Expressway energy collaborative scheduling method, system, equipment and medium

Abstract

The invention discloses an expressway energy collaborative scheduling method, system and device and a medium, mainly relates to the technical field of energy collaborative scheduling, and is used for solving the problems that a traditional prediction model is difficult to capture sudden load fluctuation characteristics and a rigid control strategy causes mismatching of an adjustment instruction and a real-time demand in an existing scheme. Comprising the steps that when sudden load fluctuation occurs in a load prediction curve, compensation adjustment is conducted on the sudden load fluctuation through a self-adaptive control algorithm, and a power grid adjustment instruction set containing voltage / current double parameters of priorities is output; a preset optimal charging and discharging objective function and constraint conditions of each energy storage unit are obtained, and the constraint conditions comprise the real-time meteorological data, the priority of the power grid adjustment instruction and the rated capacity of each energy storage unit; using a preset optimal charging and discharging objective function and constraint conditions to obtain charging and discharging power; and determining the charging and discharging power of each energy storage unit based on a preset distribution weight.

Description

Technical Field

[0001] This application relates to the field of energy collaborative scheduling technology, and in particular to a method, system, equipment and medium for energy collaborative scheduling of highways. Background Technology

[0002] In existing technologies, highway energy dispatch mainly adopts a centralized control mode based on static rules. Traditional solutions typically rely on SCADA systems to collect basic electricity consumption data and achieve load forecasting through linear regression or ARIMA time series models, with the forecasting dimension limited to single time series analysis. Grid regulation often employs PID control strategies with fixed thresholds, while energy storage unit dispatch commonly uses linear programming algorithms, with the peak-valley difference in electricity prices as the core optimization objective. Although these technologies can achieve basic energy management, the data acquisition dimension is limited to electrical parameters (such as voltage and current), lacking multi-dimensional data fusion with traffic flow and meteorological conditions; load forecasting models are mostly single-point forecasts, failing to consider the spatial correlation of the road network; and energy storage dispatch often ignores dynamic constraints such as battery degradation.

[0003] Existing technologies have the following drawbacks: First, traditional forecasting models struggle to capture the characteristics of sudden load fluctuations; second, rigid control strategies lead to a mismatch between regulation commands and real-time demand, especially in scenarios involving multiple energy storage units working together, which can easily cause "over-regulation" or "under-regulation." These problems constrain the economic efficiency and reliability of highway energy systems. Summary of the Invention

[0004] This application provides a method, system, equipment, and medium for coordinated energy dispatching on highways to address the problems of existing solutions where traditional prediction models struggle to capture sudden load fluctuations and rigid control strategies lead to mismatches between regulation commands and real-time demand.

[0005] Firstly, this application provides a method for coordinated energy dispatching on highways, the method comprising: The energy usage of each section of the highway is collected in real time by traffic data collectors. Combined with historical energy consumption database, an LSTM neural network is used to establish a load prediction model that integrates spatiotemporal features and outputs a load prediction curve with a preset time granularity. When sudden load fluctuations occur in the load forecast curve, the sudden load fluctuations are compensated and regulated by the adaptive control algorithm, and the grid regulation command set containing priority voltage / current dual parameters is output. The preset optimal charging and discharging objective function and constraints for each energy storage unit are obtained. The constraints include real-time meteorological data, the priority of grid regulation commands, and the rated capacity of each energy storage unit. The charging and discharging power is obtained using the preset optimal charging and discharging objective function and constraints. The charging and discharging power of each energy storage unit is determined based on the preset allocation weights.

[0006] In one implementation of this application, energy usage data for each section of a highway is collected in real time using a traffic data collector. Combined with a historical energy consumption database, an LSTM neural network is used to establish a load forecasting model that fuses spatiotemporal features, outputting a load forecasting curve with a preset time granularity. Specifically, this includes: Traffic data collectors are deployed along various sections of the highway to collect energy usage data in real time at preset time intervals. After the data is normalized by edge computing nodes, it is spatiotemporally aligned with the historical energy consumption database of the power grid. A two-layer LSTM neural network architecture is adopted. The first layer extracts the time series features of energy use in a single road segment, and the second layer fuses the spatial correlation of vehicle energy use in adjacent road segments through an attention mechanism. The two-layer LSTM neural network architecture outputs a load prediction curve for a future preset time granularity.

