to the target operation grid. According to the scheduling strategies, timings, and parameters determined by the target scheduling decisions, targeted operation scheduling is carried out on the mobile energy storage and charging equipment deployed within the target grid. This includes real-time control and adjustment of the charging and discharging power, operating mode, scheduling location, and switching status of the target mobile energy storage and charging equipment. It also involves synchronously matching the load demand, photovoltaic output status, and carbon emission constraints of the target grid to complete the execution of scheduling actions within or across grids, thereby achieving refined and adaptive optimization scheduling of the mobile energy storage and charging system in the target operation grid.

[0022] In one possible implementation, step S200 further includes: Step S210: Obtain the target historical electricity consumption record of the target grid.

[0023] Step S220: Analyze the target historical electricity consumption records to obtain the historical electricity consumption per unit time window, and form the target historical electricity consumption time series.

[0024] Step S230: Perform incremental conversion analysis on the target historical electricity consumption time series to obtain the target historical load time series.

[0025] Step S240: Perform polynomial regression fitting on the real-time unit load in the target multi-source information and the target historical load time series to obtain the target historical load curve.

[0026] Step S250: Based on the unit time window, construct multiple prediction time points and combine them with the target historical load curve to obtain multiple prediction loads.

[0027] Step S260: Based on the correspondence between the multiple predicted loads and the multiple predicted time points, form the target load curve.

[0028] Specifically, by using the smart electricity consumption collection terminals, power grid metering management system and energy storage operation database deployed in the target operating area, the electricity consumption data of the target grid within the preset historical statistical period is retrieved in a targeted manner, including the electricity consumption, power consumption, load fluctuation status and equipment electricity consumption details at each time node. After data integrity verification and outlier removal, a target historical electricity consumption record covering the complete historical period and characterizing the real electricity consumption characteristics of the target grid is formed.

[0029] Using a preset unit time window as the statistical granularity, the target historical electricity consumption records are analyzed and accumulated segment by segment. The total electricity consumption within each unit time window is extracted and the numerical normalization process is completed. All historical electricity consumption of unit time windows arranged in chronological order are combined sequentially to form a target historical electricity consumption time series that can reflect the electricity consumption pattern of the target grid and has a chronological order.

[0030] Incremental conversion analysis is performed on the target historical electricity consumption time series using differential operation between adjacent time windows. The difference in electricity consumption between the current unit time window and the previous unit time window is calculated sequentially. This difference is used as the load change increment. Then, power conversion is performed by combining the grid rated power and the time window coefficient to convert the electricity consumption dimension data into load power dimension data. Finally, the target historical load time series representing the load amplitude change at each moment of the target grid is generated.

[0031] The real-time unit load and historical load time series of the target from multi-source information are preprocessed by time axis alignment, outlier removal, and amplitude normalization. A multivariate polynomial regression model is constructed with time as the independent variable and load as the dependent variable. The historical load time series of the target is used as the fitting sample set, and the real-time unit load is substituted into the model as the correction constraint term. The least squares method is used to calculate the regression coefficient that minimizes the sum of squared residuals, and the optimal polynomial function expression is determined. The entire time series is smoothed and fitted using this function, and finally, a continuous, complete target historical load curve that fits the historical pattern and the real-time state is output.

[0032] Using a defined unit time window as a fixed time interval, multiple consecutive forecast time points are generated and arranged sequentially according to the equal time interval rule within the forecast period. The time variables of each forecast time point are substituted into the polynomial function corresponding to the fitted target historical load curve for numerical calculation. The forecast load amplitude corresponding to each forecast time point is directly solved, thereby obtaining multiple forecast loads that correspond one-to-one with each forecast time point.

[0033] Multiple predicted time points are used as the horizontal axis, and multiple predicted load values ​​corresponding to them are used as the vertical axis. They are arranged in an orderly manner and connected point by point in a unified time coordinate system. Linear interpolation or smooth interpolation methods are used to fit curves to adjacent predicted points, fill in the intermediate load values ​​within the time interval, and finally generate a continuous, smooth and complete target load curve that represents the future load change trend of the target grid.

