A tunnel ventilation and lighting multi-system space-time coordination intelligent regulation method and system

Abstract

The application discloses a tunnel ventilation and lighting multi-system space-time coordination intelligent regulation method and system, first, through the multi-type sensors arranged in the tunnel, tunnel environment parameters, traffic flow parameters and equipment operation parameters are collected in real time; then, the collected data is preprocessed and fused by using an edge computing unit, and short-time traffic flow prediction is carried out based on an LSTM network; then, a dynamic energy consumption model of the tunnel ventilation and lighting system is constructed, and an improved multi-objective particle swarm optimization algorithm is adopted to solve the optimal coordination regulation strategy in the space-time dimension under the rigid constraint of meeting the tunnel safe operation specification; finally, the regulation strategy is verified by safety simulation through a digital twin, and after verification, the regulation strategy is executed to the physical equipment. The problems of independent control, poor energy-saving effect and high safety risk of the tunnel ventilation and lighting system in the prior art are solved, fine and intelligent control of tunnel energy is realized, and under the premise of ensuring operation safety, the system average energy-saving rate reaches 20%-28%.

Description

Technical Field

[0001] This invention relates to the field of tunnel energy management technology, and in particular to a method and system for spatiotemporal coordinated intelligent control of multiple tunnel ventilation and lighting systems. Background Technology

[0002] Currently, my country has the world's longest highway tunnel network. Tunnels, as major energy consumers, generally suffer from problems in energy management, including unclear total energy consumption, ambiguous energy structure, crude energy-saving methods, and lagging equipment control. Ventilation and lighting systems are the primary sources of energy consumption in tunnels, accounting for over 70% of total energy consumption. In existing technologies, tunnel ventilation and lighting systems mostly employ independent control methods. Ventilation systems are typically controlled based on fixed time intervals or simple CO / VI concentration thresholds, while lighting systems typically use time-segmented fixed brightness control or simple dimming control based on traffic flow. This control method has the following drawbacks: each system operates independently, lacking collaborative optimization and failing to achieve system-level global optimal energy saving; control strategies are mostly static or based on simple rules of automation, unable to adapt to complex and changing traffic flow and environmental conditions; low control precision, only able to monitor total energy consumption, unable to delve into the dynamic analysis and optimization of each critical loop; and lagging safety control, usually taking measures only after safety indicators exceed limits, posing certain safety risks.

[0003] In recent years, some newly built or renovated large tunnels have begun pilot installations of sub-metering devices and simple monitoring systems. However, their functions are mostly limited to data collection and display, lacking in-depth analysis and intelligent control capabilities. Some research institutions and enterprises have begun to focus on tunnel energy conservation, but there is still significant room for improvement in the accuracy of energy management models, the adaptability of control strategies, and the depth of system integration. Internationally, Europe and Japan started earlier in the field of tunnel energy conservation, developing relatively complete lighting and ventilation design standards. They have also applied frequency conversion control technology based on traffic flow and pollutant concentration in some tunnels. However, their systems are mostly based on individual control or simple linkage, and the hardware and software costs are high. They lack comprehensive solutions suitable for my country's complex geological conditions, long tunnel complexes, and diverse traffic flow characteristics. Therefore, developing a method and system capable of achieving spatiotemporal coordinated intelligent control of multiple tunnel ventilation and lighting systems, minimizing energy consumption while ensuring tunnel operational safety, is a pressing technical problem in this field. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of the prior art and provide a method and system for spatiotemporal collaborative intelligent control of multiple tunnel ventilation and lighting systems. It aims to achieve spatiotemporal collaborative intelligent global optimization of multiple tunnel ventilation and lighting systems under safety constraints, and to achieve refined energy saving, cost reduction and efficiency improvement and inherent safety through AI, digital twins and cloud-edge collaboration.

