An urban rail transit energy management system considering the integration of new energy and energy storage

Abstract

The present invention provides an urban rail transit energy management system that takes into account the integration of new energy and energy storage, including: a new energy power generation unit, an energy storage unit, a regenerative braking energy recovery unit, and an energy management and control platform; the new energy power generation unit and the energy storage unit are connected to the rail transit power supply system; the energy management and control platform is connected to the power supply network and communicates and controls the above-mentioned units, monitoring, analyzing, and optimizing the energy flow in the rail transit power supply system to coordinate the utilization of new energy power generation, energy storage charging and discharging, and regenerative braking energy, improve energy utilization efficiency, and reduce the demand for purchased electricity from the power grid. The present invention effectively reduces the system's peak load and the cost of purchasing electricity from the power grid, while improving overall energy utilization efficiency, ensuring that the rail transit system achieves energy conservation and consumption reduction while operating safely and stably.

Description

Technical Field

[0001] The present invention belongs to the technical field of urban rail transit, and in particular relates to an urban rail transit energy management system that takes into account the access of new energy and energy storage. Background Art

[0002] Currently, urban rail transit energy management systems primarily focus on monitoring traditional power consumption and simple dispatch management. These systems offer relatively limited functionality and have yet to fully consider the integrated application of new energy and energy storage technologies. In actual operation, rail transit power supply systems often struggle to effectively balance load fluctuations due to the large fluctuations in renewable energy output and the limited capacity of energy storage equipment. This leads to significant power supply inefficiencies and energy waste. Traditional systems lack coordination in key areas such as renewable energy generation forecasting, energy storage scheduling, and regenerative braking energy recovery. They also lack a comprehensive control mechanism based on real-time data and intelligent algorithms, making them unable to simultaneously meet the requirements for safe and efficient power supply and energy conservation and consumption reduction. Summary of the Invention

[0003] The purpose of the present invention is to address the deficiencies in the above-mentioned background technology and to provide an urban rail transit energy management system that takes into account the access of new energy and energy storage. The system can optimize the charging and discharging strategies and regenerative braking energy recovery parameters of the energy storage device based on the prediction of new energy output and the monitoring of real-time load, effectively reduce the system peak load and the grid power purchase cost, and at the same time improve the overall energy utilization efficiency, ensuring that the rail transit system achieves energy saving and consumption reduction while operating safely and stably.

[0004] The technical solution adopted by the present invention is: an urban rail transit energy management system that takes into account the access of new energy and energy storage, including: a new energy power generation unit, an energy storage unit, a regenerative braking energy recovery unit, and an energy management and control platform; the new energy power generation unit and the energy storage unit are connected to the rail transit power supply system; the energy management and control platform is connected to the power supply system and communicates and controls the above-mentioned units, monitors, analyzes and optimizes the energy flow in the rail transit power supply system to coordinate the utilization of new energy power generation power, energy storage charging and discharging, and regenerative braking energy, improve energy utilization efficiency and reduce the demand for purchased electricity from the power grid.

[0005] In the above technical solution, the data network architecture of the energy management control platform adopts a three-layer architecture design, including:

[0006] The terminal perception layer collects equipment operation data, environmental data, and passenger flow data through various on-site intelligent instruments and sensors, and implements data monitoring;

[0007] The edge computing layer uses intelligent communication gateways and switches to send data collected by terminals to the energy management control platform through a dedicated network channel according to standard protocols, and transmits the platform's control commands to each control unit.

[0008] The system application layer, with the energy management control platform as the core, realizes the storage of all subway station data, the display of equipment operating status, and the issuance of control instructions.

[0009] In the above technical solution, the control and management architecture of the energy management control platform adopts a three-level hierarchical architecture, including:

[0010] At the field level, field instruments and sensors are used to collect equipment operating conditions and environmental data and perform local control.

[0011] At the station level, monitor, statistically analyze and optimize energy consumption within the station;

[0012] At the line level, a control center is set up to centrally monitor and make dispatching decisions on the energy of the entire line, and issue control instructions to each station level to achieve energy optimization and coordinated control across the entire line.

[0013] In the above technical solution, the energy management control platform is used to predict the output power of the new energy power generation unit and optimize the charging and discharging scheduling of the energy storage unit according to the prediction results to smooth out the fluctuations in the new energy power generation output.

[0014] In the above technical solution, the energy management control platform is used to reduce the peak power load of the rail transit system through demand management, thereby reducing the basic electricity charge, and adjust the charging and discharging strategy of the energy storage unit according to the time-of-use electricity price, storing electricity during off-peak hours and releasing electricity during peak hours to reduce the overall electricity cost.

[0015] In the above technical solution, the energy management control platform is used to dynamically adjust the operating threshold of the regenerative braking energy recovery unit based on the real-time power flow conditions of the traction power supply network to ensure line energy balance and maximize the utilization rate of braking energy recovery.

[0016] In the above technical solution, the energy management and control platform establishes a power flow calculation model that includes the AC distribution network and the DC traction network. This model is used to analyze the power flow of the rail transit power supply system that includes renewable energy and energy storage, providing a decision-making basis for energy scheduling.

[0017] The power flow analysis process includes:

[0018] (a) Initialization phase: Initialize the parameters of key devices according to preset parameters, and set the initial values of the rectifier unit output voltage, DC traction voltage, and AC side node voltage;

[0019] (b) AC distribution network calculation: Based on the AC distribution network parameters of the rail transit power supply system, the node admittance matrix Y is formed and the voltage of each AC node is calculated;

[0020] (c) First traction voltage iteration: Initially, the traction voltage in the DC traction network is calculated based on the initially set rectifier output voltage. At subsequent times, the rectifier output voltage obtained from the previous round of calculations is used to iteratively calculate the traction voltage in the DC traction network. This is done until the traction voltage changes from two consecutive iterations meet the preset first convergence condition, thus converging the DC traction network portion of the power flow calculation model.

[0021] (d) Preliminary rectifier output power calculation: Based on the initially set rectifier unit output voltage at the initial moment, and the updated rectifier unit output voltage obtained in the previous round of calculation at subsequent moments, the initial output power of the rectifier unit is calculated based on the power flow calculation model;

[0022] (e) Second traction voltage iteration and power update:

[0023] ① Update the rectifier output voltage based on the power flow calculation results of the AC side node voltage and the rectifier DC side voltage;

[0024] ② Based on the updated rectifier output voltage, the traction voltage in the DC traction network is iteratively calculated until the change in traction voltage from two consecutive iterations meets the preset convergence condition, thus converging the DC traction network portion of the power flow calculation model.

[0025] ③ Calculate the output power of the current round of rectifier units based on the converged power flow calculation model and the updated rectifier output voltage;

[0026] ④ Compare the rectifier unit output power calculated in the current round with the initial output power recorded in step (d). If the difference meets the preset power convergence condition, the power flow calculation at this moment is considered complete; otherwise, return to step (d) and repeat the above iterative process until the convergence condition is met.

[0027] (f) Time step update:

[0028] When the traction voltage and rectifier unit output power at this moment reach the preset convergence conditions, the traction voltage and AC node voltage obtained at the current moment are used as the initial values for the power flow calculation at the next moment, and steps (c) to (e) are repeated until the entire power flow analysis is completed.

[0029] In the above technical solution, the energy storage unit includes a battery energy storage subunit and a flywheel energy storage subunit; the energy management control platform adopts a reinforcement learning algorithm optimized by proximal strategy to perform intelligent scheduling and control of the energy storage unit, and dynamically adjusts the charging and discharging power of the battery and flywheel according to real-time load, renewable energy power generation and electricity price information to maximize economic benefits and minimize the amount of electricity purchased from the power grid.

[0030] In the above technical solution, the energy management control platform is also used to collect station environmental parameters and passenger flow information, and adjust the operation of the station's air-conditioning and ventilation-related environmental control loads according to changes in electricity prices, so as to reduce the energy consumption of the environmental control system while ensuring passenger comfort.

[0031] In the above technical solution, the energy management platform adopts a phased optimization and control strategy, including: a local optimization stage that prioritizes identifying and optimizing energy-consuming equipment with the greatest energy-saving potential, followed by an overall optimization stage that coordinates the energy efficiency of each station, and a global optimization stage that comprehensively considers multiple factors such as fluctuations in renewable energy output, changes in electricity prices, passenger flow, and energy storage life; the control strategies generated by the optimization in each stage are iteratively superimposed step by step to gradually approach the optimal solution for the energy utilization efficiency of the entire system.

