Smart energy scheduling-based energy exchange fusion operation management platform

Abstract

The invention, which belongs to the technical field of the smart energy and intelligent traffic system, discloses a cross-energy fusion operation management platform based on smart energy scheduling, comprising: a monitoring subsystem for collecting energy consumption data, traffic flow data and meteorological environment data in real time; the analysis subsystem is used for carrying out load prediction, generating capacity prediction, carbon emission accounting and traffic-energy consumption correlation analysis based on the multi-source heterogeneous data acquired by the monitoring subsystem; the regulation and control subsystem is used for dynamically generating an energy scheduling strategy based on an intelligent scheduling algorithm and realizing source-network-load-storage dynamic balance; wherein the monitoring subsystem, the analysis subsystem and the regulation and control subsystem carry out service decoupling and data interaction through a RESTful API (Application Program Interface). Through innovative data fusion, multi-task conjoint analysis and intelligent collaborative scheduling, accurate matching and optimization of energy supply and demand are realized, and the system toughness, the economic benefit and the green level are improved.

Description

Technical Field

[0001] This invention relates to the field of smart energy and intelligent transportation systems, and more specifically to a transportation and energy integration operation management platform based on smart energy dispatch. Background Technology

[0002] Currently, with the advancement of the "dual carbon" target and the deepening of the construction of new power systems, building a smart energy system that coordinates and interacts with the energy source, grid, load, and storage has become crucial. Meanwhile, urban transportation systems, as important energy consumers and load influencing factors, have a close and complex dynamic relationship with the energy system in terms of their operational status.

[0003] However, current energy management systems and traffic management systems are mostly built and operate independently, resulting in the following major drawbacks: First, energy consumption data (such as service area electricity consumption and photovoltaic power generation) and traffic operation data (such as mainline traffic flow and the number of vehicles staying in service areas) belong to different systems, with heterogeneous formats and inconsistent spatiotemporal benchmarks, forming "information silos" that are difficult to correlate and analyze. Second, existing management is mostly based on historical experience and simple rules, lacking the ability to accurately predict load and power generation using artificial intelligence models, and failing to quantitatively analyze the dynamic impact of traffic flow changes on energy demand, leading to insufficient forward-looking decision-making. Third, energy dispatch is often limited to internal microgrids, failing to link with traffic conditions (such as congestion and high traffic volume during holidays), and unable to mitigate load fluctuations through proactive traffic guidance (such as charging guidance) to maximize the global optimization benefits of "source-grid-load-storage". Although some energy management systems or traffic monitoring platforms have emerged in existing technologies, few operation management platforms deeply couple the two and achieve two-way closed-loop collaboration through data-driven approaches.

[0004] Therefore, how to provide a smart energy dispatch-based integrated energy operation and management platform is a problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0005] In view of this, the present invention provides an energy-integrated operation and management platform based on intelligent energy dispatch. For specific scenarios of highways, it achieves precise matching and optimization of energy supply and demand through innovative data fusion, multi-task joint analysis and intelligent collaborative dispatch, thereby improving system resilience, economic efficiency and green level.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: A smart energy dispatch-based integrated energy operation and management platform includes: The monitoring subsystem is used to collect energy consumption data, traffic flow data, and meteorological environmental data in real time. The analysis subsystem is used for load forecasting, power generation forecasting, carbon emission accounting, and traffic-energy consumption correlation analysis based on multi-source heterogeneous data collected by the monitoring subsystem. The control subsystem is used to dynamically generate energy dispatch strategies based on intelligent dispatch algorithms to achieve dynamic balance between energy sources, grid, load, and storage. The monitoring subsystem, the analysis subsystem, and the control subsystem are decoupled and interact with each other through a RESTful API.

[0007] Furthermore, the monitoring subsystem includes a multimodal data acquisition unit, a data cleaning and preprocessing unit, and a spatiotemporal alignment unit; The multimodal data acquisition unit is used to collect energy consumption data, traffic flow data and meteorological environment data, and transmit them back through a dual-mode asynchronous transmission channel composed of 5G communication and LoRa communication. The data cleaning and preprocessing unit is used to remove outliers from the collected data based on the improved K-means clustering algorithm and to smooth the traffic flow data using a sliding window mean filtering algorithm. The spatiotemporal alignment unit is used to align the spatiotemporal dimensions of the cleaned and preprocessed multi-source data through timestamp synchronization protocols and geographic information system coordinate mapping.