[0005] According to one objective of the present invention, the present invention provides a method for spatiotemporal coordinated intelligent control of multiple systems for tunnel ventilation and lighting, comprising the following steps: S1. Data Acquisition and Preprocessing: Real-time acquisition of tunnel environmental parameters, traffic flow parameters, and equipment operating parameters is achieved through multiple types of sensors deployed within the tunnel, and the acquired data is cleaned, calibrated, and fused. S2. Short-term traffic flow prediction: Based on preprocessed historical traffic flow data, a short-term traffic flow prediction model is constructed using a long short-term memory network to output the predicted traffic flow of each section of the tunnel within the next 15-60 minutes. S3. Dynamic energy consumption modeling: Based on tunnel structural parameters, equipment technical parameters and real-time operating data, construct minute-level dynamic energy consumption models for the tunnel ventilation system and lighting system respectively; S4. Multi-objective collaborative optimization: With the optimization objectives of minimizing overall system energy consumption and maximizing operational comfort, and with constraints of minimum tunnel illumination requirements, maximum CO concentration requirements, and equipment operation safety requirements, an improved multi-objective particle swarm optimization algorithm is used to solve the coordinated control strategy of ventilation and lighting for each section and time period of the tunnel in the next 15-60 minutes. S5. Digital Twin Security Verification: Input the solved control strategy into the pre-constructed digital twin of the tunnel energy system to perform safety and effect simulation and verify whether the control strategy meets all safety constraints. S6. Strategy Distribution and Execution: The control strategy that has passed security verification is distributed to the edge control unit, which then controls the operation of the fans and lights in the tunnel in real time. S7. Effect Evaluation and Model Update: Collect actual operation data after regulation in real time, evaluate energy-saving effect and safety indicators, and update traffic flow prediction model and energy consumption model regularly based on evaluation results.

[0006] Furthermore, in step S1, the multiple types of sensors include smart meters, power sensors, CO / VI detectors, vehicle detectors, brightness sensors, temperature sensors, and wind speed sensors; the data fusion processing adopts a weighted average method and a Kalman filter algorithm, and the data acquisition and transmission accuracy is not less than 99.9%.

[0007] Furthermore, in step S2, the input of the LSTM short-term traffic flow prediction model includes historical traffic flow data, time feature data, and weather feature data, and the output is traffic flow prediction values ​​at three time scales: 15 minutes, 30 minutes, and 60 minutes, with a prediction error of no more than 10%.

[0008] Furthermore, in step S3, the dynamic energy consumption model of the ventilation system considers the effects of fan operating frequency, number of fans, CO concentration in the tunnel, VI value and traffic flow; the dynamic energy consumption model of the lighting system considers the effects of lamp brightness, number of lamps, natural light brightness outside the tunnel and traffic flow.

[0009] Furthermore, in step S4, the improved multi-objective particle swarm optimization algorithm introduces adaptive inertia weights and crowding factors to improve the diversity of solutions while ensuring convergence speed; the control strategy includes the operating frequency of each fan, the number of fans turned on, and the brightness level and number of lights turned on for each lighting circuit.

[0010] Furthermore, in step S5, the digital twin of the tunnel energy system is constructed at a 1:1 scale and includes the tunnel geometry, digital models of all ventilation and lighting equipment, and dynamic simulation models of energy flow; all control strategies must be verified in the digital twin before they can be issued and executed, and the delay between the generation and issuance of control strategies does not exceed 5 seconds.

[0011] Furthermore, step S7 also includes an equipment health status assessment step: by analyzing the operating current, vibration, temperature and other operating condition data of the equipment, an equipment energy efficiency degradation and health status assessment model is established to realize early fault warning of the equipment, with a warning accuracy rate of not less than 85%.