[0032] The beneficial effects of the present invention are as follows: the present invention integrates new energy power generation devices, energy storage devices and regenerative braking energy recovery units, and is uniformly monitored and dispatched by an energy management control platform, thereby achieving coordinated optimization of various energy units in the rail transit power supply system, thereby improving overall energy utilization efficiency, reducing dependence on external power grid purchases, and improving power supply stability and economy.

[0033] Furthermore, the present invention adopts a three-tier architecture design consisting of a terminal perception layer, an edge computing layer, and a system application layer, which clearly divides the work between field data collection, data transmission, and central control. This structure ensures real-time and accurate data collection and transmission, facilitates remote centralized monitoring and refined control, and thus improves the responsiveness of energy management and system reliability.

[0034] Furthermore, the present invention utilizes a three-level control architecture: field, station, and line-level. This effectively connects local control with centralized dispatching. The field level enables local control, the station level optimizes local energy consumption, and the line-level control center makes global dispatching decisions. This enables coordinated energy optimization across the entire line, enhancing overall system stability and energy savings.

[0035] Furthermore, the present invention combines a new energy power prediction module with an energy storage scheduling module to accurately predict the output of new energy (such as photovoltaic) power generation, and optimizes the charging and discharging strategy of the energy storage device accordingly, thereby smoothing out fluctuations in new energy output, improving the stability and reliability of the power supply system, and ensuring efficient matching of energy supply and demand.

[0036] Furthermore, the present invention is provided with demand management and peak-valley electricity price response functions, which can monitor and control the peak power load of the rail transit system in real time, and use electricity price fluctuations for dynamic adjustment, thereby reducing basic electricity charges and overall electricity costs and maximizing economic benefits.

[0037] Furthermore, the present invention dynamically adjusts the charge and discharge thresholds of the DC regenerative braking energy recovery device to ensure the recovery of residual energy generated during the train braking process, thereby improving the energy recovery utilization rate, reducing energy waste, and further reducing the system's dependence on external power supply.

[0038] Furthermore, the present invention establishes a hybrid AC / DC power flow calculation model to perform real-time analysis of the power flows of the AC and DC networks within the rail transit power supply system, providing accurate data support for energy scheduling. This data-based analysis can help the system optimize energy distribution, improve the accuracy of scheduling decisions, and enhance overall operational efficiency.

[0039] Furthermore, the present invention adopts a reinforcement learning scheduling algorithm based on proximal policy optimization (PPO) to intelligently schedule energy storage devices and dynamically adjust the charging and discharging power according to real-time load, renewable energy output and electricity price information, thereby maximizing economic benefits, reducing the power grid's electricity purchase costs, and enhancing the system's adaptive control capabilities.

[0040] Furthermore, the present invention is equipped with an environmental control energy consumption optimization function, which collects environmental parameters, passenger flow data and equipment operating status in the station in real time, and dynamically adjusts the operating strategies of environmental control equipment such as air conditioning, ventilation, and exhaust in combination with electricity price information, thereby achieving reduced energy consumption of the environmental control system and reduced overall operating costs while ensuring passenger comfort and safety.

[0041] Furthermore, the present invention adopts a phased energy optimization control strategy (including three stages: local optimization, overall optimization and global optimization) to gradually iterate and optimize the system scheduling strategy, so that the optimization measures at all levels are gradually superimposed to approach the optimal state of energy utilization efficiency of the entire system, thereby achieving comprehensive energy saving and consumption reduction and improving the economy of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 The overall architecture diagram of the urban rail transit energy management system considering the integration of new energy and energy storage;

[0043] Figure 2 This is the overall software architecture diagram of the energy management control platform;

[0044] Figure 3 A schematic diagram of an iterative process of an energy-saving control algorithm applied in a specific embodiment;

[0045] Figure 4 Schematic diagram of the urban rail transit power supply system considering the integration of photovoltaic and energy storage;

[0046] Figure 5a A schematic diagram of a simplified device model of an equivalent circuit model of an urban rail DC traction power supply system according to a specific embodiment;

[0047] Figure 5b This is a schematic diagram of the overall model of the equivalent circuit model of the urban rail DC traction power supply system in a specific embodiment;

[0048] Figure 6 This is a flow chart for calculating AC / DC hybrid power flow in urban rail transit considering the integration of photovoltaic and energy storage;

[0049] Figure 7 Flowchart of the reinforcement learning-based energy scheduling algorithm applied in a specific embodiment. DETAILED DESCRIPTION

[0050] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments to facilitate a clear understanding of the present invention, but they do not constitute a limitation to the present invention.

[0051] like Figure 1 As shown, the present invention provides an urban rail transit energy management system that takes into account the access of new energy and energy storage, including: a new energy power generation unit, an energy storage unit, a regenerative braking energy recovery unit, and an energy management control platform; the new energy power generation unit and the energy storage unit are connected to the rail transit power supply system; the energy management control platform is connected to the power supply network and communicates and controls the above units, monitors, analyzes and optimizes the energy flow in the rail transit power supply system to coordinate the utilization of new energy power generation power, energy storage charging and discharging, and regenerative braking energy, improve energy utilization efficiency and reduce the demand for purchased electricity from the power grid.

[0052] Specifically, the data network architecture of the energy management control platform adopts a three-layer architecture design, including:

[0053] The terminal perception layer uses various on-site intelligent instruments and sensors to collect and monitor equipment operating data, environmental data, and passenger flow data. The system primarily uses various on-site intelligent instruments and sensors to collect and monitor equipment and environmental data. By integrating multiple sensors, the system can monitor equipment energy consumption, operating status, and environmental changes in real time. This sensor data is used for subsequent analysis and diagnosis, helping to identify potential energy waste or equipment failures. Basic on-site data collection equipment includes multi-function electricity meters and intelligent chiller control systems. It also integrates with the station's environmental and equipment monitoring systems, integrated automation systems, safety and control systems, and video surveillance data for centralized management.

[0054] The edge computing layer mainly collects data from on-site intelligent instruments and sensors through intelligent communication gateways, switches and other equipment, and sends it to the energy management system (EMS) through a dedicated network channel according to standard protocols, and transmits the platform's control commands to each control unit.

[0055] The system's application layer, centered around the energy management and control platform, stores all subway station data, displays equipment operating status, and issues control commands. The energy management and control platform provides remote, centralized monitoring of multiple subway stations. Local staff can monitor all operating conditions via a web interface, enabling unmanned or understaffed operation.

[0056] Specifically, the control and management architecture of the energy management control platform adopts a three-level hierarchical architecture, including:

[0057] At the field level, field instruments and sensors are used to collect equipment operating conditions and environmental data and perform local control.

[0058] At the station level, energy consumption within the station is monitored, statistically analyzed, and optimized. The station-level energy management system mainly includes the BAS system, power quality monitoring system, and ISCS system to collect energy and environmental data within the station, and realize station-level online monitoring, statistics, analysis, and diagnosis functions.

[0059] At the line level, a control center centrally monitors energy consumption and makes dispatch decisions for the entire line, distributing control instructions to each station to achieve coordinated energy optimization across the entire line. The line level enables data sharing between various systems, analyzing and displaying energy consumption data for the entire line. Energy usage indicators are calculated based on the existing indicator system, providing data support for energy consumption level assessments and supporting decision-making information for train scheduling and overall dispatching.

[0060] like Figure 1 As shown, the architecture of this embodiment fully embodies a multi-layered collaborative model encompassing "site-station-line / control center": the site layer is responsible for low-level perception and preliminary control, the station layer implements local energy consumption management and execution control, and the control center centrally performs advanced data analysis and scheduling decisions. A dedicated subway network tightly connects these layers, forming a scalable, secure, reliable, and comprehensive rail transit energy management system. This provides strong technical support for the intelligent scheduling and energy-saving strategies of this invention, primarily comprising site-level networks, subway networks, and line-level / control center networks.

[0061] Site-level networks include:

[0062] Field sensors, intelligent instruments and control equipment are deployed in each station, depot or parking lot, such as multi-function electricity meters, environmental sensors, passenger flow detectors, and chiller intelligent control systems.

[0063] These field devices are interconnected via industrial switches or intelligent communication gateways to form a station-level local area network. The gateway is responsible for packaging the collected data according to a unified protocol and performing edge computing (such as simple threshold determination and fault detection) when necessary.

[0064] The station-level network can be locally configured with one or more station servers to monitor and conduct statistical analysis on energy consumption data within the station, and can automatically adjust adjustable loads such as air conditioning, lighting, and ventilation according to energy-saving control instructions issued by superiors.