[0007] In one implementation of this application, when a sudden load fluctuation occurs in the load forecast curve, an adaptive control algorithm is used to compensate and regulate the sudden load fluctuation, outputting a set of grid regulation instructions containing priority voltage / current dual parameters, specifically including: Set a dynamic threshold algorithm to detect fluctuation amplitude; Predicted loads with fluctuations exceeding a preset threshold are defined as sudden loads. Based on the fuzzy PID adaptive controller, the current grid bus voltage and frequency, the fluctuation of sudden load in the load forecast curve is converted into grid compensation demand; among which, grid compensation demand includes voltage compensation and current compensation.

[0008] In one implementation of this application, based on a fuzzy PID adaptive controller, the current grid bus voltage and frequency, the fluctuation of sudden loads in the load forecast curve is converted into grid compensation demand, specifically including: Collect bus voltage and frequency, and then obtain voltage deviation and voltage offset. It adopts a dual-input voltage deviation and frequency deviation single-output structure; Detect abrupt gradient changes in the predicted curve using a sliding window algorithm; Based on the calculation formulas for the relationship between voltage compensation and abrupt gradient, and the calculation formulas for the relationship between current compensation and abrupt gradient, voltage compensation and current compensation are generated.

[0009] In one implementation of this application, the preset allocation weights include: charging allocation weights and discharging allocation weights; Obtain the preset optimal charge / discharge objective function and constraints for each energy storage unit; use the preset optimal charge / discharge objective function and constraints to obtain the charge / discharge power; determine the charge / discharge power of each energy storage unit based on preset allocation weights, specifically including: The preset optimal charge / discharge objective function and constraints can be obtained through the preset interface; Using the MOEA / D algorithm engine, the output charging and discharging power is calculated based on the preset optimal charging and discharging objective function and constraints. Obtain the charging allocation weight and discharging allocation weight of each energy storage unit, and determine the charging and discharging power of each energy storage unit based on the charging allocation weight, discharging allocation weight, and charging and discharging power; wherein, the sum of the charging allocation weights is 1, and the sum of the discharging allocation weights is 1.

[0010] In one implementation of this application, the preset optimal charging and discharging objective function is: ; in, , , Preset weighting coefficients; Let be the discharge power of the i-th energy storage unit; To predict load demand; Total operating cost; The deviation of the state of charge of an energy storage unit; The constraints include: real-time meteorological data, the priority of power grid regulation commands, and the rated capacity of each energy storage unit; The constraints for real-time meteorological data are as follows: When the real-time wind speed is greater than 8 m / s, the discharge power and charging power decrease by 20%. The priority constraint for power grid regulation commands is as follows: The response time requirement for the preset emergency power grid regulation command is less than the preset first time. The response time requirement for important power grid regulation commands is set to be less than the second preset time. The response time requirement for conventional power grid regulation commands is set to be less than the preset third time. And the preset first time is less than the preset second time is less than the preset third time; The rated capacity constraints for each energy storage unit are as follows: The rate of change of charging and discharging power shall not exceed 10% / s of the rated power.

[0011] Secondly, this application provides a highway energy collaborative dispatch system, the system comprising: The curve module collects real-time energy usage data for each section of the highway using traffic data collectors. Combined with a historical energy consumption database, it uses an LSTM neural network to establish a load forecasting model that fuses spatiotemporal features, outputting a load forecasting curve with a preset time granularity. The command module compensates for sudden load fluctuations in the load forecasting curve using an adaptive control algorithm, outputting a set of grid regulation commands with priority voltage and current parameters. The allocation module obtains the preset optimal charging / discharging objective function and constraints for each energy storage unit. These constraints include real-time meteorological data, the priority of grid regulation commands, and the rated capacity of each energy storage unit. Using the preset optimal charging / discharging objective function and constraints, it obtains the charging / discharging power. Based on preset allocation weights, it determines the charging / discharging power of each energy storage unit.