[0008] Furthermore, the energy consumption data includes energy consumption data for highway service areas, toll stations, tunnel lighting, and charging piles; and the collected traffic flow data includes traffic flow and vehicle speed data for the highway mainline and ramps.

[0009] Furthermore, the analysis subsystem includes a multi-task joint modeling unit, a traffic-energy consumption correlation mining unit, and a distributed computing unit; The multi-task joint modeling unit adopts a spatiotemporal attention mechanism model based on the Transformer architecture to simultaneously process load forecasting, power generation forecasting, and carbon emission accounting tasks. The traffic-energy consumption correlation mining unit uses the Granger causality test algorithm to quantify the dynamic correlation between traffic flow and energy consumption, and constructs a traffic flow-energy consumption elasticity coefficient matrix. The distributed computing unit, built on the Apache Flink stream processing framework, is used for real-time parallel computing of the multi-task joint modeling unit and the traffic-energy consumption correlation mining unit.

[0010] Furthermore, the traffic-energy consumption correlation mining unit also includes dynamic correlation analysis of traffic flow and energy consumption of service areas and toll stations, as well as analysis of electric vehicle charging behavior.

[0011] Furthermore, the multi-task joint modeling unit includes: An LSTM-XGBoost hybrid model is used to combine time series features with external meteorological factors for prediction; A LSTM network modified by the Clear Sky model is used for photovoltaic power generation prediction, and a CNN-LSTM model based on numerical weather prediction data is used for wind power prediction. The system performs multi-dimensional carbon emission accounting by industry, region, and energy type using a pre-set IPCC default emission factor library.

[0012] Furthermore, the control subsystem includes an intelligent scheduling algorithm engine, a strategy verification and simulation unit, and a real-time control interface unit; The intelligent scheduling algorithm engine is used to execute the following optimization process to generate a dynamic energy scheduling strategy; The strategy verification and simulation unit constructs a virtual energy system model based on digital twin technology, which is used to verify the robustness of the dynamic energy dispatch strategy through Monte Carlo simulation. The real-time control interface unit, using the IEC 61850 protocol, is used for bidirectional data interaction and control command issuance with external energy management systems, highway control systems, and energy storage devices.

[0013] Furthermore, the intelligent scheduling algorithm engine includes: a) Construct a traffic-electricity coupled stochastic user equilibrium model to predict the path flow of electric vehicles in the road network and the potential vehicle arrival rate at service area charging stations; b) Construct a service quality assessment module based on queuing theory to calculate the average waiting time and user churn probability at charging stations; c) An improved genetic algorithm is used to minimize the total operating cost of the system, maximize the absorption of renewable energy, and minimize carbon emissions. The vehicle arrival rate and service quality are incorporated as constraints or optimization factors to perform multi-objective optimization.

[0014] Furthermore, the intelligent scheduling algorithm engine also includes: A master-slave iterative optimization framework is adopted, in which the distribution side generates planning candidate schemes that include priority photovoltaic consumption, and the transmission side conducts adaptive evaluation and cost accounting of the schemes. Through iteration, a collaborative scheduling scheme that ensures service quality and economic benefits is sought. Set grid security constraints, energy storage device charge and discharge rate constraints, and traffic signal timing constraints; A dynamic weight adjustment mechanism is adopted to dynamically adjust the weights of each optimization objective in the objective function based on real-time electricity prices, fluctuations in renewable energy output, and traffic congestion index.

[0015] Furthermore, it also includes a storage subsystem, which adopts a hybrid architecture of time-series database and relational database to store real-time monitoring data and analysis results data respectively, and to perform historical data backtracking analysis.

[0016] As can be seen from the above technical solution, compared with the prior art, this invention discloses a traffic-energy integration operation and management platform based on intelligent energy dispatch, effectively solving the problem of multi-source heterogeneous data fusion, providing a reliable and consistent data foundation for upper-level analysis, and is particularly suitable for the linear distribution and diverse scenarios of highways; it adopts a Transformer-based multi-task joint learning model to share cross-domain feature information and simultaneously improve the prediction efficiency and accuracy of load, power generation, and carbon accounting. The introduction of quantitative analysis methods such as Granger causality tests makes the impact of traffic on energy consumption measurable and interpretable, providing a scientific basis for refined management; the dispatch model uses traffic status as a key input and constraint, enabling energy dispatch strategies to proactively respond to traffic changes. Simultaneously, the platform can output proactive guidance suggestions for the traffic system, forming a closed-loop optimization of "energy-driven traffic management and traffic-driven energy," truly realizing the synergy of the two networks. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0018] Figure 1 This is a schematic diagram of the platform structure provided by the present invention. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some 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.