[0034] In one possible implementation, step S260 further includes: Step S261: Perform historical state analysis on the target mesh based on predetermined state characteristics to obtain target historical state characteristic parameters.

[0035] Step S262: Construct the target state transition matrix based on the target historical state feature parameters.

[0036] Step S263: Perform prior state prediction on the target state estimate obtained by analyzing the target historical state characteristic parameters and the target state transition matrix to obtain the prior estimate covariance.

[0037] Step S264: Based on the real-time load power, real-time photovoltaic output power and real-time state of charge of the target mobile energy storage and charging device in the target multi-source information, construct the target observation matrix.

[0038] Step S265: Calculate the Kalman gain matrix using the target observation matrix, and correct the prior estimated covariance based on the Kalman gain matrix to obtain the posterior estimated covariance.

[0039] Step S266: Calibrate the target load curve based on the posterior estimated covariance.

[0040] Specifically, based on preset state characteristics, the analysis includes the historical total load demand, photovoltaic output, and charging and discharging power of the target mobile energy storage device within a unit time window. The corresponding data is extracted from the historical operation database of the target grid in chronological order window by window. The average and peak values ​​of the historical total load demand within each unit time window are statistically analyzed, the output amplitude and fluctuation of the photovoltaic output are calculated, and the direction and amplitude of the charging and discharging power of the target mobile energy storage device are marked. At the same time, abnormal sampling points are removed and missing data is linearly completed. Finally, the multidimensional feature values ​​after the above statistics and processing are integrated in time series to form the target historical state feature parameters that can completely characterize the historical operation status of the target grid.

[0041] Using the target's historical state characteristic parameters as input data, a multi-dimensional state vector consisting of historical total load demand, photovoltaic output, and the charging and discharging power of the target mobile energy storage and charging equipment is used as a state node. The state change samples of all adjacent unit time windows are traversed, the transition frequency between each state vector is counted and normalized, and the state transition probability and amplitude gain coefficient are calculated. The matrix is ​​then arranged in an orderly manner according to the rule that rows correspond to the state at the previous moment and columns correspond to the state at the next moment, to construct the target state transition matrix, which is used to quantitatively describe the dynamic evolution relationship of the grid's operating state over time.

[0042] Based on the target's historical state characteristic parameters, a time-series recursive operation is performed to obtain the target state estimate at the current moment. This target state estimate is used as the initial input vector and substituted into the Kalman filter prior prediction algorithm. A matrix multiplication operation is performed with the target state transition matrix to complete the prior prediction of the state at the next moment. At the same time, combined with the system process noise covariance matrix, the prior estimation error is iteratively calculated through the covariance propagation formula. That is, the algorithm logic is that the prior estimation covariance is equal to the product of the state transition matrix and the posterior covariance matrix, and then the process noise covariance matrix is ​​superimposed. Finally, the prior estimation covariance that characterizes the uncertainty distribution of the predicted state is obtained.

[0043] The real-time load power, real-time photovoltaic output power, and real-time state of charge of the target mobile energy storage and charging equipment are taken as three-dimensional observation variables from the multi-source information of the target. According to the construction rules of the Kalman filter observation model, the observation vector is established with time series as index. Then, based on the linear mapping relationship between each observation variable and the grid system state, a target observation matrix with dimensions matching the state vector is constructed. Each column of this matrix corresponds to the observation coefficients of real-time load power, real-time photovoltaic output power, and real-time state of charge. The values ​​of each coefficient are determined by least squares fitting of historical measured data, so that the matrix can accurately reflect the quantitative correlation between the observation variables and the system state, thereby completing the construction of the target observation matrix.

[0044] The target observation matrix, prior estimated covariance, and preset observation noise covariance matrix are substituted into the Kalman gain calculation formula. The Kalman gain matrix is ​​obtained by matrix inversion and multiplication. Then, the Kalman gain matrix, the target observation matrix, and the prior estimated covariance are used for iterative correction. The posterior estimated covariance is calculated by multiplying the prior estimated covariance by the identity matrix minus the product of the Kalman gain matrix and the target observation matrix. This process updates the prediction error covariance and finally yields a more accurate posterior estimated covariance that better reflects the actual observation data.