[0012] In one implementation of this application, the instruction module includes an output unit. Used to set a dynamic threshold algorithm to detect fluctuation amplitude; Predicted loads with fluctuations exceeding a preset threshold are defined as sudden loads. Based on the fuzzy PID adaptive controller, the current grid bus voltage and frequency, the fluctuation of sudden load in the load forecast curve is converted into grid compensation demand; among which, grid compensation demand includes voltage compensation and current compensation.

[0013] Thirdly, this application provides a highway energy collaborative dispatching device, the device comprising: processor; And a memory that stores executable code, which, when executed, causes the processor to execute a highway energy coordinated scheduling method as described above.

[0014] Fourthly, this application provides a non-volatile computer storage medium storing computer instructions, which, when executed, implement a highway energy collaborative scheduling method as described above.

[0015] As can be seen from the above technical solutions, this application has the following advantages: 1. Improved the ability to predict and respond to sudden load fluctuations: A load forecasting model that fuses spatiotemporal features by constructing an LSTM neural network, combined with real-time traffic data and a historical energy consumption database, can effectively capture the nonlinear temporal characteristics of highway energy use. Compared to traditional forecasting methods, the long-term memory characteristic of LSTM can identify the spatiotemporal correlation of sudden load fluctuations, making the load forecast curve with a preset time granularity more closely match actual needs. The introduction of an adaptive control algorithm further solves the command mismatch problem of rigid strategies: when sudden fluctuations occur in the forecast curve, the algorithm dynamically generates a set of voltage / current dual-parameter adjustment commands with priority ranking, and achieves high-precision matching between commands and real-time needs through compensation adjustment, thereby reducing the risk of energy waste or supply shortages caused by forecast deviations.

[0016] 2. Optimized the dynamic collaborative scheduling efficiency of energy storage units: Based on an optimal charge / discharge objective function with multiple constraints (meteorological data, command priority, and rated capacity), the system can quantitatively assess the real-time adjustment potential of each energy storage unit and achieve precise allocation of charge and discharge power by combining preset allocation weights. This dynamic optimization mechanism ensures the power supply stability of high-importance loads through a priority response mechanism and mitigates the impact of environmental factors on energy storage efficiency by utilizing constraints such as meteorological data. The synergistic effect of dual-parameter adjustment commands and multi-objective optimization enables energy storage units to adaptively adjust their charge / discharge strategies within their rated capacity range, avoiding overcharging / over-discharging problems caused by single-parameter control, extending equipment lifespan, and improving overall energy utilization. Attached Figure Description

[0017] To more clearly illustrate the technical solution of the present invention, the accompanying drawings used in the description 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.

[0018] Figure 1 This is a flowchart of a highway energy collaborative scheduling method provided in an embodiment of this application.

[0019] Figure 2 This is a schematic diagram of the internal structure of a highway energy collaborative scheduling system provided in an embodiment of this application.

[0020] Figure 3 This is a schematic diagram of the internal structure of a highway energy collaborative scheduling device provided in an embodiment of this application. Detailed Implementation

[0021] 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 embodiments of the present invention, and not all embodiments. 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.

[0022] Those skilled in the art should understand that the embodiments described below are merely preferred embodiments of this disclosure and do not imply that this disclosure can only be implemented through these preferred embodiments. These preferred embodiments are merely used to explain the technical principles of this disclosure and are not intended to limit the scope of protection of this disclosure. Based on the preferred embodiments provided by this disclosure, all other embodiments obtained by those skilled in the art without creative effort should still fall within the scope of protection of this disclosure.