[0020] Example 1: See Figure 1 This invention discloses an energy exchange and energy integration operation management platform based on smart energy dispatch, characterized in that it includes a monitoring subsystem, an analysis subsystem, a control subsystem, and a storage subsystem, and each subsystem decouples services and interacts with data through a RESTful API.

[0021] The monitoring subsystem is used to collect and visualize energy consumption data, traffic flow data, and meteorological environmental data of transportation infrastructure in real time.

[0022] Based on the data collected and aggregated by the monitoring subsystem, the analysis subsystem performs load forecasting and power generation forecasting based on multi-source heterogeneous data, including historical traffic flow data, and simultaneously completes carbon emission accounting and traffic-energy consumption correlation analysis.

[0023] The control subsystem is used to dynamically generate energy dispatch strategies based on intelligent dispatch algorithms to achieve dynamic balance between energy sources, grid, load, and storage. Its core components are: first, predicting electric vehicle flow and charging demand through a traffic-electricity coupled stochastic user equilibrium model; second, quantifying the charging service experience using a queuing theory-based service quality assessment module; and finally, generating minute-level updated collaborative dispatch strategies with the goals of economy, low carbon emissions, and high-proportion green energy consumption, while ensuring charging service quality. These strategies are then pre-verified through digital twins and Monte Carlo simulations.

[0024] In one specific embodiment, the monitoring subsystem is the platform's data perception layer, responsible for the real-time and reliable acquisition and preprocessing of multi-source heterogeneous data. Specifically, it includes: Multimodal data acquisition unit: Connects to various sensing devices along the highway via an IoT interface. The energy consumption data it collects includes: power consumption data from service area buildings, charging piles, toll station electromechanical equipment, and tunnel lighting and ventilation systems; power generation and charging / discharging data from photovoltaic inverters and energy storage converters. The traffic flow data it collects includes: traffic volume, vehicle speed, and vehicle type data collected from highway mainline and ramp checkpoints, and microwave vehicle detectors; and service area parking space monitoring data. Data is transmitted back via a dual-mode asynchronous transmission channel composed of 5G and LoRa, balancing the real-time requirements of critical data with the low power consumption of massive nodes.

[0025] Data cleaning and preprocessing unit: To address noise and anomalies in the original data, an improved K-means clustering algorithm is used, introducing a local density factor to identify discrete points and remove outliers; for traffic flow data with high-frequency fluctuations, a sliding window mean filtering algorithm is used for smoothing to extract effective trend information.

[0026] Spatiotemporal alignment unit: To ensure the fusion of multi-source data, this unit uses the high-precision Network Time Protocol (NTP) to stamp all data with a unified timestamp, and based on GIS coordinates, maps each data point to specific road segments, service areas, or equipment coordinates to achieve precise alignment of data in the spatiotemporal dimension, such as associating the energy consumption of a service area with the vehicle dwell data within it during a certain time period.

[0027] In one specific embodiment, the analysis subsystem is the core cognitive hub of the platform, responsible for extracting patterns and predicting the future from fused data. Specifically, it includes: Distributed computing unit: Built on the Apache Flink stream processing framework, it provides high-throughput, low-latency real-time parallel computing capabilities for upper-layer analysis tasks, supporting the real-time processing of tens of thousands of data records per second.

[0028] Multi-task joint modeling unit: Employs a spatiotemporal attention mechanism model based on the Transformer architecture. This model takes spatiotemporally aligned multidimensional feature vectors, such as historical load, real-time traffic flow, and meteorological factors, as input. Through a self-attention mechanism, it captures long-range dependencies between different features and simultaneously outputs load forecasting results, photovoltaic / wind power generation forecasting results, and carbon emission estimates based on the current energy structure. As a preferred implementation, this unit can integrate multiple advanced models: for example, a hybrid LSTM-XGBoost model can be used for load forecasting, combining sequential features with external factors; photovoltaic power generation forecasting uses an LSTM network modified by the Clear Sky physics model to improve forecasting accuracy under clear sky conditions; and wind power forecasting uses a CNN-LSTM model driven by numerical weather prediction data.