[0045] The covariance of the a posteriori estimate is used as a quantitative indicator of the prediction error. The correction coefficient for each prediction time point on the target load curve is determined based on the magnitude of the covariance. The larger the covariance, the lower the prediction reliability and the larger the correction magnitude at that point, and vice versa. The correction coefficient is weighted with the prediction load value at the corresponding time, and the amplitude of the target load curve is corrected point by point. At the same time, abnormal fluctuation points are smoothed, and finally a calibrated target load curve with higher accuracy that fits the actual operating state is obtained.

[0046] In one possible implementation, step S261 further includes: The predetermined state characteristics include the historical total load demand of the target grid within a unit time window, photovoltaic output, and the charging and discharging power of the target mobile energy storage device.

[0047] Specifically, the predetermined state characteristics include the historical total load demand, photovoltaic output, and charging / discharging power of the target grid within a unit time window. The unit time window is a pre-defined fixed-duration statistical interval used for segmented collection and analysis of grid operation data. The historical total load demand refers to the sum of active power consumed by all electrical loads within the target grid within this time window, reflecting the overall regional electricity demand and load level. Photovoltaic output refers to the actual active power output of distributed photovoltaic power stations within the target grid within this time window, reflecting the real-time supply capacity and fluctuation characteristics of photovoltaic power generation. The charging / discharging power of the target mobile energy storage device refers to the real-time power value of the mobile energy storage device connected to the target grid during charging or discharging within this time window; charging power is negative and discharging power is positive, used to characterize the energy interaction intensity and operating status between energy storage and the grid. These characteristics together constitute a set of key parameters for characterizing the historical operating state of the target grid.

[0048] In one possible implementation, step S265 further includes: Extract the target user distribution dispersion from the target multi-source information.

[0049] The posterior estimated covariance is adjusted using the dispersion of the target user distribution as a regulating factor.

[0050] Specifically, the coordinates of power consumption nodes and load data from the target multi-source information are collected traversally. Using the geographical coordinates of each node as samples, the spatial distribution standard deviation is calculated using a sample standard deviation algorithm. Simultaneously, the coefficient of variation is calculated using the load power of each node as samples. The spatial distribution standard deviation and the load coefficient of variation are normalized and then weighted and fused using the formula D=α⋅σ. xy +β⋅C v The dispersion D of the target user distribution is calculated, where α and β are weighting coefficients, and σ xy C represents the standard deviation of the spatial distribution. v This is the load power variation coefficient, which enables accurate quantitative extraction of the dispersion of the target user distribution.

[0051] The dispersion of the target user distribution is normalized to the [0, 1] interval to obtain an adaptive adjustment factor. This adjustment factor is then superimposed with a preset benchmark adjustment coefficient to construct the final correction coefficient. The correction coefficient is then multiplied element by element with the posterior estimated covariance matrix to adaptively scale the covariance value. The more dispersed the user distribution, the larger the covariance magnitude to reflect higher state uncertainty; the more concentrated the distribution, the smaller the covariance. Finally, the adaptive adjustment of the posterior estimated covariance is completed.

[0052] In one possible implementation, step S200 further includes: The predetermined carbon emission mechanism refers to constructing a target carbon potential transfer model for the target operating area based on carbon emission flow theory, and obtaining carbon potential data corresponding to a unit time window in conjunction with the target total carbon emissions to form the target carbon emission curve.