[0023] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0024] The technical solutions proposed in the embodiments of this application will be described in detail below with reference to the accompanying drawings.

[0025] The embodiment provides a method for coordinated energy scheduling on highways, such as Figure 1 As shown in the embodiments of this application, the method mainly includes the following steps: Step 110: Collect energy usage data for each section of the highway in real time using a traffic data collector, combine it with a historical energy consumption database, and use an LSTM neural network to establish a load prediction model that integrates spatiotemporal features, outputting a load prediction curve with a preset time granularity.

[0026] In some embodiments, energy usage data for each section of a highway is collected in real time using a traffic data collector. Combined with a historical energy consumption database, an LSTM neural network is used to establish a load forecasting model that fuses spatiotemporal features, outputting a load forecasting curve with a preset time granularity. Specifically, this includes: Traffic data collectors are deployed along various sections of the highway to collect energy usage data in real time at preset time intervals. After the data is normalized by edge computing nodes, it is spatiotemporally aligned with the historical energy consumption database of the power grid. A two-layer LSTM neural network architecture is adopted. The first layer extracts the time series features of energy use in a single road segment, and the second layer fuses the spatial correlation of vehicle energy use in adjacent road segments through an attention mechanism. The two-layer LSTM neural network architecture outputs a load prediction curve for a future preset time granularity.

[0027] Based on the above description, this step achieves a substantial improvement in the accuracy of highway energy load forecasting through the collaborative application of traffic data collectors and LSTM neural networks. Specifically, traffic data collectors deployed on various highway sections collect energy usage data in real time at a preset time granularity. Combined with the normalization and spatiotemporal alignment of the data by edge computing nodes, this provides a highly timely and standardized input for load forecasting. In the adopted two-layer LSTM neural network architecture, the first layer captures the dynamic changes in energy consumption of individual road sections through time series feature extraction, solving the problem of insufficient sensitivity of traditional models to short-term fluctuations. The second layer introduces an attention mechanism to integrate the spatial correlation between adjacent road sections, which can identify the cross-regional impact of traffic parameters such as traffic density and speed on energy demand, thereby reflecting the spatial coupling characteristics of road network-level energy consumption in the forecast curve. This spatiotemporal feature fusion mechanism enables the load forecast curve at the preset time granularity to not only include single-point time series trends but also integrate the spatial dependencies brought about by the road network topology, making it particularly suitable for scenarios of sudden energy consumption fluctuations caused by events such as traffic accidents and sudden weather changes. Edge computing's localized data processing reduces cloud transmission latency, ensuring the real-time response capability of the predictive model, while access to historical energy consumption databases optimizes the stability of prediction results through long-term pattern learning. The entire technology chain, from data acquisition and processing to modeling output, forms a closed-loop optimization, providing a reliable decision-making basis for subsequent adaptive control strategies.

[0028] Step 120: When a sudden load fluctuation occurs in the load forecast curve, the sudden load fluctuation is compensated and regulated by an adaptive control algorithm, and a set of grid regulation instructions containing priority voltage / current dual parameters is output.

[0029] Specifically, when sudden load fluctuations occur in the load forecast curve, an adaptive control algorithm is used to compensate for and regulate these fluctuations, outputting a set of grid regulation commands containing priority voltage and current dual parameters, including: Set a dynamic threshold algorithm to detect fluctuation amplitude; Predicted loads with fluctuations exceeding a preset threshold are defined as sudden loads. Based on the fuzzy PID adaptive controller, the current grid bus voltage and frequency, the fluctuation of sudden load in the load forecast curve is converted into grid compensation demand; among which, grid compensation demand includes voltage compensation and current compensation.