[0065] The metro network includes:

[0066] The local area network of each station, depot or parking lot is interconnected with the line-level or control center network through dedicated communication lines (i.e. "metro network").

[0067] The subway network usually takes the form of a private network or virtual private network (VPN), with security encryption and bandwidth guarantee to meet the real-time and reliability requirements of the rail transit system.

[0068] The line-level / control center network includes:

[0069] Core switches and server clusters are deployed in the control center or line-level data center, and are equipped with large-capacity storage devices.

[0070] These servers host the main functional modules of the energy management control platform (system application layer), including:

[0071] Database and historical data storage: Centrally store real-time and historical energy consumption data, equipment status information, etc. from each station.

[0072] Visual monitoring and alarm: Provides a graphical interface for operation and maintenance personnel to view the energy consumption and equipment status of the entire line in real time. If there is an abnormal energy consumption or equipment failure, an alarm can be automatically triggered.

[0073] Advanced scheduling and optimization algorithms: such as AC / DC hybrid power flow calculation, PPO reinforcement learning scheduling, phased optimization strategy, demand response, etc., issue control instructions in real time by analyzing the data sent from each station.

[0074] After the control platform generates a scheduling strategy (such as energy storage charging and discharging plans, regenerative braking energy recovery threshold adjustment, station air conditioning and ventilation optimization, etc.), these instructions are transmitted down through the subway network to the gateway or server of each station, and then specifically executed by the station server.

[0075] The terminal perception layer (site level) is mainly responsible for on-site data collection and preliminary monitoring. It uses sensors and smart meters to obtain real-time information on energy-consuming equipment, environmental parameters, and passenger flow, and can perform local control.

[0076] The edge computing layer (station / line level) consists of intelligent gateways and switches, responsible for data aggregation, transmission, filtering, and some edge-side computing (such as simple alarm detection). It also maintains two-way communication with the control center, distributing commands from the control center to devices at each station.

[0077] The system application layer (control center) deploys the core energy management system, including data storage, visualization, alarm management, scheduling optimization algorithms, etc., to achieve comprehensive energy consumption analysis and intelligent scheduling across the entire line.

[0078] The present invention implements basic functions, statistical analysis and energy-saving control based on the above architecture.

[0079] In terms of basic functions, the system offers data collection and storage, real-time monitoring, statistical analysis, query and display, report printing, user rights management, alarm indication, and software updates. Through data collection and storage, the system can collect and store various energy data in real time; the real-time monitoring function allows for comprehensive tracking of equipment and energy usage; the statistical analysis function supports comprehensive analysis of energy data, helping users understand energy consumption trends; the query and display function provides flexible data query and display, allowing users to easily view historical and real-time energy consumption data; the report printing function automatically generates various reports and provides the necessary analysis results; user rights management ensures access control for different roles; the alarm indication function issues an alarm when energy consumption is abnormal, ensuring a timely response; and the software update function ensures that the system remains up-to-date to adapt to changing needs.

[0080] Regarding statistical analysis, the system provides functions such as energy monitoring, energy regulation, energy monitoring, energy statistics, energy disclosure, energy early warnings and alarms, indicator management, and energy billing. The energy monitoring and regulation functions help maintain real-time visibility into the system's operating status and ensure compliance with energy usage. Energy monitoring supports precise tracking of energy consumption and accurate calculation of each device's energy usage. The energy statistics and disclosure functions facilitate the display and analysis of energy consumption data, enhancing system transparency. The energy early warning and alarm functions can issue alarms when energy consumption exceeds standards or equipment fails, preventing unnecessary losses in advance. The indicator management function helps users set and monitor key energy efficiency indicators, providing data support for energy-saving goals. The energy billing module calculates and manages electricity costs based on energy usage, ensuring control of system operation and maintenance costs.

[0081] Regarding energy-saving control, the system features environmental control and energy-saving, intelligent peak shaving, energy storage scheduling, smart lighting, an expert system, load shifting, and algorithm updates. The environmental control and energy-saving module reduces energy consumption by optimizing control of air conditioning, ventilation, and other equipment. The intelligent peak shaving function adjusts power consumption based on load demand to avoid peak loads. The energy storage scheduling function improves energy efficiency by intelligently managing the battery energy storage system. Smart lighting optimizes the operation of lighting equipment through intelligent control, reducing unnecessary energy consumption. The expert system provides intelligent decision-making support through big data analysis, helping managers optimize energy use. The load shifting function balances loads across different time periods, reducing electricity costs. The algorithm update function continuously optimizes the energy-saving algorithm, improving the system's energy-saving effectiveness.

[0082] The terminal perception layer collects real-time data on equipment operating status, environmental parameters, and passenger flow through various intelligent instruments and sensors installed on-site in rail transit systems (such as multi-function electricity meters, temperature sensors, humidity sensors, energy consumption sensors, and passenger flow counters). These devices are typically embedded directly in on-site control units or on-site collection terminals and provide preliminary data monitoring and preprocessing capabilities.

[0083] The edge computing layer aggregates data collected on-site through intelligent communication gateways and industrial switches using standard communication protocols (such as MODBUS and OPC UA). The aggregated data is then uploaded to the central energy management and control platform via a dedicated network channel. The aggregated data undergoes preliminary processing on the edge devices. If an abnormality is detected (such as a sudden increase in energy consumption or equipment failure), an early warning signal is immediately triggered and the alarm information is transmitted to the system application layer.

[0084] At the system's application layer, data is stored in servers and storage repositories at the control center, forming a system database that provides the foundation for subsequent statistical analysis and scheduling decisions. Central monitoring software automatically generates alarm reports by comparing real-time data with preset thresholds. It also displays abnormal conditions through a graphical interface, facilitating timely response and maintenance by operations and maintenance personnel. After data collection and storage, the energy management and control platform runs a statistical analysis module that comprehensively processes historical and real-time data, generating energy consumption trend charts, comparative analysis reports, and various statistical reports. The system also provides a user-friendly query and display interface, allowing operators to query data and display results as needed. The report printing module automatically outputs corresponding analysis reports to inform decision-making. The management software within the control center implements hierarchical permission management for different users (such as field operators, dispatchers, and administrators), ensuring that each role can only access and operate data and control functions within their authorized scope. A software update module ensures that the platform regularly receives the latest software and algorithm updates to adapt to changing operational requirements and technological advancements, ensuring long-term, efficient and stable system operation.

[0085] like Figure 2 As shown, this embodiment can further divide the data processing architecture from bottom to top into: field equipment layer, data acquisition layer, data preprocessing layer, algorithm model layer and intelligent control layer.

[0086] The field device layer includes:

[0087] Flywheel energy storage is mainly used to quickly absorb or release energy, such as storing regenerative energy during train braking, or releasing energy to assist in power supply during peak power consumption.

[0088] The cooling system provides cooling for air conditioners or other equipment with high heat dissipation requirements to maintain safe and stable operation of the equipment.

[0089] The fresh air system controls the air circulation and quality in stations and tunnels, providing a comfortable environment and removing harmful gases.

[0090] Photovoltaic, distributed photovoltaic power generation system is a new energy power generation unit that converts solar energy into electrical energy to provide part of the power supply for the rail transit system.

[0091] Energy storage refers to battery energy storage units, which are used to charge during periods of low electricity prices or sufficient photovoltaic power generation, and discharge during periods of peak electricity prices or sudden load increases to smooth out power supply fluctuations.

[0092] Thermo-hygrometer detects ambient temperature and humidity in real time, providing reference data for the regulation of air conditioning or ventilation systems.

[0093] Electricity meters monitor the energy consumption or power generation at key nodes, accurately measure energy flow, and provide data support for subsequent energy consumption analysis and control strategies.

[0094] Low-level field devices are the physical foundation for the system to acquire real-time data and execute control instructions. They are responsible for directly interacting with energy or the environment, providing input for upper-level data collection and decision-making, and executing corresponding regulatory measures.

[0095] The data collected by the data collection layer includes:

[0096] Environmental data and outdoor environmental data: from temperature and humidity meters, fresh air systems, etc., reflecting the station and external meteorological conditions.

[0097] Equipment energy consumption data: collected by electricity meters, energy storage and flywheel equipment, recording equipment operating power, voltage, current, etc.

[0098] Air conditioning status data and air conditioning control data: including air conditioning operation mode, temperature setting, fan speed and other information, providing key input for subsequent optimization and scheduling.

[0099] The data acquisition layer collects the real-time status, energy consumption information and environmental parameters of the underlying equipment through sensors and metering devices of on-site equipment, and packages them into a standardized format, laying the data foundation for subsequent data preprocessing and algorithm analysis.