[0053] Specifically, the predetermined carbon emission mechanism refers to constructing a target carbon potential transfer model for the target operating area based on carbon emission flow theory, and coordinating with the target total carbon emissions to obtain carbon potential data corresponding to a unit time window, thereby forming a target carbon emission curve. Among them, the carbon emission flow theory is used to characterize the source, distribution and transmission patterns of carbon emissions in the power system along with the transmission of electricity, linking carbon emissions on the generation side with electricity consumption behavior on the load side, and realizing the quantitative tracking of carbon emissions in the power grid; the target operating area is the power supply range in which the energy storage system participates in the dispatch, including the complete operating area consisting of power sources, power grid, loads and energy storage equipment; the target carbon potential transfer model takes the carbon emissions, power flow distribution and energy flow direction of each node in the area as inputs, constructs the mapping relationship between carbon potential and power flow, and simulates the transmission and transfer process of carbon emissions between different nodes and branches; the target total carbon emissions are the total amount of carbon emissions generated by power generation, electricity consumption and equipment losses in the target operating area within the statistical period; the unit time window is a pre-set fixed time interval used to collect and calculate carbon potential data in segments; the carbon potential data is a quantitative indicator characterizing the regional carbon emission intensity and carbon transfer potential energy within the unit time window, reflecting the regional carbon emission level and trend during that period; by arranging the carbon potential data of continuous unit time windows in a time sequence, the target carbon emission curve reflecting the dynamic change pattern of carbon emissions in the target operating area is finally formed.

[0054] In one possible implementation, step S300 further includes: Step S310: Obtain any point in time within the unit time window.

[0055] Step S320: Sequentially match the arbitrary load and arbitrary carbon emission corresponding to any time point in the target load curve and the target carbon emission curve.

[0056] Step S330: Perform a variation-weighted calculation on the arbitrary load and the arbitrary carbon emissions to obtain the arbitrary scheduling demand index corresponding to the arbitrary time point.

[0057] Step S340: When the arbitrary scheduling demand index is at a predetermined threshold, a scheduling signal is issued.

[0058] Step S350: Based on the scheduling signal, calculate the scheduling fitness corresponding to the predetermined scheduling strategy using the predetermined scheduling fitness evaluation function.

[0059] Step S360: Compare the scheduling fitness to obtain the scheduling strategy corresponding to the maximum scheduling fitness, and use it as the target scheduling decision.

[0060] Specifically, a pre-defined unit time window is used as the analysis interval. According to the time series sampling rules, a moment is randomly selected from the start time to the end time of the time window as the analysis object. The accurate timestamp of that moment is obtained and used as the reference time point for subsequent load, carbon emission matching and scheduling demand calculations, ensuring that the scheduling analysis has a clear time series positioning.

[0061] Using the timestamp of any obtained time point as the matching benchmark, firstly, time series retrieval and interpolation calculation are performed in the generated target load curve to locate and extract the real-time load value corresponding to that time point, which is recorded as arbitrary load; then, the same time series matching method is used in the target carbon emission curve, and the real-time carbon emission value corresponding to the same time point is obtained through curve interpolation and point alignment, which is recorded as arbitrary carbon emission, thereby completing the synchronous matching of load data and carbon emission data at the same time.

[0062] First, calculate the coefficient of variation of any load relative to the baseline load and the coefficient of variation of any carbon emission relative to the baseline carbon emission. Then, assign the first weight to the load coefficient of variation and the second weight to the carbon emission coefficient of variation according to the scheduling priority. Add the two weighted results to obtain the scheduling demand index. Specifically, the calculation is completed by the formula: Scheduling demand index = load coefficient of variation × first weight + carbon emission coefficient of variation × second weight, where the coefficient of variation is the absolute value of the difference between the real-time value and the baseline value divided by the baseline value, thereby quantifying the scheduling urgency at that point in time.

[0063] The calculated arbitrary scheduling demand index is compared with a preset threshold range. The threshold judgment logic determines whether the index falls within the preset trigger range. If the arbitrary scheduling demand index is greater than or equal to the lower threshold and less than or equal to the upper threshold, the scheduling trigger condition is met at the current moment. The system immediately generates and outputs a scheduling signal to start the subsequent scheduling strategy evaluation and decision-making process. If it is not within the preset threshold range, scheduling is not triggered, and the current running state remains unchanged.