[0012] According to another objective of the present invention, the present invention provides a multi-system spatiotemporal coordinated intelligent control system for tunnel ventilation and lighting, comprising: The perception layer includes various types of sensors deployed inside the tunnel to collect real-time tunnel environmental parameters, traffic flow parameters, and equipment operating parameters. Edge computing and control layer: This includes multiple intelligent edge control units deployed in various substations or key nodes within the tunnel. These units are used to preprocess and fuse data collected by the perception layer, perform short-term traffic flow prediction, receive and execute control strategies sent from the cloud, and achieve real-time control of fans and lights. The cloud-based big data analysis and optimization platform includes a data storage module, a model training module, a multi-objective optimization module, and a digital twin simulation module. It is used to store historical data, train and update prediction models and energy consumption models, solve for the optimal coordinated control strategy, and perform safety simulation verification of the control strategy. Visualized integrated management and control platform: including energy panorama dashboard, energy consumption analysis module, equipment monitoring module, energy efficiency benchmarking module and early warning alarm module, used to provide managers with intuitive display of system operation status and decision support.

[0013] Furthermore, the intelligent edge control unit deeply integrates the reliability of industrial PLC with the computing power of edge server, supports Modbus, OPC UA or MQTT industrial protocol conversion, and has a built-in safety pre-simulation module that continues to run according to preset safety policies when the network is interrupted, with system availability of no less than 99.5%.

[0014] Furthermore, the digital twin simulation module enables dynamic simulation of energy flow, visual simulation and deduction of control strategies, and pre-evaluation of control effects; the visual integrated management and control platform provides visualization tools such as three-dimensional heat maps of energy consumption, radar charts of equipment energy efficiency benchmarking, and simulation curves of energy-saving strategy effects.

[0015] The technical solution of this invention achieves cross-temporal and global collaborative control of tunnel ventilation and lighting systems through multi-source data acquisition, AI short-term traffic flow prediction, dynamic energy consumption modeling, and multi-objective collaborative optimization, combined with digital twin safety simulation verification. It upgrades traditional independent and extensive control to intelligent closed-loop management under safety constraints, effectively solving problems such as independent system operation, limited energy-saving effect, and delayed safety response. Under the premise of strictly meeting the rigid indicators of tunnel operation safety, it significantly improves the overall energy saving rate and equipment operating efficiency, reduces energy consumption and operation and maintenance costs, and ensures that the control process is safe, reliable, and timely. It provides stable and feasible technical support for the green, low-carbon, intelligent, and efficient operation of tunnels. Attached Figure Description

[0016] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0017] Figure 1 This is an overall architecture diagram of the system according to an embodiment of the present invention; Figure 2 This is a flowchart of the method according to an embodiment of the present invention; Figure 3 This is a flowchart illustrating the digital twin security verification process according to an embodiment of the present invention. Detailed Implementation

[0018] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] Example 1 like Figure 1 As shown in the figure, this embodiment provides a multi-system spatiotemporal collaborative intelligent control system for tunnel ventilation and lighting. The system adopts a cloud-edge collaborative architecture, including four layers: a perception layer, an edge computing and control layer, a cloud big data analysis and optimization platform, and a visual integrated management and control platform.

[0020] The perception layer includes various types of sensors deployed within the tunnel, specifically including: Smart meters and power sensors: used to collect power parameters such as voltage, current, power, and power consumption of various fans and lighting circuits; CO / VI detector: used to collect carbon monoxide concentration and visibility inside the tunnel; Vehicle detector: Used to collect traffic flow parameters such as traffic volume, vehicle speed, and vehicle type in various sections of the tunnel; Brightness sensor: Used to collect brightness values ​​inside and outside the tunnel; Temperature and wind speed sensors: used to collect temperature and wind speed inside the tunnel.

[0021] The edge computing and control layer includes multiple Intelligent Edge Control Units (TEUs) deployed in various substations or key nodes within the tunnel. TEUs deeply integrate the reliability of industrial PLCs with the computing power of edge servers, supporting conversions of multiple industrial protocols such as Modbus, OPCUA, and MQTT, and enabling the acquisition and fusion of multi-source heterogeneous data. TEUs have a built-in safety pre-simulation module, allowing them to continue operating according to preset safety policies even during network outages, ensuring tunnel operational safety. The main functions of the Intelligent Edge Control Unit (TEU) include: Preprocessing and fusing the data collected by the perception layer; Perform short-term traffic flow forecasting; Receive and execute control strategies issued from the cloud; To achieve real-time control of fans and lighting fixtures; Local data caching and resume interrupted downloads.