[0100] The data processes processed by the data preprocessing layer include:

[0101] Data cleaning: Filter out noise, duplication, or outliers to ensure data quality.

[0102] Data conversion: Convert data in different protocols or formats into a unified standard (such as unified timestamps, units, etc.) to facilitate subsequent analysis.

[0103] Detection (or “anomaly detection”): Automatically identifies abnormal conditions such as sudden increases in energy consumption, equipment failures, or sensor failures, and generates alerts.

[0104] Data completion: interpolate or infer missing or discontinuous data to maintain data continuity as much as possible.

[0105] The data preprocessing layer is the "data quality gatekeeper". It cleans, normalizes, identifies anomalies and corrects the collected multi-source heterogeneous data, providing highly reliable and consistent input data for the algorithm model, and avoiding scheduling decision errors caused by poor original data.

[0106] The algorithm models used in the algorithm model layer include:

[0107] Expert Experience Pool: Based on operational maintenance experience and historical statistical information, it provides reference thresholds or parameter ranges (temperature alarm values, humidity alarm values, equipment loss changes, air conditioning parameter adjustment ranges, etc.) for the system in boundary or special scenarios. For example, the upper / lower limits of air conditioning temperature, humidity warning values, and equipment life assessment are provided.

[0108] Photovoltaic power generation prediction module: Utilizes historical power generation data, real-time meteorological information, and photovoltaic power curves to make short-term predictions of photovoltaic output, providing a reference for energy storage and load scheduling.

[0109] Environmental control energy consumption prediction model: Combines indoor and outdoor environmental data, passenger flow changes, air conditioning operating conditions, etc. to predict the demand for environmental control loads such as air conditioning and ventilation in the future, avoiding blind cooling / heating waste.

[0110] Environmental Control Energy Consumption Optimization Algorithm: Based on predictions and the set comfort range, the algorithm dynamically adjusts the operating strategies of equipment such as air conditioning and fresh air systems to reduce environmental control system energy consumption while maintaining passenger comfort. By monitoring environmental parameters such as temperature, humidity, and air quality on platforms and in train cars, as well as passenger flow fluctuations, equipment operating status, and energy prices in real time, the algorithm dynamically adjusts the operating strategies of environmental control equipment. This algorithm combines energy demand prediction with equipment performance optimization to rationally allocate and utilize energy resources, while ensuring passenger comfort and safety, thereby reducing the overall energy consumption of the environmental control system. By optimizing the operating efficiency of equipment such as air conditioning, ventilation, and exhaust, the algorithm effectively reduces peak energy consumption during subway operations and fully utilizes energy resources during off-peak hours, achieving globally optimal energy management and ultimately achieving the goals of cost savings and improved operational efficiency.

[0111] Energy storage intelligent peak shaving algorithm: For battery energy storage or flywheel energy storage equipment, based on electricity prices, load forecasts, and photovoltaic power generation forecasts, a charging and discharging plan is formulated to reduce the maximum load peak, profit from the peak-valley electricity price difference, and ensure system stability.

[0112] The algorithmic model layer is the "core brain" of system decision-making. It provides decision-making solutions for the upper layer's "intelligent control" through the collaborative work of various predictive models, optimization algorithms, and expert experience knowledge bases. This layer focuses on computational analysis of multi-source data to form scientific and accurate optimization strategies.

[0113] The functions implemented by the intelligent control layer include:

[0114] Air conditioning energy saving: Combined with environmental control energy consumption prediction and optimization algorithms, it dynamically adjusts air conditioning temperature settings, fan speed, etc. to achieve a balance between energy saving and comfort.

[0115] Ventilation energy saving: According to the indoor and outdoor air quality and passenger flow, the opening degree or operation time of the fresh air system is controlled to reduce the energy consumption of the ventilation system.

[0116] Photovoltaic power generation: Utilize photovoltaic forecast results to determine photovoltaic power generation grid connection or energy storage charging strategies, maximizing the proportion of renewable energy self-generation and self-use.

[0117] Energy storage peak shaving: When electricity prices or loads are at their peak, stored energy is released to reduce electricity purchases from the external grid; charging occurs during off-peak hours or when there is surplus photovoltaic power, smoothing system load fluctuations and saving costs.

[0118] Flywheel energy storage and braking energy feedback: absorbs and stores the train's braking energy and releases it at the appropriate time to further reduce traction energy consumption.

[0119] The intelligent control layer implements the optimization strategy output by the algorithm model layer. Through real-time scheduling of equipment such as air conditioning, energy storage, photovoltaics and flywheel energy storage, it dynamically matches load demand and energy supply to achieve the goals of peak shaving and valley filling, saving energy consumption, reducing operating costs and improving overall energy efficiency.

[0120] In this embodiment, the system data processing process proceeds from bottom to top: the underlying equipment operates in real time and generates large amounts of data; the data acquisition layer collects data from multiple sources; the data preprocessing layer cleans, converts, and detects anomalies in the data; the algorithm model layer conducts comprehensive analysis and decision-making based on expert experience pools, predictive models, and optimization algorithms; and the intelligent control layer executes scheduling strategies (for air conditioning, ventilation, photovoltaics, energy storage, flywheel feedback, etc.), providing real-time feedback on execution results and updating data, forming a closed loop that continuously iterates to improve system energy efficiency. Through this five-layer structure, the present invention achieves multi-objective coordinated optimization of renewable energy utilization, energy storage peak regulation, regenerative braking energy recovery, and environmental control while ensuring the safe operation of rail transit. This significantly reduces operating energy consumption and electricity costs, and improves the overall energy efficiency and sustainable development level of urban rail transit systems.

[0121] Preferably, the BAS system utilizes an environmental control energy consumption prediction model and an environmental control energy consumption optimization algorithm to conduct state searches based on the equipment's operating status and duration, and adjusts the control strategy according to the target state. The controller can employ various operational modes, such as adjusting chiller frequency, adjusting water supply temperature, and optimizing water pump frequency. Data analysis identifies and prioritizes operations with the greatest potential, thereby improving the energy efficiency of the air conditioning system. The system intelligently schedules based on factors such as load data, equipment performance, and energy costs to avoid energy waste. It also dynamically adjusts the air conditioning system's operating mode based on real-time indoor temperature and humidity data, enhancing the equipment's autonomous operation capabilities and ensuring a balance between energy conservation and comfort. The system can identify hazardous, promising, and feasible states, prioritizing hazardous states and concentrating computing resources on promising and feasible states, gradually finding the optimal control path.

[0122] Preferably, the present invention adopts a photovoltaic power generation system as a new energy power generation device, converts solar energy into direct current through multiple photovoltaic modules, and converts direct current into alternating current through an inverter for use by the rail transit system. The system is combined with a photovoltaic power generation prediction module to estimate future photovoltaic power generation in advance by combining meteorological data, historical power generation data and load demand forecasts. This prediction result provides support for power dispatching and energy storage management to ensure the stability and reliability of power supply. Based on the energy storage intelligent peak shaving algorithm, the system adjusts the charging and discharging strategy of the battery energy storage system according to the changes in photovoltaic power generation, stores excess electricity for use in periods of insufficient sunlight, and dynamically adjusts the collaboration between photovoltaic power generation and energy storage system according to real-time load demand. For example, during peak load periods, the photovoltaic system can give priority to charging energy storage equipment to ensure that the energy storage system releases enough electricity to support rail transit operation during peak power consumption periods, thereby achieving load balancing and optimizing power dispatching.

[0123] Preferably, the flywheel energy storage system is mainly used for energy recovery and scheduling of rail transit. The system uses an intelligent energy storage peak shaving algorithm or a flywheel energy storage braking energy feedback algorithm to convert the train's excess kinetic energy into the rotational energy of the flywheel for storage during the train braking process. When the train needs to accelerate or encounters a peak in power demand, the flywheel energy storage system will quickly release the stored energy to reduce dependence on the external power grid. Flywheel energy storage can provide high-power power output in a short period of time, has the characteristics of efficient energy storage and rapid release, and significantly improves the overall energy utilization of the traction system.