[0064] Upon receiving the dispatch signal, each preset dispatch strategy is read sequentially. The energy storage charging and discharging power, load offset, carbon emission increment, and grid loss value corresponding to each strategy are input into the preset dispatch fitness evaluation function. This function first normalizes the three indicators of load smoothing effect, carbon emission reduction, and energy storage operation efficiency, and then weights and sums them according to the load weight W1, carbon emission weight W2, and energy storage efficiency weight W3. The fitness value F is calculated using the formula F=W1・F1+W2・F2+W3・F3, where F1 is the load fluctuation suppression rate, F2 is the carbon emission reduction rate, and F3 is the energy storage energy conversion efficiency. The calculation is performed for each strategy one by one to obtain the corresponding dispatch fitness.

[0065] The calculated scheduling fitness values ​​corresponding to each predetermined scheduling strategy are traversed and compared. The strategy with the largest scheduling fitness value is selected by sorting the values. This strategy represents the optimal scheduling scheme under the current load, carbon emission and energy storage operation conditions. It is determined and output as the target scheduling decision of this scheduling process, which is used to guide the target mobile energy storage and charging equipment and the power grid to perform corresponding charging, discharging and load regulation operations.

[0066] In one possible implementation, step S350 further includes: Step S351: Read the predetermined renewable energy benchmark, wherein the predetermined renewable energy benchmark includes a variety of energy types with proportional identifiers.

[0067] Step S352: Obtain the target renewable energy information of the target grid and perform similarity analysis with the various energy types with proportional identifiers to obtain the target similarity.

[0068] Step S353: Use the target similarity as a coefficient to perform a weighted calculation on the value of the predetermined scheduling action to obtain an estimated value of the target scheduling action.

[0069] Step S354: Use the estimated value of the target scheduling action as the scheduling fitness.

[0070] Specifically, the pre-configured renewable energy benchmark is read from the system's preset parameter library. This benchmark defines a variety of energy types and their corresponding ideal proportions in a standardized manner, covering typical clean energy types such as photovoltaic, wind power, hydropower, and biomass energy. Each energy type is marked with its recommended installed capacity ratio, output ratio, and other quantitative proportion indicators in the target grid, which are used as reference standards for subsequent energy structure matching and scheduling value assessment.

[0071] Data on the actual output ratio of various renewable energy sources such as photovoltaic, wind power, and hydropower in the target grid at the current scheduling time is collected to form a real-time energy ratio vector with unified dimensions. Then, the cosine similarity of this vector with various standard energy type vectors with ratio identifiers in the predetermined renewable energy benchmark is calculated. By calculating the ratio difference of each dimension separately and summing it after normalization, the target similarity between the real-time energy structure and the benchmark energy structure is quantified, thereby reflecting the degree of closeness between the current renewable energy structure and the benchmark structure.

[0072] The obtained target similarity is used as a weighting coefficient and weighted by the value of the basic scheduling action corresponding to each predetermined scheduling action. Specifically, the basic value of the scheduling action is scaled and corrected by directly multiplying the target similarity by the value of the predetermined scheduling action, so as to obtain the estimated value of the target scheduling action that takes into account the matching degree of the current renewable energy structure, making the value assessment result more in line with the actual energy structure characteristics of the power grid.

[0073] The calculated estimated value of the target scheduling action is directly assigned as the scheduling fitness of the corresponding scheduling strategy. This serves as a quantitative indicator to measure the overall merits of the scheduling strategy in adapting to the current renewable energy structure and meeting load and carbon emission constraints, providing a unified evaluation basis for subsequent selection of the optimal scheduling decision.

[0074] In one possible implementation, step S350 further includes: The predetermined scheduling action value includes the intra-grid scheduling action value and the cross-grid scheduling action value.