[0029] Traffic-Energy Correlation Mining Unit: To quantitatively reveal the impact of traffic behavior on energy consumption, this unit employs the Granger causality test algorithm. Based on time series analysis, this algorithm determines whether changes in specific traffic variables, such as service area entrance traffic flow, statistically lead changes in energy variables, such as changes in total service area load, thereby confirming the causal relationship. Through continuous calculation, the platform can construct a dynamic traffic flow-energy consumption elasticity coefficient matrix, quantifying the sensitivity of traffic conditions at different road segments and time periods to regional energy consumption. This unit's analysis particularly focuses on the dynamic correlation between traffic flow and service area / toll station energy consumption, as well as mining behavioral characteristics such as electric vehicle charging time preferences and single-charge volume distribution.

[0030] In one specific embodiment, the control subsystem serves as the platform's decision-making and execution hub, responsible for generating and validating optimization strategies. Specifically, it includes: Intelligent Scheduling Algorithm Engine: This engine executes a phased optimization process. The first phase constructs a traffic-electricity coupled stochastic user equilibrium model, classifying users in the traffic network into electric vehicles and gasoline vehicles. By solving for the equilibrium state, it predicts the flow of electric vehicles heading to each service area, thereby calculating the potential vehicle arrival rate at charging stations. The second phase constructs a service quality assessment module based on queuing theory. Based on vehicle arrival rate, the number of charging piles, and service rate, it simulates and calculates the average waiting time for users at charging stations and the probability of leaving due to long queues, quantifying service quality into optimizable indicators. The third phase employs an improved genetic algorithm for multi-objective optimization. This algorithm primarily aims to minimize the total system operating cost, maximize the local consumption ratio of renewable energy (such as photovoltaics), and minimize total carbon emissions, while using the arrival rate predicted in the first step and the service quality requirements calculated in the second step as key constraints. To further improve grid coordination efficiency, the algorithm adopts a master-slave iterative optimization framework: the distribution side (service area microgrid) generates multiple priority scheduling candidate schemes for absorbing photovoltaic power, and the transmission side (upper-level grid) evaluates the feasibility and global cost of these schemes. Through iterative feedback, it seeks the optimal coordinated scheduling scheme for the entire system. The engine has a built-in dynamic weight adjustment mechanism that can dynamically adjust the weights of each objective based on real-time electricity prices, photovoltaic power output fluctuations, and traffic congestion indices, generating dynamic energy scheduling strategies updated every minute.

[0031] Strategy Verification and Simulation Unit: To ensure the security and robustness of the scheduling strategy, this unit constructs a high-fidelity virtual highway energy system model based on digital twin technology. Before the strategy is issued, various possible random disturbances are injected into the digital twin environment using the Monte Carlo simulation method. In this embodiment, these include load forecasting deviations and sudden weather changes. The strategy is simulated thousands of times, and it is only released when the strategy can maintain system stability and achieve the optimization objectives in the vast majority of scenarios.

[0032] Real-time control interface unit: Serving as the interface between the platform and the physical world, it uses IEC 61850 and vehicle-to-everything (V2X) or dedicated data interfaces for bidirectional data interaction with external energy management systems, highway traffic guidance systems, and energy storage devices. It also sends coordinated guidance suggestions to the highway traffic control system (TSC) via a dedicated data interface; in this embodiment, this refers to guidance during charging discount periods. The platform is positioned as an operation management platform, primarily issuing strategy suggestions or setpoint commands to lower-level control systems rather than directly executing device-level control, thereby ensuring system flexibility and safety.

[0033] In one specific embodiment, the storage subsystem adopts a hybrid architecture of time-series database and relational database.

[0034] Time-series databases are used for efficient storage and rapid retrieval of massive amounts of real-time monitoring sequence data; in this embodiment, the data is power per second and traffic flow per minute. Relational databases are used to store structured data such as model parameters, analysis reports, and strategy logs, and support complex historical data backtracking analysis and multidimensional report generation.