[0030] Among them, the adaptive controller based on fuzzy PID, the current bus voltage and frequency of the power grid, converts the fluctuations of sudden loads in the load forecast curve into power grid compensation requirements, specifically including: Collect bus voltage and frequency, and then obtain voltage deviation and voltage offset. It adopts a dual-input voltage deviation and frequency deviation single-output structure; Detect abrupt gradient changes in the predicted curve using a sliding window algorithm; Based on the calculation formulas for the relationship between voltage compensation and abrupt gradient, and the calculation formulas for the relationship between current compensation and abrupt gradient, voltage compensation and current compensation are generated.

[0031] Based on the above description, this step effectively improves the dynamic response capability and regulation accuracy of the power grid to sudden load fluctuations by combining adaptive control algorithms with dynamic threshold detection. The dynamic threshold algorithm identifies sudden loads by setting a preset amplitude threshold, avoiding misjudgment of small fluctuations by traditional fixed thresholds, while ensuring timely capture of critical fluctuations. The fuzzy PID-based adaptive controller collects bus voltage and frequency data in real time, constructs a dual-input (voltage deviation, frequency deviation) single-output structure, and dynamically adjusts PID parameters using fuzzy logic, enabling the control strategy to adapt to compensation requirements of different fluctuation intensities. The introduction of the sliding window algorithm enables quantitative analysis of the gradient of sudden changes in the load prediction curve. Combining the calculation formulas for voltage / current compensation and the gradient of sudden changes, the fluctuation is transformed into precise voltage and current compensation, forming a dual-parameter regulation instruction set with priority. This compensation mechanism, through the coordinated regulation of voltage and current dual parameters, solves the over-compensation or under-compensation problems that may occur with single-parameter regulation, and ensures the power supply stability of critical loads through priority ranking. Real-time feedback of bus voltage and frequency allows the system to dynamically correct the compensation amount according to the actual state of the power grid, avoiding secondary fluctuations caused by prediction errors. The overall solution achieves closed-loop optimization of the entire process from sudden load identification to compensation command generation through the coordination of three levels: dynamic detection, fuzzy adaptive control, and dual-parameter output. It provides a solution for power grid dispatch that balances response speed and regulation accuracy.

[0032] Step 130: Obtain the preset optimal charging and discharging objective function and constraints for each energy storage unit; use the preset optimal charging and discharging objective function and constraints to obtain the charging and discharging power; determine the charging and discharging power of each energy storage unit based on the preset allocation weights.

[0033] The constraints include real-time meteorological data, the priority of power grid regulation commands, and the rated capacity of each energy storage unit.

[0034] In some embodiments, the preset allocation weights include: charging allocation weights and discharging allocation weights; Obtain the preset optimal charge / discharge objective function and constraints for each energy storage unit; use the preset optimal charge / discharge objective function and constraints to obtain the charge / discharge power; determine the charge / discharge power of each energy storage unit based on preset allocation weights, specifically including: The preset optimal charge / discharge objective function and constraints can be obtained through the preset interface; Using the MOEA / D algorithm engine, the output charging and discharging power is calculated based on the preset optimal charging and discharging objective function and constraints. Obtain the charging allocation weight and discharging allocation weight of each energy storage unit, and determine the charging and discharging power of each energy storage unit based on the charging allocation weight, discharging allocation weight, and charging and discharging power; wherein, the sum of the charging allocation weights is 1, and the sum of the discharging allocation weights is 1.

[0035] The preset optimal charge / discharge objective function is: ; in, , , Preset weighting coefficients; Let be the discharge power of the i-th energy storage unit; To predict load demand; Total operating cost; The deviation of the state of charge of an energy storage unit; The constraints include: real-time meteorological data, the priority of power grid regulation commands, and the rated capacity of each energy storage unit; The constraints for real-time meteorological data are as follows: When the real-time wind speed is greater than 8 m / s, the discharge power and charging power decrease by 20%. The priority constraint for power grid regulation commands is as follows: The response time requirement for the preset emergency power grid regulation command is less than the preset first time. The response time requirement for important power grid regulation commands is set to be less than the second preset time. The response time requirement for conventional power grid regulation commands is set to be less than the preset third time. And the preset first time is less than the preset second time is less than the preset third time; The rated capacity constraints for each energy storage unit are as follows: The rate of change of charging and discharging power shall not exceed 10% / s of the rated power.