[0124] The algorithm model layer also optimizes energy consumption monitoring. With a comprehensive meter configuration, the system enables comprehensive online monitoring of energy consumption data for the entire rail transit system. The system supports precise metering by category, item, and region, and automatically generates energy consumption reports, providing detailed data analysis. Real-time monitoring of key energy consumption components, such as vehicle traction and regenerative braking systems, allows for year-on-year, month-on-month, and horizontal comparative analysis, helping managers understand energy usage trends and promptly identify anomalies. The system automatically calculates and benchmarks energy consumption indicators, triggering alerts when preset thresholds are exceeded, ensuring accurate energy efficiency management. The energy consumption forecasting module predicts future energy consumption based on historical data and real-time information, supporting the scientific formulation of quota indicators. The system also provides peak and valley time-of-use metering, accurately distinguishing energy consumption differences between different time periods, helping to optimize energy use and reduce costs. Furthermore, the system supports maximum power metering, cost accounting, transfer station cost allocation, and external unit cost calculation, comprehensively improving energy management efficiency. Through intelligent data analysis and report generation, the system provides strong support for energy optimization and energy conservation and emission reduction, and promotes the development of rail transit systems in an efficient and sustainable direction.

[0125] Preferably, this embodiment, based on an intelligent energy storage peak-shaving algorithm and an environmental control energy consumption optimization algorithm, implements intelligent scheduling of the power supply system. This optimizes energy usage, reduces the maximum power demand of the lines, and thus reduces basic electricity bills. The system comprehensively evaluates the benefits of photovoltaic power generation and energy storage systems and, through intelligent scheduling, rationally arranges power generation and load, further improving the system's economic efficiency. Intelligent scheduling can shift some load during peak demand periods to periods with lower electricity prices, thereby maximizing electricity efficiency and reducing overall operating costs. The system also features intelligent load regulation, monitoring the load status of each station and its three-phase power supply in real time to ensure proper load distribution and avoid overload. By dynamically detecting load imbalances, the system proactively issues adjustment recommendations, optimizes power distribution, reduces energy loss in the power supply system, and improves power utilization. Furthermore, the power supply system is equipped with load rate monitoring and early warning functions to monitor equipment capacity utilization in real time. If high-load equipment is detected, the system automatically issues an early warning signal, enabling operators to conduct timely inspections and maintenance, preventing equipment overload and ensuring a stable and reliable power supply system. Through these intelligent scheduling and monitoring functions, the power supply system achieves more efficient and economical energy management.

[0126] like Figure 3 As shown, this embodiment uses a phased optimization strategy to improve energy efficiency, reduce waste, and ensure sustainable system operation. The algorithm can be divided into three development stages, gradually transitioning from local optimization to global optimization, ultimately achieving a comprehensive improvement in energy utilization.

[0127] During the local optimization phase, the system first collects operational data, energy consumption information, and environmental parameters (such as air conditioning energy consumption, lighting energy consumption, and equipment load) from various on-site devices and sensors. The "Data Analysis" module cleans, compiles, and visualizes this raw data, helping managers understand the current system energy consumption profile and identify high-energy-consuming devices or weak links. Based on the collected energy consumption data and device characteristics, digital models are built, including device energy consumption models, load forecasting models, and energy storage / new energy models. During this phase, the system utilizes various algorithms (such as regression analysis, neural networks, and expert systems) to construct a mathematical description of device energy consumption or performance, laying the foundation for subsequent energy-saving optimization and integrated scheduling. After optimizing a specific device type or site, the system evaluates the energy-saving effects by comparing energy consumption indicators before and after optimization to quantify the savings. The evaluation results are fed back to the core algorithm layer to facilitate parameter updates and policy adjustments in subsequent iterations.

[0128] During the local optimization phase, energy-saving algorithms are iterated for typical energy-consuming equipment or processes, such as lighting, chillers, composite energy storage, and medium-voltage energy feed, to identify the optimal energy-saving control strategy for each. In this phase, the algorithm uses previously described data and models (such as energy consumption forecasts and equipment characteristic models) to conduct in-depth local optimization of specific equipment.

[0129] The system will iterate the operating parameters (temperature setting, frequency adjustment, start and stop time, etc.) of the selected equipment type (such as a chiller) multiple times:

[0130] 1. Initial strategy: based on expert experience or historical average strategy.

[0131] 2. Calculate energy-saving potential: Evaluate the energy consumption level of the current strategy through simulation or actual operation comparison.

[0132] 3. Update strategy: Use digital models or heuristic algorithms to adjust control parameters. If the energy saving effect is improved and the comfort / safety constraints are not violated, the new strategy is accepted.

[0133] 4. Iterative loop: Continuously repeat the measurement and update to gradually approach the optimal or suboptimal local energy consumption level.

[0134] When several equipment categories have completed local energy-saving iterations and obtained better strategies, the system will summarize these results and provide optimized subsystem control solutions for the next level of "comprehensive optimization", that is, global optimization.

[0135] Globally optimize external or system-level dynamic factors such as "wind power and photovoltaics, electrochemical energy storage, real-time electricity prices, weather, and passenger flow", and focus on system-level energy scheduling, including renewable energy access, energy storage equipment charging and discharging strategies, passenger peak energy consumption management, and real-time electricity price response.

[0136] The system combines optimized strategies for local devices with real-time information about the external environment (such as renewable energy fluctuations, electricity prices, and passenger flow) to achieve optimal overall energy allocation through multi-objective or multi-stage scheduling algorithms. For example, load balancing and peak shaving: flexibly dispatching energy storage devices and controllable loads during different time periods to avoid peak loads; price response: minimizing grid purchases during periods of high electricity prices or charging and storing energy during periods of low electricity prices or surplus photovoltaic power generation; and multi-objective optimization: balancing energy conservation, economic efficiency, equipment lifespan, and system safety and stability.

[0137] Global optimization ultimately outputs a system-level energy-saving scheduling plan, covering the interaction between various types of equipment and external energy sources: for example, photovoltaic output is given priority to supplying loads, surplus power is stored in batteries, discharge is used to subsidize loads when electricity prices are high, and air conditioning operation is appropriately shifted when the load is too high.

[0138] Furthermore, the first phase is a local optimization phase, which involves load selection and optimization based on the "wooden barrel principle." The energy-saving algorithm uses this principle to identify the load types with the greatest energy-saving potential in the system, prioritizing those with the greatest potential and impact for optimization and implementing local energy-saving measures. Through detailed evaluation and prioritization of load types, the system can focus on high-energy-consuming devices or weak links in the system, using multiple iterations of the optimization algorithm to gradually improve local energy efficiency.

[0139] In this embodiment, the optimization goal of the first phase is to optimize the energy consumption of a single device or site to reduce its energy consumption. Mathematically, this can be expressed as minimizing the total energy consumption of a device over a given period. For example, for high-energy-consuming devices such as air conditioners, the goal is to minimize the energy consumption integral E:

[0140] ;

[0141] Among them, P 设备 represents the power consumption of the device, which depends on the control decision u(t) (such as the air conditioner temperature setpoint, fan speed, etc.). t represents the time, and T represents the control period. By optimizing the control decision u(t), the total energy consumption E is minimized.

[0142] In local optimization, constraints such as comfort and equipment operation safety need to be met, including but not limited to:

[0143] Environmental comfort constraints: For example, the station air conditioning temperature must be kept within the passenger comfort range. This ensures energy savings without compromising passenger comfort.

[0144] Equipment safety constraints: Ensure that equipment operates within its rated safety range. For example, the air conditioner compressor cannot start and stop too frequently, and the motor cannot be overloaded, to avoid shortening equipment life or causing failures.

[0145] Therefore, the above optimization problem is essentially a constrained minimization problem. We can introduce Lagrange multipliers or penalty function methods to deal with constraints to ensure that comfort and safety requirements are not violated during the optimization process.

[0146] During the local optimization phase, the system analyzes energy consumption data and leverages expert experience to identify the load types with the highest energy consumption and greatest potential for energy savings (this embodies the "barrel principle" of prioritizing addressing the weakest link). Once a target load, such as the air conditioning or lighting system, is identified, a dedicated optimization algorithm is employed to optimize it.

[0147] For example, for an environmentally controlled air conditioning system, an air conditioning dynamics model (i.e., an environmental control energy consumption prediction model) is established to predict future temperatures and loads. The environmental control energy consumption prediction model uses indoor and outdoor environmental data, passenger flow information, and equipment operating status to predict the demand for environmental control loads such as air conditioning and ventilation over a period of time. For example, a model can be constructed:

[0148] ;

[0149] Where T(t), H(t), Q(t) and N(t) represent temperature, humidity, air quality and passenger flow data respectively.

[0150] The environmental control energy consumption prediction model uses machine learning / regression models to predict energy consumption for air conditioning and other equipment based on environmental conditions. Input variables include historical environmental control system energy consumption data; environmental parameters (such as outdoor temperature and humidity, and station passenger flow); and current air conditioning / ventilation equipment settings. The model fits historical data to map the relationship between environmental changes and energy consumption, providing a reference for decision-making.