[0035] Example 2: This embodiment is based on the application of the integration of transportation and energy in a highway reconstruction and expansion demonstration project, specifically in a relevant section of the Shen-Su-Zhe-Wan Expressway.

[0036] The Shensu-Zhewan Expressway, from Changxing West Interchange to the Zhejiang-Anhui provincial border, is approximately 28.761 kilometers long. The section from the starting point to Yaojiaqiao (approximately 1.211 kilometers) is currently a six-lane expressway with a design speed of 120 km / h. The section from Yaojiaqiao to the end point (approximately 27.55 kilometers) was originally a four-lane expressway with a design speed of 100 km / h, but the operating speed for passenger cars has now been increased to 120 km / h. Based on the project's construction needs, and considering the actual conditions, implementation conditions, and future development requirements of the Shensu-Zhewan Expressway reconstruction and expansion project, and following the integrated transportation and energy solution approach, this project will be developed into a pilot demonstration project for integrated transportation and energy solutions in Zhejiang's expressway reconstruction and expansion.

[0037] Zhejiang possesses excellent conditions for the integrated development of transportation and energy, and can fully leverage its late-mover advantage to achieve leapfrog development. As an open economic province, Zhejiang enjoys unique advantages in terms of policy and market. First, as one of the first 13 pilot areas for the construction of a strong transportation nation, it has fertile ground for cultivating pilot demonstration projects. Second, the approval process for distributed photovoltaic (PV) registration and grid connection remains relatively lenient, which is conducive to the development and construction of roadside PV. Third, the ownership and market share of new energy vehicles far exceed the national average, and charging demand will continue to grow rapidly. Fourth, with its developed industry and commerce, and numerous ports, the demand for short-distance road transportation is huge, making it suitable for carrying out electric heavy-duty truck demonstrations. Zhejiang Provincial Transportation Investment Group is well-positioned to pioneer and explore high-quality development of integrated transportation and energy.

[0038] By leveraging the Shensu-Zhejiang-Anhui project, the transportation and energy integration operation and management platform will be promoted and implemented, creating a demonstration project for the integration of highway reconstruction and expansion with new energy, and providing replicable and scalable experience for the integration of transportation and energy in highways in Zhejiang Province and even the whole country.

[0039] The platform as a whole adopts a cloud-edge collaborative microservice architecture. Lightweight data collection and preprocessing modules of the monitoring subsystem are deployed at the edge, close to data sources such as service areas and toll stations. All services of the analysis, control, and storage subsystems are deployed in the cloud, utilizing cloud computing's elastic resources for deep computing and global optimization. All services are registered, discovered, and invoked through an API gateway.

[0040] Data acquisition phase: Photovoltaic strings and smart meters are deployed on the rooftops of the service area of ​​the demonstration project, and AC charging piles are installed at charging stations. Data from these devices is aggregated to the local IoT gateway, i.e., the edge server, via RS-485 or Ethernet. Traffic flow data is transmitted to the regional edge server via fiber optic cables from checkpoint cameras and microwave vehicle detectors deployed on the main line and ramps. The multimodal data acquisition unit has a built-in protocol adapter that encapsulates various data types into a unified JSON format.

[0041] Edge Preprocessing Stage: Data cleaning and preprocessing units running on edge servers perform batch processing every 5 minutes. For example, for charging pile power data, an improved K-means algorithm is used to mark data points whose instantaneous power values ​​deviate from historical normal cluster centers by more than 3 standard deviations as outliers and remove them. For mainline vehicle speed data, a 15-minute sliding window is used for mean filtering to eliminate random fluctuations.

[0042] Specifically, the data source is the equipment's historical data itself, based on the statistical 3σ criterion, and its selection is based on the optimal balance point in engineering practice. During project implementation, the platform automatically completes this calculation and threshold setting process based on the actual historical data of the specific equipment.

[0043] Spatiotemporal alignment phase: All edge nodes synchronize with the central NTP server. Based on the precise latitude and longitude preset in the equipment ledger, the spatiotemporal alignment unit aggregates data such as the total electricity consumption, photovoltaic power generation, and number of vehicles entering Service Area A during the time period of 09:00-09:05 into the spatial entity "Service Area A" and the time window of "09:00-09:05", completing the integration preparation.