[0075] Specifically, the predetermined scheduling action value includes the intra-grid scheduling action value and the cross-grid scheduling action value. The predetermined scheduling action value refers to the comprehensive quantitative result of various scheduling actions, such as energy storage charging and discharging, load regulation, and energy interaction, in terms of economy, low carbon emissions, and stability, used to directly evaluate the merits of the scheduling actions. The intra-grid scheduling action value refers to the value generated by executing scheduling actions only within the current target grid, considering only load balancing, carbon emission control, renewable energy consumption, and equipment operating costs within the region, without involving external grid interactions. The cross-grid scheduling action value refers to the value generated by cross-grid scheduling behaviors such as energy transmission, carbon emission coordination, and load support between the current target grid and adjacent regional power grids, including comprehensive quantitative indicators such as cross-grid energy loss, transmission revenue, inter-regional carbon emission coordination benefits, and tie-line scheduling costs.

[0076] Example 2, based on the same inventive concept as the operation grid optimization scheduling method for mobile storage and charging systems in the foregoing examples, such as... Figure 2 As shown, this application provides an operational grid optimization scheduling system for mobile storage and charging systems. The system and method embodiments in this application are based on the same inventive concept. The system includes: The target multi-source information acquisition module 10 is used to divide the target operation area into a target operation grid set and to monitor and obtain the target multi-source information of the target grid in the target operation grid set in real time.

[0077] The target carbon emission curve acquisition module 20 is used to predict and analyze the target multi-source information to obtain the target load curve, and to call a predetermined carbon emission mechanism to perform carbon emission analysis on the target load curve to obtain the target carbon emission curve.

[0078] The target scheduling decision generation module 30 is used to perform scheduling optimization analysis of the target grid using the target load curve and the target carbon emission curve as scheduling constraints and a predetermined scheduling fitness evaluation function as the scheduling evaluation method, so as to form a target scheduling decision.

[0079] The operation scheduling execution module 40 is used to perform operation scheduling execution on the target mobile storage and charging equipment in the target grid according to the target scheduling decision.

[0080] Furthermore, the system is also used to implement the following functions: The process involves: acquiring the target historical electricity consumption records of the target grid; analyzing the target historical electricity consumption records to obtain the historical electricity consumption per unit time window, forming a target historical electricity consumption time series; performing incremental transformation analysis on the target historical electricity consumption time series to obtain a target historical load time series; performing polynomial regression fitting between the real-time unit load in the target multi-source information and the target historical load time series to obtain a target historical load curve; constructing multiple prediction time points based on the unit time window, and combining them with the target historical load curve to obtain multiple prediction loads; and forming the target load curve based on the correspondence between the multiple prediction loads and the multiple prediction time points.

[0081] Furthermore, the system is also used to implement the following functions: Historical state analysis is performed on the target grid based on predetermined state characteristics to obtain target historical state characteristic parameters; a target state transition matrix is ​​constructed based on the target historical state characteristic parameters; prior state prediction is performed on the target state estimate obtained from the analysis of the target historical state characteristic parameters and the target state transition matrix to obtain the prior estimate covariance; a target observation matrix is ​​constructed based on the real-time load power, real-time photovoltaic output power and real-time state of charge of the target mobile energy storage and charging equipment in the target multi-source information; the Kalman gain matrix is ​​introduced into the target observation matrix to calculate the Kalman gain matrix, and the prior estimate covariance is corrected according to the Kalman gain matrix to obtain the posterior estimate covariance; the target load curve is calibrated based on the posterior estimate covariance.

[0082] Furthermore, the system is also used to implement the following functions: The predetermined state characteristics include the historical total load demand of the target grid within a unit time window, photovoltaic output, and the charging and discharging power of the target mobile energy storage device.

[0083] Furthermore, the system is also used to implement the following functions: Extract the target user distribution dispersion from the target multi-source information; adjust the posterior estimation covariance using the target user distribution dispersion as an adjustment factor.

[0084] Furthermore, the system is also used to implement the following functions: The predetermined carbon emission mechanism refers to constructing a target carbon potential transfer model for the target operating area based on carbon emission flow theory, and obtaining carbon potential data corresponding to a unit time window in conjunction with the target total carbon emissions to form the target carbon emission curve.