[0151] The environmental control energy consumption optimization algorithm dynamically adjusts control parameters (such as chiller frequency and water supply temperature) based on the prediction results to minimize energy consumption while ensuring passenger comfort. Its optimization goal can be written as:

[0152] ;

[0153] Constraints include: comfort (for example, belonging to a specific temperature and humidity range), equipment operation safety (preventing overload and frequent start and stop), actual load demand meeting basic operating requirements, air quality and safety constraints: ventilation and air exchange meet standards.

[0154] The expert experience database provides initial strategies and adjustment rules (such as adjusting temperature setpoints and start and stop times based on historical data and expert knowledge). This is then combined with real-time data for iterative optimization. Through multiple iterations, the energy efficiency of the equipment is gradually improved. Once optimization is complete for one device, a similar process is repeated for the next high-energy-consuming device, gradually reducing the energy consumption of each local load.

[0155] Taking the subway station air conditioning system as an example, energy consumption analysis revealed that air conditioning refrigeration units account for 40% of the station's total energy consumption, representing the greatest potential for energy savings. Expert experience determined that the passenger comfort temperature is around 26°C. The system then uses an environmental control energy consumption optimization algorithm to adjust air conditioning operation: during periods of low passenger flow, the supply air temperature is appropriately increased by 1°C (ensuring a station temperature between 5°C and 27°C), reducing compressor output. The station hall is also pre-cooled before peak passenger flow to avoid excessive cooling energy consumption during peak hours. Following this optimization, the station's daily air conditioning power consumption decreased by approximately 10%. Throughout this process, the temperature remained consistently within comfort standards (for example, around 26°C), and the equipment operated within a safe range without overload. This demonstrates that the local optimization algorithm achieves energy savings by fine-tuning individual devices.

[0156] Furthermore, the second phase involves overall optimization, refining load control strategies and improving energy efficiency. As the load optimization algorithms are continuously refined, the system accumulates a wealth of data and experience, gradually improving overall energy utilization. Through global scheduling and resource optimization, the needs of different load types are comprehensively considered to improve the energy efficiency of the entire system. The system not only optimizes the use of individual loads but also intelligently allocates energy across them, balancing the energy needs of different regions and devices. This effectively avoids excessive energy consumption and waste, achieving more efficient energy utilization.

[0157] The primary task of the overall optimization phase is to coordinate and control loads within multiple stations or regions, achieving energy resource balance among these loads and thus improving overall energy efficiency across the entire network. This phase focuses on energy allocation and scheduling, ensuring that local optimization results are coordinated and consistent across the entire system. It does not specifically address price responsiveness or peak-load shifting (these issues will be introduced in the global optimization phase).

[0158] The photovoltaic power generation prediction model is used to estimate the photovoltaic output in future periods, providing data support for energy distribution among various loads in the region, so that the system can make full use of renewable energy when allocating energy.

[0159] PV power generation prediction models are trained using historical power and meteorological factors (e.g., time series models or neural networks). Future weather forecasts are used as input to output predicted power. Input variables may include: historical PV power generation data; meteorological data (future solar irradiation, temperature, etc.); and time (sunshine duration, season). The PV power prediction values generated by the PV power generation prediction model must not be negative and must not exceed installed capacity. Reliability under weather uncertainty must be considered to avoid overestimating power generation.

[0160] Utilizing the data and forecasts obtained during each local optimization phase, the energy demand and equipment operating status of each station are summarized. The optimization objective can be to minimize fluctuations in network-wide energy consumption or maximize the utilization of renewable energy. This phase primarily utilizes linear programming or heuristic scheduling algorithms to optimize resources and balance energy consumption across regions, achieving load smoothing across the entire network. Since dynamic electricity prices are not considered during this phase, the scheduling process primarily determines the energy allocation ratio for each region based on photovoltaic power generation forecasts, load forecasts, and historical data distribution.

[0161] The inputs of the optimization objective function in this stage include: load data after local optimization of each station, historical statistical data, and local prediction results (such as energy consumption forecasts for air conditioning, lighting, ventilation and other equipment and photovoltaic power generation forecasts).

[0162] The outputs of the objective function solution include: the network-wide scheduling plan, that is, the energy allocation plan for each region, the smoothed load curve, and the adjustment suggestions for the control parameters of each load.

[0163] Constraints include:

[0164] Energy balance constraint: ensures the balance of supply and demand across the entire network at all times;

[0165] Equipment capacity constraints: The power control of each device cannot exceed its maximum tolerance;

[0166] Comfort and safety: Ensure that the distribution plan balances the load without affecting passenger comfort and safe operation of equipment.

[0167] Furthermore, the third stage involves global optimization, achieving systematic energy management under the coupled multi-factor approach. The algorithm considers a more complex multi-factor coupling relationship, including the volatility of renewable energy generation, the safety and lifespan of electrochemical energy storage, real-time electricity prices, real-time passenger flow, and meteorological factors. Renewable energy generation is highly random and volatile, while the safety and lifespan of electrochemical energy storage systems also require comprehensive consideration. By combining real-time electricity prices, weather, and passenger flow data, the system can intelligently adjust energy usage strategies to achieve optimal resource allocation, thereby improving overall energy utilization and ensuring the stability and sustainability of energy supply.

[0168] The global optimization stage requires accurate understanding of the operating status of the entire power supply network (such as the voltage, current, and power distribution of each node). Figure 4The urban rail transit power supply system used in this embodiment is demonstrated. The subway power supply system can be divided into two parts: the AC power supply network and the DC traction network. The AC power supply network mainly includes a switchgear, a traction step-down substation, a step-down substation, photovoltaic power generation, and battery energy storage. The switchgear receives medium-voltage AC power from the preceding medium-voltage AC substation. Typically, a switchgear is located at the head and tail of each line. Each station is equipped with a traction step-down substation or a step-down substation. The traction step-down substation includes a traction substation and a step-down substation. The traction substation rectifies the medium-voltage AC power to the DC traction network voltage using a rectifier transformer and a rectifier device, and then supplies power to the DC traction network. The step-down substation steps down the medium-voltage AC power to low-voltage AC using a transformer, and then supplies power to low-voltage loads such as lighting, elevators, and air conditioning. Flywheel energy storage is connected to the DC traction network. It can be seen that the urban rail transit power supply system has typical distributed characteristics.

[0169] Figure 5a-5b This embodiment provides an equivalent circuit model of a DC traction power supply system for urban rail transit. Figure 5a As shown, the rectifier unit can be equivalent to a voltage source with unidirectional current flow, where U s is the no-load voltage of the rectifier unit, i s is the output current, R s is the equivalent resistance of the rectifier unit; the flywheel energy storage is equivalent to two unidirectional voltage sources, U ce is the charging threshold, R ce is the charging resistor, U de is the discharge threshold, R de is the discharge resistor; the train can be equivalent to a current source, I t is the train current, U brk is the braking resistor action threshold, R as is the auxiliary energy consumption equivalent resistance. Figure 5b This is the overall model of the urban rail traction power supply system. The rectifier unit is connected in parallel with the flywheel energy storage to form a simplified substation model. The size of the up (down) variable resistor and the rail resistance depends on the distance between the up (down) train and the substation.

[0170] Establish a hybrid power flow calculation model (in Figure 6 ), Figure 4 Figure 5 and Figure 5 provide the necessary physical parameters and structural information to support global energy scheduling decisions and ensure that strategies such as energy regulation and peak shaving and valley filling are effectively implemented across the entire network.

[0171] like Figure 6 As shown, the power flow calculation method for AC / DC hybrid urban rail transit adopted in this embodiment includes the following steps:

[0172] Step A: Based on the port characteristics of key equipment such as rectifier units, trains, and flywheel energy storage, construct their equivalent circuit models. Using a modular modeling approach, establish a complete DC traction power supply system power flow calculation model for the DC traction power supply system.

[0173] Step B: According to the AC side network and specific parameters of the line power supply system, the node admittance matrix Y is formed; assuming that the initial voltage of each node is , then the real and imaginary parts of the voltage satisfy:

[0174] ;

[0175] In the line power supply system, the photovoltaic power generation system is configured as a PQ node, and the power value of this node is determined according to the current photovoltaic power generation status; the battery energy storage system is a PQ node, and the power value of this node is confirmed according to the instructions obtained by the scheduling algorithm.

[0176] Step C: At t=0, let the traction voltage be , then the traction current Satisfies the formula:

[0177] ;

[0178] Where n is the number of substations.

[0179] At subsequent moments, the traction voltage calculated in the previous round is used as .