[0022] The cloud-based big data analytics and optimization platform includes: Data storage module: Employs distributed database technology to store historical operating data, model parameters, and control strategies; Model training module: used to train and update traffic flow prediction models, energy consumption models, and equipment health assessment models; Multi-objective optimization module: An improved multi-objective particle swarm optimization algorithm is used to solve for the optimal cooperative control strategy; Digital twin simulation module: Constructs a digital twin of the tunnel energy system to perform safety simulation verification of the control strategy.

[0023] The visualized integrated management and control platform includes: Energy Panorama Dashboard: Displays the tunnel's overall energy consumption, energy-saving effects, and key operational indicators; Energy consumption analysis module: provides time-sharing, zone-based, and subsystem-based energy consumption analysis and three-dimensional energy consumption heat maps; Equipment monitoring module: Real-time monitoring of the operating status of all equipment, providing equipment energy efficiency benchmarking radar chart; Energy efficiency benchmarking module: Provides horizontal benchmarking analysis with design values, historical values, and similar tunnels; Early warning and alarm module: Enables early warning of abnormal energy consumption and equipment failure.

[0024] Example 2 like Figure 2 As shown, this embodiment provides a method for spatiotemporal coordinated intelligent control of multiple systems for tunnel ventilation and lighting based on the embodiment system, including the following steps: Step S1: Data Acquisition and Preprocessing Multiple types of sensors deployed within the tunnel collect real-time tunnel environmental parameters, traffic flow parameters, and equipment operating parameters. The collection frequency is determined based on the parameter type: power parameters and equipment operating parameters are collected once every 1 minute, while environmental parameters and traffic flow parameters are collected once every 5 minutes.

[0025] The collected data is preprocessed, including: Data cleaning: removing outliers and missing values; Data calibration: Calibrate sensor data to eliminate systematic errors; Data fusion: Weighted average method and Kalman filter algorithm are used to fuse multi-source data to improve the accuracy and reliability of the data.

[0026] Step S2: Short-term traffic flow prediction Based on preprocessed historical traffic flow data, a short-term traffic flow prediction model is constructed using a Long Short-Term Memory (LSTM) network. LSTM is a special type of recurrent neural network (RNN) that can effectively solve the gradient vanishing and gradient exploding problems during long sequence training, making it very suitable for handling time series prediction problems.

[0027] The model's inputs include: Historical traffic flow data: Traffic flow data every 5 minutes over the past 24 hours; Time-related data: hour, day of the week, whether it is a holiday, etc.; Weather characteristic data: weather conditions, temperature, rainfall, etc.

[0028] The model outputs traffic flow predictions for the next 15, 30, and 60 minutes. The model is trained using the Adam optimizer with a learning rate of 0.001, a batch size of 32, and 100 training epochs. After training, the model's short-term traffic flow prediction error does not exceed 10%.

[0029] Step S3: Dynamic Energy Consumption Modeling Based on tunnel structural parameters, equipment technical parameters, and real-time operating data, minute-level dynamic energy consumption models for the tunnel ventilation system and lighting system are constructed respectively.

[0030] The dynamic energy consumption model of the ventilation system is shown in equation (1): E vent (t)= (1); Among them, E vent (t) represents the energy consumption of the ventilation system at time t, n represents the number of fans, and P fan,i f is the power of the i-th wind turbine. i (t) represents the operating frequency of the i-th wind turbine, N i (t) represents the number of the i-th fan that is turned on, CO(t) represents the CO concentration in the tunnel, VI(t) represents the visibility in the tunnel, Q(t) represents the traffic flow, and Δt represents the time step.

[0031] The dynamic energy consumption model of the lighting system is shown in equation (2): E light (t)= (2); Among them, E light (t) represents the energy consumption of the lighting system at time t, m represents the number of lighting circuits, and P lamp,j Let L be the power of the j-th lighting circuit. j (t) represents the brightness level of the j-th lighting circuit, M j (t) represents the number of lights turned on in the j-th lighting circuit, L out Q(t) represents the natural light intensity outside the tunnel, Q(t) represents the traffic flow, and Δt represents the time step.