[0036] Based on the above description, this step achieves the scientific planning and dynamic adjustment of the energy storage unit's charging and discharging strategy through the synergistic application of multi-objective optimization algorithms and weight allocation mechanisms. When using the MOEA / D algorithm engine to solve for the optimal charging and discharging objective function, multi-dimensional constraints are integrated, including real-time meteorological data (such as the impact of temperature and sunlight on energy storage efficiency), grid regulation command priorities (ensuring the continuity of power supply to critical loads), and the rated capacity of each energy storage unit (preventing overcharging and over-discharging). This ensures that the solution meets both grid regulation requirements and equipment safety operation requirements. The introduction of charging / discharging allocation weights (each summing to 1) enables on-demand power allocation: higher-priority energy storage units receive greater regulation margins, while meteorological data constraints prevent excessive use of inefficient units under extreme conditions. The configuration function of the preset interface allows operators to flexibly adjust the objective function parameters (such as economic efficiency, response speed, and other optimization objectives) according to actual scenarios, while the multi-objective decomposition characteristics of MOEA / D ensure the balanced achievement of different optimization objectives. This allocation mechanism, while ensuring that the total charging and discharging power meets the target, dynamically adjusts the output ratio of each unit through weighting coefficients. This avoids local overload caused by traditional equal-distribution strategies and prioritizes the activation of energy storage units with fast response times based on the real-time grid status (such as frequency deviation). The overall scheme forms a complete decision-making chain from global optimization to individual execution through three progressive processing levels: constraint modeling, multi-objective optimization solution, and weight allocation. This provides a quantifiable and configurable solution for collaborative scheduling of energy storage under complex operating conditions.

[0037] In addition, this application Figure 2 This application provides an embodiment of a highway energy collaborative scheduling system. For example... Figure 2 As shown in the embodiments of this application, the system mainly includes: The curve module 210 is used to collect energy usage data of each section of the highway in real time through a traffic collector, combine it with a historical energy consumption database, and use an LSTM neural network to establish a load prediction model that integrates spatiotemporal features, and output a load prediction curve with a preset time granularity.

[0038] Curve module 210 includes curve units, Used to deploy traffic data collectors on various sections of highways to collect energy usage data in real time at preset time granularity. After the data is normalized by edge computing nodes, it is spatiotemporally aligned with the historical energy consumption database of the power grid. A two-layer LSTM neural network architecture is adopted. The first layer extracts the time series features of energy use in a single road segment, and the second layer fuses the spatial correlation of vehicle energy use in adjacent road segments through an attention mechanism. The two-layer LSTM neural network architecture outputs a load prediction curve for a future preset time granularity.

[0039] The instruction module 220 is used to compensate and regulate sudden load fluctuations through an adaptive control algorithm when such fluctuations occur in the load forecast curve, and outputs a set of grid regulation instructions containing priority voltage / current dual parameters.

[0040] Instruction module 220 includes an output unit, Used to set a dynamic threshold algorithm to detect fluctuation amplitude; Predicted loads with fluctuations exceeding a preset threshold are defined as sudden loads. Based on the fuzzy PID adaptive controller, the current grid bus voltage and frequency, the fluctuation of sudden load in the load forecast curve is converted into grid compensation demand; among which, grid compensation demand includes voltage compensation and current compensation.

[0041] The allocation module 230 is used to obtain the preset optimal charging and discharging objective function and constraints of each energy storage unit, wherein the constraints include real-time meteorological data, the priority of grid regulation commands, and the rated capacity of each energy storage unit; using the preset optimal charging and discharging objective function and constraints, the charging and discharging power is obtained; and based on the preset allocation weight, the charging and discharging power of each energy storage unit is determined.