[0180] According to the power flow calculation model of the DC traction power supply system, based on the initially set rectifier unit output voltage at the initial moment, the rectifier unit output voltage obtained by the previous round of calculation is used in subsequent moments to calculate the traction voltage as follows: , then the voltage correction is: .

[0181] Repeated calculation of traction voltage Until the following formula is satisfied, the DC part of the power flow calculation model converges:

[0182] .

[0183] Step D: According to the output voltage of each rectifier unit and output current , calculate the output power of the rectifier units at each site :

[0184] ;

[0185] Among them, at the initial moment, the output voltage of the rectifier unit adopts the corresponding initial setting value, and at subsequent moments, the output voltage of the rectifier unit updated by the previous round of calculation is adopted. The output current of the rectifier unit Obtained through analysis of the current converged power flow calculation model.

[0186] The AC output power of the rectifier unit in the DC traction substation meets the following requirements:

[0187] ;

[0188] Where α is the commutation angle.

[0189] Step E: Voltage correction equation according to the Newton-Raphson method:

[0190] ;

[0191] P i ′ and Q i ′ usually represents the active power imbalance and reactive power imbalance of node ii.

[0192] Calculate ΔP AC(i) , △Q AC(i) , △U AC(i) , where G ij Y in the admittance matrix ij The real part, B ij Y in the admittance matrix ij The imaginary part of Y ij =G ij +jB ij .

[0193] Calculate each element in the Jacobian matrix (H ij 、N ij 、J ij 、L ij 、R ij 、S ij ), when i≠j:

[0194] .

[0195] According to the formula ;

[0196] Calculate the voltage correction amount, and the new node voltage satisfies the formula:

[0197] ;

[0198] Repeat this step until the voltage correction value satisfies the convergence formula:

[0199] .

[0200] Step F: Based on the power flow calculation results of the node voltage of the AC part of the DC traction power supply system and the calculation of the DC side voltage of the rectifier, the rectifier output voltage is obtained. In this embodiment, the rectifier output voltage formula is:

[0201] ;

[0202] where R sw is the commutation resistance, satisfying the formula:

[0203] ;

[0204] Among them L sw is the commutation inductor.

[0205] Recalculate the traction voltage iteratively based on the rectifier unit output voltage With traction current , until the following convergence conditions are met: .

[0206] Calculate the output power of the rectifier units at each site at this time and power variation , the power change satisfies the formula: ;

[0207] in, Use the calculation result of step D.

[0208] Repeat steps D-F until the power variation satisfies the following convergence conditions: .

[0209] Step G: Use the traction voltage and node voltage obtained from the previous power flow calculation as the initial values for the next long-term power flow calculation. Repeat steps C-F to perform power flow calculation until the calculation is completed.

[0210] The global optimization phase further integrates dynamic factors, such as random fluctuations in renewable energy generation, the safety and lifespan of electrochemical energy storage, real-time electricity prices, passenger flow, and meteorological conditions, based on overall scheduling. The goal is to achieve optimal energy scheduling across the entire system by coupling these factors, minimizing the grid's power purchase costs and maximizing economic benefits.

[0211] During this phase, the system needs to integrate various dynamic factors (such as renewable energy output fluctuations, load variations, energy storage charging and discharging characteristics, real-time electricity prices and passenger flow, and weather conditions) to develop an optimal dispatching strategy for the entire system. The AC / DC hybrid power flow calculation method constructs a comprehensive model of the AC and DC traction grid, calculating the voltage, current, and power distribution at key nodes in real time, providing accurate grid operating status data for global dispatching decisions. This data is crucial for ensuring supply and demand balance, reducing grid power purchase costs, and ensuring stable system operation when implementing peak shaving, energy storage scheduling, and multi-factor coupled optimization. Therefore, as part of the algorithm model layer, this power flow calculation method primarily supports the global optimization phase.

[0212] like Figure 7 As shown, this stage adopts the reinforcement learning (PPO) scheduling algorithm (i.e., energy storage intelligent peak shaving algorithm): construct a Markov decision process (MDP), define the state s t Contains real-time load, photovoltaic power generation, energy storage SOC, electricity price, passenger flow and weather information; action a t is the charging and discharging power of the energy storage device (battery and flywheel). The immediate reward function is designed as:

[0213] ;

[0214] Among them, P grid (t) represents the power purchased from the grid. A value greater than 0 indicates power purchase from the grid, while a value less than 0 indicates power sales to the grid. π(t) represents the electricity price during the time period. This design ensures that the system can make appropriate decisions based on electricity prices in different scenarios: charging (power purchase) is encouraged when prices are low, while discharging (power sales) is encouraged when prices are high. Therefore, the key is to ensure that the system's action space and load data allow for negative values, so that power sales behavior can be reflected in the mathematical model.

[0215] The reinforcement learning goal is to maximize the cumulative discounted reward:

[0216] .

[0217] The PPO algorithm updates the policy parameters through a “cutting” mechanism, and its objective function is:

[0218] ;

[0219] in is the strategy ratio, and A is the advantage function.

[0220] In the reinforcement learning framework, photovoltaic prediction results, load forecast, energy storage status and external electricity price signals are used as input to ensure that the intelligent body can adjust the energy storage charging and discharging strategy according to the real-time status to achieve peak shaving and valley filling and maximize economic benefits.

[0221] Specifically, the inputs to the reinforcement learning (PPO) scheduling algorithm include real-time load data, photovoltaic power generation forecasts, energy storage device SOC, electricity price information, passenger flow data, and meteorological data. Its output includes energy storage scheduling decisions optimized through reinforcement learning, specifically the charge and discharge power values of energy storage devices at each moment, which in turn determine the minimum scheduling plan for purchasing / selling electricity from the grid for the entire network. Its constraints include:

[0222] Energy storage device constraints: For example, the battery SOC must be kept within a certain range, and the charge and discharge power must not exceed the device rating;

[0223] Energy balance: Ensure supply and demand balance within each time period. Ensure load = grid power supply + energy storage discharge - energy storage charging + photovoltaic power generation at all times. Peak shaving does not affect the power supply of critical loads.

[0224] Multi-objective balance: Penalties for energy storage life and safety are introduced into the reward function to prevent the strategy from over-consuming energy storage equipment.

[0225] In summary, the local optimization stage uses expert experience and environmental control energy consumption prediction and optimization algorithms to perform local regulation on individual loads (such as air conditioning systems), reduce equipment energy consumption, and meet comfort and safety constraints.

[0226] The overall optimization stage integrates local optimization results and balances the energy consumption of each station across the entire network through data fusion and resource scheduling algorithms to achieve load balance and improve overall energy utilization efficiency. This stage does not focus on electricity price response or peak shaving regulation, but only on energy distribution optimization.

[0227] The global optimization phase incorporates multiple factors (such as electricity prices, passenger flow, and weather conditions) and a reinforcement learning (PPO) scheduling algorithm to globally optimize the charging and discharging strategies of energy storage equipment, achieving peak load shifting and minimizing electricity purchase costs. This phase comprehensively considers energy storage safety, equipment lifespan, and economic benefits to develop a globally optimal scheduling strategy.

[0228] Figure 7 The following demonstrates the solution process for an energy storage scheduling algorithm based on a reinforcement learning (Proximal Policy Optimization, PPO) framework. This algorithm uses flywheel storage power and battery storage power as primary action outputs and updates the policy through an actor-critic architecture. The following explains its working principles step by step:

[0229] 1. Environment

[0230] Input: system status, such as the SOC (state of charge) of the flywheel and battery, current load demand, real-time electricity prices, passenger flow information, weather conditions, etc.

[0231] Output: After the algorithm selects a set of energy storage charging and discharging actions, the environment will feedback the new state (updated SOC, load, etc.) and immediate reward (based on the grid purchase cost or economic benefits).

[0232] The environment is expanded through the "Consideration of multi-factor characteristics (battery life, real-time electricity prices, meteorological information)" module to ensure that scheduling decisions comprehensively reflect external uncertainties.

[0233] 2. Actor_new and Actor_old

[0234] The Actor network is responsible for outputting actions based on the current state. Here, the action is {flywhee-power,battery-power}, which is the charge and discharge power of the flywheel energy storage and battery energy storage.

[0235] "Actor_new" represents the updated policy network in the current iteration, and "Actor_old" represents the policy network from the previous iteration. The PPO algorithm first fixes the old policy (Actor_old) in each training round, then interacts with the environment to collect data, and then updates the new policy (Actor_new) based on this data during the optimization phase.

[0236] 3. Critic Network (Critic_new)

[0237] The Critic network is used to evaluate the value (Value Function or QFunction) of the current state or state-action pair to assist in calculating the advantage function.