[0032] Step S4: Multi-objective collaborative optimization With the optimization objectives of minimizing overall system energy consumption and maximizing operational comfort, and with constraints such as minimum tunnel illumination requirements, maximum CO concentration requirements, and equipment operational safety requirements, a multi-objective optimization problem is constructed: Objective function: minF=[E total (t),C(t)](3); Among them, E total (t)=E vent(t)+E light C(t) represents the total energy consumption of the system, and C(t) represents the operational comfort index.

[0033] Constraints: L in (t)≥L min (4); CO(t)≤CO max (5); f min ≤f i (t)≤f max (6); 0≤N i (t)≤N i,max (7); L min,j ≤L j (t)≤L max,j (8); 0≤M j (t)≤M j,max (9); Among them, L in (t) represents the illuminance inside the tunnel, L min To meet the minimum illumination requirements for tunnels, CO max To meet the maximum CO concentration requirement for the tunnel, f min and f max These represent the minimum and maximum operating frequencies of the fan, respectively, N. i,max L represents the maximum number of wind turbines that can be operated at any given time. min,j and L max,j M represents the minimum and maximum brightness levels of the j-th lighting circuit, respectively. j,max This represents the maximum number of lights that can be turned on in the j-th lighting circuit.

[0034] An improved multi-objective particle swarm optimization (MOPSO) algorithm is used to solve the above multi-objective optimization problem. The improved MOPSO algorithm introduces adaptive inertia weights and a crowding factor, which can improve the diversity of solutions while maintaining convergence speed. The specific steps of the algorithm are as follows: Step 401: Initialize the particle swarm, with each particle representing a possible control strategy; Step 402: Calculate the fitness value for each particle; Step 403: Update the individual optimal solution and the global optimal solution; Step 404: Update the particle's velocity and position; Step 405: Calculate the fitness value of the updated particles; Step 406: Update external files; Step 407: Determine if the termination condition has been met. If it has, output the Pareto optimal solution set; otherwise, return to step 3.

[0035] A compromise solution is selected from the Pareto optimal solution set as the final control strategy. The control strategy includes the operating frequency, number of fans activated, brightness level, and number of lighting circuits activated for each tunnel section and time period within the next 15-60 minutes.

[0036] Step S5: Digital Twin Security Verification like Figure 3 As shown, the obtained control strategy is input into a pre-constructed digital twin of the tunnel energy system for safety and effect simulation. The digital twin is constructed at a 1:1 scale and includes digital models of the tunnel geometry, all ventilation and lighting equipment, and a dynamic simulation model of energy flow.

[0037] The specific steps of the simulation are as follows: Step 501: Input the control strategy into the digital twin; Step 502: Run the digital twin to simulate the tunnel's operating status over the next 15-60 minutes; Step 503: Monitor safety indicators such as tunnel illumination and CO concentration during the simulation process; Step 504: Determine whether all safety indicators meet the constraints; Step 505: If the condition is met, output "Verification passed"; if the condition is not met, output "Verification failed" and return to step S4 to re-solve the control strategy.

[0038] All control strategies must be verified in the digital twin before they can be deployed and executed, ensuring absolute security of the control process. The delay between the generation and deployment of control strategies does not exceed 5 seconds.

[0039] Step S6: Strategy Issuance and Execution The control strategy, which has passed safety verification, is sent to the edge control unit. Based on the control strategy, the edge control unit controls the operation of the fans and lights within the tunnel in real time. For the lighting system, a real-time dimming strategy of "lights on when vehicles approach, lights off when vehicles leave" is adopted; for the ventilation system, an intelligent variable frequency control strategy based on pollutant concentration and traffic flow prediction is used.

[0040] Step S7: Effect Evaluation and Model Update Real-time collection of actual operating data after regulation is used to evaluate energy-saving effects and safety indicators. Energy-saving effects are evaluated by comparing energy consumption data before and after regulation; safety indicators are evaluated by monitoring parameters such as illuminance and CO concentration inside the tunnel.