[0085] Furthermore, the system is also used to implement the following functions: Obtain any time point within a unit time window; sequentially match any load and any carbon emission corresponding to the arbitrary time point in the target load curve and the target carbon emission curve; perform a variation-weighted calculation on the arbitrary load and the arbitrary carbon emission to obtain an arbitrary scheduling demand index corresponding to the arbitrary time point; when the arbitrary scheduling demand index is at a predetermined threshold, issue a scheduling signal; based on the scheduling signal, use the predetermined scheduling fitness evaluation function to calculate the scheduling fitness corresponding to the predetermined scheduling strategy; compare the scheduling fitness to obtain the scheduling strategy corresponding to the maximum scheduling fitness, which is used as the target scheduling decision.

[0086] Furthermore, the system is also used to implement the following functions: Read the predetermined renewable energy benchmark, wherein the predetermined renewable energy benchmark includes multiple energy types with proportional identifiers; obtain the target renewable energy information of the target grid, and perform similarity analysis with the multiple energy types with proportional identifiers to obtain the target similarity; use the target similarity as a coefficient to perform a weighted calculation on the value of the predetermined scheduling action to obtain the estimated value of the target scheduling action; use the estimated value of the target scheduling action as the scheduling fitness.

[0087] Furthermore, the system is also used to implement the following functions: The predetermined scheduling action value includes the intra-grid scheduling action value and the cross-grid scheduling action value.

[0088] It should be noted that the order of the embodiments described above is for descriptive purposes only and does not represent the superiority or inferiority of the embodiments. Specific embodiments of this specification have been described above. Furthermore, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0089] The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

[0090] This specification and accompanying drawings are merely illustrative examples of this application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Therefore, if such modifications and variations fall within the scope of this application and its equivalents, this application intends to include such modifications and variations.

Claims

1. An operational grid optimization scheduling method for mobile energy storage and charging systems, characterized in that, include: The target operating area is divided into target operating grid sets, and the target multi-source information of the target grids in the target operating grid sets is obtained through real-time monitoring; The target load curve is obtained by predictive analysis of the target multi-source information, and the carbon emission analysis of the target load curve is performed by a predetermined carbon emission mechanism to obtain the target carbon emission curve. Using the target load curve and the target carbon emission curve as scheduling constraints, and a predetermined scheduling fitness evaluation function as the scheduling evaluation method, the scheduling optimization analysis of the target grid is performed to form a target scheduling decision; Based on the target scheduling decision, the operation scheduling of the target mobile storage and charging equipment in the target grid is executed.

2. The operation grid optimization scheduling method for mobile storage and charging systems as described in claim 1, characterized in that, The target load curve is obtained by predictive analysis of the multi-source information of the target, including: Obtain the target historical electricity consumption records of the target grid; Analyzing the target's historical electricity consumption records yields the historical electricity consumption per unit time window, forming a time series of the target's historical electricity consumption. Incremental conversion analysis is performed on the target historical electricity consumption time series to obtain the target historical load time series; The real-time unit load in the target multi-source information is fitted with the target historical load time series to obtain the target historical load curve; Multiple forecast time points are constructed based on the unit time window, and multiple forecast loads are obtained by combining the target historical load curve; The target load curve is formed based on the correspondence between the multiple predicted loads and the multiple predicted time points.

3. The operation grid optimization scheduling method for mobile storage and charging systems as described in claim 2, characterized in that, Based on the correspondence between the multiple predicted loads and the multiple predicted time points, the target load curve is formed, and then the process further includes: Based on predetermined state characteristics, the target mesh is analyzed for historical state to obtain target historical state characteristic parameters. Construct a target state transition matrix based on the target historical state feature parameters; A prior state prediction is performed on the target state estimate obtained by analyzing the target historical state characteristic parameters and the target state transition matrix to obtain the prior estimate covariance. Based on the real-time load power, real-time photovoltaic output power, and real-time state of charge of the target mobile energy storage and charging device from the target multi-source information, a target observation matrix is ​​constructed. The Kalman gain matrix is ​​calculated using the target observation matrix, and the prior estimated covariance is corrected based on the Kalman gain matrix to obtain the posterior estimated covariance. The target load curve is calibrated based on the posterior estimated covariance.