[0044] Real-time computing and storage: The distributed computing unit Flink job continuously consumes the aligned data stream from the monitoring subsystem, writing part of it to the time-series database for real-time querying, and sending the other part to the analysis model.

[0045] Joint Prediction: The Transformer model in the multi-task joint modeling unit has been trained using historical data. The model is triggered every 15 minutes, inputting feature sequences such as service area load, traffic flow, temperature, and irradiance from the past 24 hours, and outputs in parallel load forecasts, photovoltaic power generation forecasts, and corresponding carbon emission intensity trends for the next 6 hours. Preferably, for photovoltaic forecasting, the ClearSky model is first used to calculate the theoretical maximum power generation, and then an LSTM is used to learn the deviation relationship between actual power generation and the theoretical value, such as the influence of clouds and dust, thereby improving prediction accuracy.

[0046] Association Mining: A separate Flink job runs the logic of the traffic-energy consumption association mining unit. For example, it continuously analyzes the time series of "entry traffic" and "total load" for each service area, using Granger causality tests with a lag order of 4 and a significance level of α=0.05. If the test passes, the elasticity coefficient is calculated. Based on this, the platform can generate insights such as: During the National Day holiday, the elasticity coefficient of energy consumption of service area B to its entry traffic is 0.8, and traffic changes lead energy consumption changes by approximately 30 minutes.

[0047] Strategy Generation: At fixed daily times and trigger events, when a charging peak is predicted, the intelligent scheduling algorithm engine is activated. It acquires the latest predictions and resilience coefficients from the analysis subsystem. The initial optimization target weights are set as follows: cost 0.5, green energy consumption 0.3, and low carbon 0.2. The improved genetic algorithm's chromosome encoding includes energy storage charging and discharging plans every 15 minutes for the next 24 hours, interruptible load switching plans, and "charging guidance periods" issued to the traffic system. The algorithm searches the solution space while satisfying all constraints, specifically including reducing the adjustable power limit of charging piles at midday if afternoon congestion is predicted in service area C.

[0048] Strategy Validation: The generated strategy is immediately sent to the strategy validation and simulation unit. The digital twin model loads the current real system state, energy storage SOC, current photovoltaic output, etc., and imports 1000 possible random weather and load deviation scenarios that may occur in the next 24 hours for Monte Carlo simulation. Validation is passed only if the strategy can ensure that the system does not exceed limits and the overall cost does not exceed the limit in more than 95% of the scenarios.

[0049] Command Issuance: Verified strategies are executed and issued by the real-time control interface unit. For example, the energy storage charging and discharging schedule is issued to the energy storage power station's EMS via the IEC 61850 MMS protocol; information inducing discounts for charging at service area D during 10:00-14:00 is pushed to the highway travel service app and roadside information boards via HTTPS API. The platform itself does not directly close or open switches, but rather drives the lower-level system to execute commands, completing the closed loop.

[0050] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to the method section.

[0051] 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 smart energy dispatch-based integrated energy operation and management platform, characterized in that, include: The monitoring subsystem is used to collect energy consumption data, traffic flow data, and meteorological environmental data in real time. The analysis subsystem is used for load forecasting, power generation forecasting, carbon emission accounting, and traffic-energy consumption correlation analysis based on multi-source heterogeneous data collected by the monitoring subsystem. The control subsystem is used to dynamically generate energy dispatch strategies based on intelligent dispatch algorithms to achieve dynamic balance between energy sources, grid, load, and storage. The monitoring subsystem, the analysis subsystem, and the control subsystem are decoupled and interact with each other through a RESTful API.

2. The energy exchange and integrated operation management platform based on intelligent energy dispatch according to claim 1, characterized in that, The monitoring subsystem includes a multimodal data acquisition unit, a data cleaning and preprocessing unit, and a spatiotemporal alignment unit; The multimodal data acquisition unit is used to collect energy consumption data, traffic flow data and meteorological environment data, and transmit them back through a dual-mode asynchronous transmission channel composed of 5G communication and LoRa communication. The data cleaning and preprocessing unit is used to remove outliers from the collected data based on the improved K-means clustering algorithm and to smooth the traffic flow data using a sliding window mean filtering algorithm. The spatiotemporal alignment unit is used to align the spatiotemporal dimensions of the cleaned and preprocessed multi-source data through timestamp synchronization protocols and geographic information system coordinate mapping.