[0041] Based on the evaluation results, the traffic flow prediction model and energy consumption model are updated monthly, and the parameters of the multi-objective optimization algorithm are adjusted quarterly to ensure that the system performance is always at its optimal level.

[0042] In addition, this step includes equipment health status assessment: by analyzing operating data such as equipment current, vibration, and temperature, a model for assessing equipment energy efficiency degradation and health status is established to achieve early warning of equipment failures. When the equipment health status falls below a preset threshold, the system automatically issues a warning message to remind management personnel to perform maintenance. The accuracy rate of equipment anomaly warnings is no less than 85%.

[0043] This invention has been deployed and debugged as a complete system in two demonstration tunnels. Data shows that, while ensuring operational safety (illuminance and CO concentration fully meeting standards), the demonstration tunnels achieved an average annual comprehensive energy saving rate of 20%-28%. Specifically, the lighting system achieved an energy saving rate exceeding 40%, and the ventilation system exceeded 25%. The system availability reached over 99.5%, achieving significant economic and social benefits.

[0044] Compared with the prior art, the present invention has the following beneficial effects: This invention proposes for the first time a new paradigm of tunnel energy management based on multi-system spatiotemporal collaborative dynamic optimization under safety constraints. It optimizes the ventilation and lighting systems as a whole, achieving a leap from single-point energy saving to system-level global optimal energy saving, with an average system energy saving rate of 20%-28%.

[0045] This invention deeply integrates digital twin technology into the control loop, forming a three-step control process of "simulation deduction, strategy pre-screening, and safe execution". It moves safety from the "protection" stage to the "decision-making" stage, achieving inherent safety and effectively avoiding the risks of direct control.

[0046] This invention adopts a cloud-edge collaborative architecture, which pushes some computing tasks to the edge, significantly reducing system response latency. The generation and distribution of control strategies take no more than 5 seconds, which can meet the requirements of real-time tunnel control.

[0047] This invention employs AI-based intelligent decision-making technology, upgrading from rule-based automation to AI model-based intelligence, which can adapt to complex and ever-changing traffic flow and environmental conditions, and improves the adaptability and robustness of the control strategy.

[0048] This invention enables refined management and control, moving from monitoring total energy consumption to dynamic analysis and optimization of each critical circuit, accurately locating energy consumption anomalies and improving energy management levels.

[0049] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for spatiotemporal coordinated intelligent control of multiple systems for tunnel ventilation and lighting, characterized in that, Includes the following steps: S1. Data Acquisition and Preprocessing: Real-time acquisition of tunnel environmental parameters, traffic flow parameters, and equipment operating parameters is achieved through multiple types of sensors deployed within the tunnel, and the acquired data is cleaned, calibrated, and fused. S2. Short-term traffic flow prediction: Based on preprocessed historical traffic flow data, a short-term traffic flow prediction model is constructed using a long short-term memory network to output the predicted traffic flow of each section of the tunnel within the next 15-60 minutes. S3. Dynamic energy consumption modeling: Based on tunnel structural parameters, equipment technical parameters and real-time operating data, construct minute-level dynamic energy consumption models for the tunnel ventilation system and lighting system respectively; S4. Multi-objective collaborative optimization: With the optimization objectives of minimizing overall system energy consumption and maximizing operational comfort, and with constraints of minimum tunnel illumination requirements, maximum CO concentration requirements, and equipment operation safety requirements, an improved multi-objective particle swarm optimization algorithm is used to solve the coordinated control strategy of ventilation and lighting for each section and time period of the tunnel in the next 15-60 minutes. S5. Digital Twin Security Verification: Input the solved control strategy into the pre-constructed digital twin of the tunnel energy system to perform safety and effect simulation and verify whether the control strategy meets all safety constraints. S6. Strategy Distribution and Execution: The control strategy that has passed security verification is distributed to the edge control unit, which then controls the operation of the fans and lights in the tunnel in real time. S7. Effect Evaluation and Model Update: Collect actual operation data after regulation in real time, evaluate energy-saving effect and safety indicators, and update traffic flow prediction model and energy consumption model regularly based on evaluation results.