4. The operation grid optimization scheduling method for mobile storage and charging systems as described in claim 3, characterized in that, The predetermined state characteristics include the historical total load demand of the target grid within a unit time window, photovoltaic output, and the charging and discharging power of the target mobile energy storage device.

5. The operation grid optimization scheduling method for mobile storage and charging systems as described in claim 3, characterized in that, The target observation matrix is ​​introduced to calculate the Kalman gain matrix, and the prior estimated covariance is corrected based on the Kalman gain matrix to obtain the posterior estimated covariance. The process then includes: Extract the target user distribution dispersion from the target multi-source information; The posterior estimated covariance is adjusted using the dispersion of the target user distribution as a regulating factor.

6. The operation grid optimization scheduling method for mobile storage and charging systems as described in claim 1, characterized in that, The predetermined carbon emission mechanism refers to constructing a target carbon potential transfer model for the target operating area based on carbon emission flow theory, and obtaining carbon potential data corresponding to a unit time window in conjunction with the target total carbon emissions to form the target carbon emission curve.

7. The operational grid optimization scheduling method for mobile storage and charging systems as described in claim 2, characterized in that, Using the target load curve and the target carbon emission curve as scheduling constraints, and a predetermined scheduling fitness evaluation function as the scheduling evaluation method, the scheduling optimization analysis of the target grid is performed to form a target scheduling decision, including: Get any point in time within a unit time window; Match any load and any carbon emission corresponding to any time point in the target load curve and the target carbon emission curve in sequence. The arbitrary scheduling demand index corresponding to any time point is obtained by performing a variation-weighted calculation on the arbitrary load and the arbitrary carbon emissions. When the arbitrary scheduling demand index is at a predetermined threshold, a scheduling signal is issued; Based on the scheduling signal, the scheduling fitness corresponding to the predetermined scheduling strategy is calculated using the predetermined scheduling fitness evaluation function. The scheduling strategy corresponding to the maximum scheduling fitness is obtained by comparing the scheduling fitness values, and this strategy is used as the target scheduling decision.

8. The operation grid optimization scheduling method for mobile storage and charging systems as described in claim 7, characterized in that, Based on the scheduling signal, the scheduling fitness corresponding to the predetermined scheduling strategy is calculated using the predetermined scheduling fitness evaluation function, including: Read the predetermined renewable energy benchmark, wherein the predetermined renewable energy benchmark includes multiple energy types with proportional identifiers; Obtain the target renewable energy information of the target grid and perform similarity analysis with the various energy types with proportional identifiers to obtain the target similarity; The value of the predetermined scheduling action is calculated by weighting the target similarity as a coefficient to obtain the estimated value of the target scheduling action. The estimated value of the target scheduling action is used as the scheduling fitness.

9. The operation grid optimization scheduling method for mobile storage and charging systems as described in claim 8, characterized in that, The predetermined scheduling action value includes the intra-grid scheduling action value and the cross-grid scheduling action value.

10. An operational grid optimization scheduling system for mobile storage and charging systems, characterized in that, The system is used to implement the operational grid optimization scheduling method for mobile storage and charging systems according to any one of claims 1-9, the system comprising: The target multi-source information acquisition module is used to divide the target operation area into a target operation grid set, and to monitor and obtain the target multi-source information of the target grids in the target operation grid set in real time; The target carbon emission curve acquisition module is used to predict and analyze the target multi-source information to obtain the target load curve, and to call a predetermined carbon emission mechanism to perform carbon emission analysis on the target load curve to obtain the target carbon emission curve. The target scheduling decision generation module is used to perform scheduling optimization analysis of the target grid using the target load curve and the target carbon emission curve as scheduling constraints and a predetermined scheduling fitness evaluation function as the scheduling evaluation method, so as to form a target scheduling decision. The operation scheduling execution module is used to perform operation scheduling execution on the target mobile storage and charging equipment in the target grid according to the target scheduling decision.

Images