[0238] In PPO, the Critic network is updated based on the interaction data between the old and new strategies, outputting an estimate of the environmental return and providing an evaluation basis for the policy gradient.

[0239] 4. Sampling and data collection

[0240] The algorithm uses Actor_old to interact with the environment over a period of time (one round or multiple time steps) to generate a batch of four-tuples of state, action, reward, and next state.

[0241] These interaction data (or "experiences") are temporarily stored and used to calculate policy gradients and value errors after the round ends.

[0242] 5. Mini-batch sampling

[0243] Divide the collected data into small batches to ensure the stability and efficiency of gradient calculation. Each small batch contains several "(state, action, reward, next state)" examples.

[0244] 6. Calculate the Advantage Function

[0245] The advantage function A(s,a) reflects the superiority of an action a relative to the average strategy action. Typical calculation methods include generalized advantage estimation (GAE).

[0246] 7. Update strategy: PPO shear mechanism

[0247] In PPO, the probability ratio of the new strategy to the old strategy is compared and limited to the range of [1−ϵ,1+ϵ] to avoid instability caused by excessive policy updates.

[0248] Loss function: .

[0249] By performing gradient descent on this objective, the updated policy parameters θnew are obtained. The critic network is also updated synchronously to better evaluate the value function.

[0250] 8. Update Actor_new and Critic_new

[0251] Based on the calculated gradient, the network parameters of Actor_new and Critic_new are updated by gradient descent. After completion, Actor_new will replace Actor_old and become the policy network for the next round of interaction.

[0252] 9. Iteration loop

[0253] Repeat the above process:

[0254] Use Actor_old to interact with the environment and collect data;

[0255] Calculate advantages and value errors based on batch data;

[0256] Use PPO shear mechanism to update Actor_new and Critic_new;

[0257] Replace Actor_old with Actor_new and enter the next round of interaction.

[0258] As the number of training rounds increases, the energy storage scheduling strategy is continuously optimized, and the intelligent agent gradually learns when to charge, when to discharge, and how much electricity to discharge under different electricity prices, loads, and SOC conditions, thereby maximizing overall economic benefits.

[0259] Through these three phases, the system gradually expands from local device optimization to network-wide coordination, ultimately incorporating dynamic external factors to achieve optimal scheduling, forming a closed-loop energy management system. Each phase is interconnected: local optimization provides basic data and preliminary results for overall scheduling, while global optimization ensures balanced energy consumption across regions. Global optimization builds on this foundation by further incorporating external factors (such as real-time electricity prices) to achieve optimal economic benefits.

[0260] This phased optimization strategy not only enables the system to achieve energy saving at the individual load level, but also realizes efficient, intelligent and economical operation of the entire rail transit power supply system through full network scheduling and multi-factor coupling.

[0261] The contents not described in detail in this specification belong to the prior art known to those skilled in the art.

Claims

1. An urban rail transit energy management system that considers the integration of new energy and energy storage, characterized by: include: A new energy power generation unit, an energy storage unit, a regenerative braking energy recovery unit, and an energy management and control platform; the new energy power generation unit and the energy storage unit are connected to the rail transit power supply system; the energy management and control platform is connected to the power supply system and communicates and controls the above-mentioned units, monitors, analyzes, and optimizes the energy flow in the rail transit power supply system to coordinate the utilization of new energy power generation, energy storage charging and discharging, and regenerative braking energy, improve energy utilization efficiency, and reduce the demand for purchased electricity from the power grid; The energy management and control platform establishes a power flow calculation model that includes the AC distribution network and the DC traction network, and uses this model to analyze the power flow of the rail transit power supply system that includes new energy and energy storage access, providing a decision-making basis for energy scheduling; The power flow analysis process includes: (a) Initialize the parameters of key devices according to preset parameters, and set the initial values of the rectifier unit output voltage, DC traction voltage, and AC side node voltage; (b) Form the node admittance matrix Y based on the AC distribution network parameters of the rail transit power supply system and calculate the voltage of each AC node; (c) at an initial moment, based on an initially set rectifier unit output voltage, and at subsequent moments, using the rectifier unit output voltage updated from the previous round of calculation, iteratively calculate the traction voltage in the DC traction network until the traction voltage change from two consecutive iterations satisfies a preset first convergence condition, thereby converging the DC traction network portion of the power flow calculation model; (d) calculating the initial output power of the rectifier unit based on the initially set output voltage of the rectifier unit at the initial moment and the output voltage of the rectifier unit updated by the previous round of calculation at subsequent moments based on the power flow calculation model; (e) updating the rectifier output voltage based on the power flow calculation results of the AC side node voltage and the rectifier DC side voltage; Based on the updated rectifier output voltage, the traction voltage in the DC traction network is iteratively calculated until the change in the traction voltage in two consecutive iterations meets a preset convergence condition, so that the DC traction network part of the power flow calculation model converges; Calculate the output power of the current round of rectifier units based on the converged power flow calculation model and the updated rectifier output voltage; Compare the rectifier unit output power calculated in the current round with the initial output power recorded in step (d). If the difference meets the preset power convergence condition, the power flow calculation at this moment is considered complete; otherwise, return to step (d) and repeat the above iterative process until the convergence condition is met. (f) When the traction voltage and rectifier unit output power at this moment have reached the preset convergence conditions, the traction voltage and AC node voltage obtained at the current moment are used as the initial values for the power flow calculation at the next moment, and steps (c) to (e) are repeated until the entire power flow analysis is completed.

2. The system according to claim 1, wherein: The data network architecture of the energy management control platform adopts a three-layer architecture design, including: The terminal perception layer collects equipment operation data, environmental data, and passenger flow data through various on-site intelligent instruments and sensors, and implements data monitoring; The edge computing layer uses intelligent communication gateways and switches to send the data collected by the terminals to the energy management control platform through a dedicated network channel according to standard protocols, and then sends the platform's control commands to each control unit; The system application layer, with the energy management control platform as the core, realizes the storage of all subway station data, the display of equipment operating status, and the issuance of control instructions.

3. The system according to claim 1, wherein: The control and management architecture of the energy management control platform adopts a three-level hierarchical architecture, including: At the field level, field instruments and sensors are used to collect equipment operating conditions and environmental data and perform local control. At the station level, monitor, statistically analyze and optimize energy consumption within the station; At the line level, a control center is set up to centrally monitor and make dispatching decisions on the energy of the entire line, and issue control instructions to each station level to achieve energy optimization and coordinated control across the entire line.

4. The system according to claim 1, wherein: The energy management control platform is used to predict the output power of the new energy power generation unit and optimize the charging and discharging scheduling of the energy storage unit based on the prediction results to smooth out the fluctuations in the new energy power generation output.

5. The system according to claim 1, wherein: The energy management control platform is used to reduce the peak power load of the rail transit system through demand management, thereby reducing the basic electricity fee, and adjust the charging and discharging strategy of the energy storage unit according to the time-of-use electricity price, storing electricity during off-peak hours and releasing electricity during peak hours to reduce the overall electricity cost.

6. The system according to claim 1, wherein: The energy management control platform is used to dynamically adjust the operating threshold of the regenerative braking energy recovery unit based on the real-time power flow conditions of the traction power supply network to ensure line energy balance and maximize the utilization rate of braking energy recovery.

7. The system according to claim 1, wherein: The energy storage unit includes a battery energy storage subunit and a flywheel energy storage subunit; the energy management control platform uses a reinforcement learning algorithm optimized by proximal strategy to intelligently dispatch and control the energy storage unit, dynamically adjusting the charging and discharging power of the battery and flywheel according to real-time load, renewable energy generation and electricity price information to maximize economic benefits and minimize the amount of electricity purchased from the power grid.

8. The system according to claim 1, characterized in that The energy management control platform is also used to collect station environmental parameters and passenger flow information, and adjust the operation of the station's air-conditioning and ventilation-related environmental control loads according to changes in electricity prices, so as to reduce the energy consumption of the environmental control system while ensuring passenger comfort.

9. The system according to claim 1, wherein: The energy management platform adopts a phased optimization and control strategy, including: a local optimization stage that prioritizes identifying and optimizing energy-consuming equipment with the greatest energy-saving potential, followed by an overall optimization stage that coordinates the energy efficiency of each station, and a global optimization stage that comprehensively considers fluctuations in renewable energy output, changes in electricity prices, passenger flow, and energy storage life. The control strategies generated by the optimization in each stage are iteratively superimposed step by step to gradually approach the optimal solution for the energy utilization efficiency of the entire system.

Images