2. The method according to claim 1, characterized in that, In step S1, the multiple types of sensors include smart meters, power sensors, CO / VI detectors, vehicle detectors, brightness sensors, temperature sensors, and wind speed sensors; the data fusion processing adopts a weighted average method and a Kalman filter algorithm, and the data acquisition and transmission accuracy is not less than 99.9%.

3. The method according to claim 1, characterized in that, In step S2, the input of the LSTM short-term traffic flow prediction model includes historical traffic flow data, time feature data and weather feature data, and the output is the traffic flow prediction value at three time scales: 15 minutes, 30 minutes and 60 minutes in the future, with a prediction error of no more than 10%.

4. The method according to claim 1, characterized in that, In step S3, the dynamic energy consumption model of the ventilation system considers the effects of fan operating frequency, number of fans, CO concentration in the tunnel, VI value and traffic flow; the dynamic energy consumption model of the lighting system considers the effects of lamp brightness, number of lamps, natural light brightness outside the tunnel and traffic flow.

5. The method according to claim 1, characterized in that, In step S4, the improved multi-objective particle swarm optimization algorithm introduces adaptive inertia weight and crowding factor to improve the diversity of solutions while ensuring convergence speed; the control strategy includes the operating frequency of each fan, the number of fans turned on, and the brightness level and number of lights turned on for each lighting circuit.

6. The method according to claim 1, characterized in that, In step S5, the digital twin of the tunnel energy system is constructed at a 1:1 scale and includes the tunnel geometry, digital models of all ventilation and lighting equipment, and dynamic simulation models of energy flow. All control strategies must be verified in the digital twin before they can be issued and executed, and the delay between the generation and issuance of control strategies shall not exceed 5 seconds.

7. The method according to claim 1, characterized in that, Step S7 also includes an equipment health status assessment step: by analyzing the operating current, vibration, temperature and other operating condition data of the equipment, an equipment energy efficiency degradation and health status assessment model is established to realize early fault warning of the equipment, with a warning accuracy rate of not less than 85%.

8. A multi-system spatiotemporal coordinated intelligent control system for tunnel ventilation and lighting, characterized in that, include: The perception layer includes various types of sensors deployed inside the tunnel to collect real-time tunnel environmental parameters, traffic flow parameters, and equipment operating parameters. Edge computing and control layer: This includes multiple intelligent edge control units deployed in various substations or key nodes within the tunnel. These units are used to preprocess and fuse data collected by the perception layer, perform short-term traffic flow prediction, receive and execute control strategies sent from the cloud, and achieve real-time control of fans and lights. The cloud-based big data analysis and optimization platform includes a data storage module, a model training module, a multi-objective optimization module, and a digital twin simulation module. It is used to store historical data, train and update prediction models and energy consumption models, solve for the optimal coordinated control strategy, and perform safety simulation verification of the control strategy. Visualized integrated management and control platform: including energy panorama dashboard, energy consumption analysis module, equipment monitoring module, energy efficiency benchmarking module and early warning alarm module, used to provide managers with intuitive display of system operation status and decision support.

9. The system according to claim 8, characterized in that, The intelligent edge control unit deeply integrates the reliability of industrial PLC with the computing power of edge server, supports Modbus, OPC UA or MQTT industrial protocol conversion, and has a built-in safety pre-simulation module. It continues to run according to the preset safety policy when the network is interrupted, and the system availability is no less than 99.5%.

10. The system according to claim 8, characterized in that, The digital twin simulation module enables dynamic simulation of energy flow, visual simulation and deduction of control strategies, and pre-evaluation of control effects; the visual integrated management and control platform provides visualization tools such as three-dimensional heat maps of energy consumption, radar charts of equipment energy efficiency benchmarking, and simulation curves of energy-saving strategy effects.

Images