3. The energy exchange and integrated operation management platform based on intelligent energy dispatch according to claim 1, characterized in that, The energy consumption data includes energy consumption data for highway service areas, toll stations, tunnel lighting, and charging piles; the collected traffic flow data includes traffic volume and vehicle speed data for the highway mainline and ramps.

4. The energy exchange and integrated operation management platform based on intelligent energy dispatch according to claim 1, characterized in that, The analysis subsystem includes a multi-task joint modeling unit, a traffic-energy consumption correlation mining unit, and a distributed computing unit; The multi-task joint modeling unit adopts a spatiotemporal attention mechanism model based on the Transformer architecture to simultaneously process load forecasting, power generation forecasting, and carbon emission accounting tasks. The traffic-energy consumption correlation mining unit uses the Granger causality test algorithm to quantify the dynamic correlation between traffic flow and energy consumption, and constructs a traffic flow-energy consumption elasticity coefficient matrix. The distributed computing unit, built on the Apache Flink stream processing framework, is used for real-time parallel computing of the multi-task joint modeling unit and the traffic-energy consumption correlation mining unit.

5. The energy exchange and integrated operation management platform based on intelligent energy dispatch according to claim 4, characterized in that, The traffic-energy consumption correlation mining unit also includes dynamic correlation analysis of traffic flow and energy consumption of service areas and toll stations, as well as analysis of electric vehicle charging behavior.

6. The energy exchange and integrated operation management platform based on intelligent energy dispatch according to claim 4, characterized in that, The multi-task joint modeling unit includes: An LSTM-XGBoost hybrid model is used to combine time series features with external meteorological factors for prediction; A LSTM network modified by the Clear Sky model is used for photovoltaic power generation prediction, and a CNN-LSTM model based on numerical weather prediction data is used for wind power prediction. The system performs multi-dimensional carbon emission accounting by industry, region, and energy type using a pre-set IPCC default emission factor library.

7. The energy exchange and integrated operation management platform based on intelligent energy dispatch according to claim 1, characterized in that, The control subsystem includes an intelligent scheduling algorithm engine, a strategy verification and simulation unit, and a real-time control interface unit. The intelligent scheduling algorithm engine is used to execute the following optimization process to generate a dynamic energy scheduling strategy; The strategy verification and simulation unit constructs a virtual energy system model based on digital twin technology, which is used to verify the robustness of the dynamic energy dispatch strategy through Monte Carlo simulation. The real-time control interface unit, using the IEC 61850 protocol, is used for bidirectional data interaction and control command issuance with external energy management systems, highway control systems, and energy storage devices.

8. The energy exchange and integrated operation management platform based on intelligent energy dispatch according to claim 7, characterized in that, The intelligent scheduling algorithm engine includes: a) Construct a traffic-electricity coupled stochastic user equilibrium model to predict the path flow of electric vehicles in the road network and the potential vehicle arrival rate at service area charging stations; b) Construct a service quality assessment module based on queuing theory to calculate the average waiting time and user churn probability at charging stations; c) An improved genetic algorithm is used to minimize the total operating cost of the system, maximize the absorption of renewable energy, and minimize carbon emissions. The vehicle arrival rate and service quality are incorporated as constraints or optimization factors to perform multi-objective optimization.

9. A smart energy dispatch-based integrated energy operation and management platform according to claim 8, characterized in that, The intelligent scheduling algorithm engine also includes: A master-slave iterative optimization framework is adopted, in which the distribution side generates planning candidate schemes that include priority photovoltaic consumption, and the transmission side conducts adaptive evaluation and cost accounting of the schemes. Through iteration, a collaborative scheduling scheme that ensures service quality and economic benefits is sought. Set grid security constraints, energy storage device charge and discharge rate constraints, and traffic signal timing constraints; A dynamic weight adjustment mechanism is adopted to dynamically adjust the weights of each optimization objective in the objective function based on real-time electricity prices, fluctuations in renewable energy output, and traffic congestion index.

10. The energy exchange and integrated operation management platform based on intelligent energy dispatch according to claim 1, characterized in that, It also includes a storage subsystem, which adopts a hybrid architecture of time-series database and relational database to store real-time monitoring data and analysis results data respectively, and to perform historical data backtracking analysis.

Images