[0042] The above are method embodiments of this application. Based on the same inventive concept, this application also provides a highway energy collaborative scheduling device. Figure 3 As shown, the device includes: a processor; and a memory storing executable code thereon, which, when executed, causes the processor to perform a highway energy collaborative scheduling method as described in the above embodiments.

[0043] Specifically, the server collects real-time energy usage data for each section of the highway using traffic data collectors. Combined with a historical energy consumption database, it uses an LSTM neural network to establish a load forecasting model that fuses spatiotemporal features, outputting a load forecasting curve with a preset time granularity. When sudden load fluctuations occur in the load forecasting curve, an adaptive control algorithm compensates for these fluctuations, outputting a set of grid regulation commands with priority voltage and current parameters. The server also obtains the preset optimal charging and discharging objective function and constraints for each energy storage unit. These constraints include real-time meteorological data, the priority of grid regulation commands, and the rated capacity of each energy storage unit. Using the preset optimal charging and discharging objective function and constraints, the charging and discharging power is obtained. Finally, based on preset weight allocation, the charging and discharging power of each energy storage unit is determined.

[0044] In addition, this application embodiment also provides a non-volatile computer storage medium storing executable instructions, which, when executed, implement the highway energy collaborative scheduling method described above.

[0045] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for coordinated energy dispatching on highways, characterized in that, The method includes: The energy usage of each section of the highway is collected in real time by traffic data collectors. Combined with historical energy consumption database, an LSTM neural network is used to establish a load prediction model that integrates spatiotemporal features and outputs a load prediction curve with a preset time granularity. When sudden load fluctuations occur in the load forecast curve, the sudden load fluctuations are compensated and regulated by the adaptive control algorithm, and the grid regulation command set containing priority voltage / current dual parameters is output. The preset optimal charging and discharging objective function and constraints for each energy storage unit are obtained. The constraints include real-time meteorological data, the priority of grid regulation commands, and the rated capacity of each energy storage unit. The charging and discharging power is obtained using the preset optimal charging and discharging objective function and constraints. The charging and discharging power of each energy storage unit is determined based on the preset allocation weights.

2. The highway energy collaborative scheduling method according to claim 1, characterized in that, Real-time energy usage data for various sections of the highway is collected by traffic data collectors. Combined with a historical energy consumption database, an LSTM neural network is used to establish a load forecasting model that fuses spatiotemporal features. The model outputs load forecast curves with a preset time granularity, specifically including: Traffic data collectors are deployed along various sections of the highway to collect energy usage data in real time at preset time intervals. After the data is normalized by edge computing nodes, it is spatiotemporally aligned with the historical energy consumption database of the power grid. A two-layer LSTM neural network architecture is adopted. The first layer extracts the time series features of energy use in a single road segment, and the second layer fuses the spatial correlation of vehicle energy use in adjacent road segments through an attention mechanism. The two-layer LSTM neural network architecture outputs a load prediction curve for a future preset time granularity.

3. The highway energy collaborative scheduling method according to claim 1, characterized in that, When sudden load fluctuations occur in the load forecast curve, an adaptive control algorithm is used to compensate for and regulate these fluctuations, outputting a set of grid regulation commands containing priority voltage and current dual parameters, specifically including: Set a dynamic threshold algorithm to detect fluctuation amplitude; Predicted loads with fluctuations exceeding a preset threshold are defined as sudden loads. Based on the fuzzy PID adaptive controller, the current grid bus voltage and frequency, the fluctuation of sudden load in the load forecast curve is converted into grid compensation demand; among which, grid compensation demand includes voltage compensation and current compensation.

4. The highway energy collaborative scheduling method according to claim 1, characterized in that, Based on a fuzzy PID adaptive controller, and using the current grid bus voltage and frequency, the fluctuations in sudden loads in the load forecast curve are converted into grid compensation requirements, specifically including: Collect bus voltage and frequency, and then obtain voltage deviation and voltage offset. It adopts a dual-input voltage deviation and frequency deviation single-output structure; Detect abrupt gradient changes in the predicted curve using a sliding window algorithm; Based on the calculation formulas for the relationship between voltage compensation and abrupt gradient, and the calculation formulas for the relationship between current compensation and abrupt gradient, voltage compensation and current compensation are generated.

5. The highway energy collaborative scheduling method according to claim 1, characterized in that, The preset allocation weights include: charging allocation weights and discharging allocation weights; Obtain the preset optimal charge / discharge objective function and constraints for each energy storage unit; use the preset optimal charge / discharge objective function and constraints to obtain the charge / discharge power; determine the charge / discharge power of each energy storage unit based on preset allocation weights, specifically including: The preset optimal charge / discharge objective function and constraints can be obtained through the preset interface; Using the MOEA / D algorithm engine, the output charging and discharging power is calculated based on the preset optimal charging and discharging objective function and constraints. Obtain the charging allocation weight and discharging allocation weight of each energy storage unit, and determine the charging and discharging power of each energy storage unit based on the charging allocation weight, discharging allocation weight, and charging and discharging power; wherein, the sum of the charging allocation weights is 1, and the sum of the discharging allocation weights is 1.

6. The highway energy collaborative scheduling method according to claim 1, characterized in that, The preset optimal charge / discharge objective function is: ； in, , , Preset weighting coefficients; Let be the discharge power of the i-th energy storage unit; To predict load demand; Total operating cost; This represents the deviation in the state of charge of an individual energy storage unit. The constraints include: real-time meteorological data, the priority of power grid regulation commands, and the rated capacity of each energy storage unit; The constraints for real-time meteorological data are as follows: When the real-time wind speed is greater than 8 m / s, the discharge power and charging power decrease by 20%. The priority constraint for power grid regulation commands is as follows: The response time requirement for the preset emergency power grid regulation command is less than the preset first time. The response time requirement for important power grid regulation commands is set to be less than the second preset time. The response time requirement for conventional power grid regulation commands is set to be less than the preset third time. And the preset first time is less than the preset second time is less than the preset third time; The rated capacity constraints for each energy storage unit are as follows: The rate of change of charging and discharging power shall not exceed 10% / s of the rated power.

7. A highway energy collaborative dispatch system, characterized in that, The system includes: The curve module is used to collect energy usage data for each section of the highway in real time through traffic data collectors. Combined with historical energy consumption databases, it uses an LSTM neural network to establish a load prediction model that fuses spatiotemporal features and outputs a load prediction curve with a preset time granularity. The instruction module is used to compensate and regulate sudden load fluctuations in the load forecast curve through an adaptive control algorithm, and outputs a set of grid regulation instructions containing priority voltage / current dual parameters. The allocation module is used to obtain the preset optimal charging and discharging objective function and constraints for each energy storage unit. The constraints include real-time meteorological data, the priority of grid regulation commands, and the rated capacity of each energy storage unit. The module obtains the charging and discharging power using the preset optimal charging and discharging objective function and constraints. Based on the preset allocation weights, the module determines the charging and discharging power of each energy storage unit.

8. The highway energy collaborative dispatch system according to claim 7, characterized in that, The instruction module includes an output unit. Used to set a dynamic threshold algorithm to detect fluctuation amplitude; Predicted loads with fluctuations exceeding a preset threshold are defined as sudden loads. Based on the fuzzy PID adaptive controller, the current grid bus voltage and frequency, the fluctuation of sudden load in the load forecast curve is converted into grid compensation demand; among which, grid compensation demand includes voltage compensation and current compensation.

9. A highway energy collaborative dispatching device, characterized in that, The device includes: processor; And a memory having executable code stored thereon, which, when executed, causes the processor to perform a highway energy collaborative scheduling method as described in any one of claims 1-6.

10. A non-volatile computer storage medium, characterized in that, It stores computer instructions, which, when executed, implement a highway energy collaborative scheduling method as described in any one of claims 1-6.

Images