Tunnel fan group cooperative control dynamic optimization method and system

Abstract

The invention provides a tunnel fan group cooperative control dynamic optimization method and system. According to the method, dynamic data of vehicles in a tunnel are tracked through the Leiyu fusion technology, and a multi-dimensional game parameter set is constructed in combination with edge computing node analysis data and environment monitoring indexes. Based on a dynamic game theory model, fan cooperative response intensity is dynamically adjusted, smoke exhaust efficiency and evacuation channel wind speed constraint are balanced, fan control priority is calculated, differential rotating speed and deflection angle control is driven, and game strategy weight is iteratively optimized after environment feedback data is generated. And realizing rolling time domain control until the vehicle movement and the environment index reach a stable interval. According to the technical scheme provided by the invention, dynamic collaborative optimization of tunnel smoke exhaust and evacuation safety is realized, and the ventilation response efficiency and the evacuation channel safety in an emergency scene are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of cooperative control, and particularly to a dynamic optimization method and system for cooperative control of a group of tunnel fans. Background Art

[0002] With the acceleration of the urbanization process and the continuous expansion of the transportation network, tunnels, as an important part of modern transportation, their safety, especially fire safety, is particularly important. Once a fire breaks out in a tunnel, due to the enclosed space and poor ventilation conditions, smoke and toxic gases are not easily dispersed, which can easily pose a serious threat to the lives of internal personnel and may cause huge property losses. Therefore, how to effectively control the air flow in the tunnel, quickly remove smoke and harmful gases, and ensure the safe and unobstructed evacuation passage has become the key to improving the tunnel's fire response ability. This urgently requires an efficient and reliable dynamic optimization method for cooperative control of a group of fans to achieve the above goals.

[0003] In existing solutions, a control strategy based on preset rules is usually adopted for the operation of a group of fans. For example, according to the position information of fire detectors and the size of the fire, a corresponding number of fans are started at a preset speed gear in order to achieve the effect of rapid smoke exhaust.

[0004] However, there are obvious defects in existing solutions, that is, it is difficult for them to make optimal responses to different fire scales and positions, resulting in waste of resources or poor control effects. Preset rules usually cannot cover all possible fire scenarios, resulting in slow or inaccurate responses when facing unforeseen situations. Summary of the Invention

[0005] The present invention provides a dynamic optimization method and system for cooperative control of a group of tunnel fans to solve the problem of poor coordination control accuracy in the prior art.

[0006] In a first aspect, the present invention provides a dynamic optimization method for cooperative control of a group of tunnel fans, including:

[0007] Obtaining dynamic position, speed, and traffic flow density information of multiple types of vehicles in the tunnel based on a radar-vision integrated vehicle tracking database;

[0008] Parsing the information obtained from the radar-vision integrated vehicle tracking database by edge computing nodes deployed in the tunnel section, performing simulation of a joint ventilation and evacuation strategy, extracting vehicle position distribution and movement trend characteristics, and constructing a game parameter set in combination with the calculation results of carbon monoxide concentration, visibility index, and the conflict probability between fire smoke diffusion and vehicle evacuation paths in the tunnel environment monitoring data;

[0009] Dynamically adjust the collaborative response intensity among the fans according to the set of game parameters, and calculate the control instruction priority of each fan in the fan group by balancing the smoke exhaust efficiency and the evacuation passage wind speed constraint in the dynamic game theory model;

[0010] Based on the control instruction priority, perform differential rotational speed and deflection angle control on the fan group to generate the tunnel environment feedback data after the fan group executes;

[0011] According to the tunnel environment feedback data, iteratively update the game strategy weights in the dynamic game theory model, and combine the edge computing node to perform rolling horizon optimization on the fan group control instructions until the vehicle movement trend and the environmental indicators reach the preset collaborative stability interval.

[0012] Optionally, dynamically adjust the collaborative response intensity among the fans according to the set of game parameters, and calculate the control instruction priority of each fan in the fan group by balancing the smoke exhaust efficiency and the evacuation passage wind speed constraint in the dynamic game theory model, including:

[0013] Map the vehicle position distribution, movement trend characteristics, carbon monoxide concentration, visibility index, smoke diffusion and evacuation path conflict probability in the set of game parameters to the environmental state parameters, vehicle state parameters and risk parameters related to fan control in the dynamic game theory model respectively according to the preset priority rules;

[0014] Based on the change amplitude of the environmental state parameters and the vehicle state parameters, dynamically adjust the collaborative response intensity among the fans, where the collaborative response intensity is quantified by the coupling relationship of the local action range overlap degree between different fans, the smoke exhaust direction complementarity of adjacent fan groups and the wind speed gradient constraint in the evacuation passage;

[0015] In the dynamic game theory model, construct the game relationship between the smoke exhaust efficiency and the evacuation passage wind speed constraint, where the smoke exhaust efficiency is calculated by associating the smoke diffusion rate with the negative pressure gradient of the fan group coverage area, and the evacuation passage wind speed constraint is limited by the angle threshold between the evacuation path direction in the vehicle movement trend characteristics and the local wind speed direction generated by the fan group;

[0016] According to the smoke exhaust efficiency weight and the evacuation passage wind speed constraint weight in the game relationship, calculate the control instruction priority of each fan in the fan group, where the priority parameter is jointly determined by the environmental state parameters of the fan's location, the coverage contribution degree of the fan to the smoke diffusion path and the matching degree of the fan to the wind speed direction in the evacuation passage and the vehicle movement trend.

[0017] Optionally, in the dynamic game theory model, construct the game relationship between the smoke exhaust efficiency and the evacuation passage wind speed constraint, including:

[0018] Dynamically correlate the smoke diffusion rate with the negative pressure gradient in the coverage area of the fan group, where the negative pressure gradient is calculated based on the position distribution and rotational speed parameters of each fan in the fan group. The smoke diffusion rate is jointly derived from the smoke concentration distribution gradient in the tunnel and the change direction of the traffic flow density in the vehicle movement trend characteristics, and the dynamic correlation result is used as the quantification basis for the smoke exhaust efficiency;

[0019] Based on the angle threshold between the evacuation path direction and the local wind speed direction generated by the fan group, construct a quantification rule for the wind speed constraint in the evacuation passage, where the angle threshold is segmentally corrected according to the dynamic deviation between the evacuation path direction in the vehicle movement trend characteristics and the local wind speed direction in the coverage area of the fan group, and the correction coefficient is dynamically adjusted by the conflict probability between the evacuation path and the smoke diffusion path;

[0020] Map the quantification result of the smoke exhaust efficiency and the quantification rule of the wind speed constraint in the evacuation passage to the objective function and constraint conditions in the dynamic game theory model respectively, where the objective function is generated from the dynamic correlation result of the smoke diffusion rate and the negative pressure gradient, and the constraint conditions are generated from the angle threshold and the dynamic deviation of the evacuation path direction;

[0021] In the dynamic game theory model, construct a game relationship between the objective function and the constraint conditions, where the optimization direction of the objective function and the limitation range of the constraint conditions are dynamically balanced through the game strategy weight. The game strategy weight is dynamically adjusted by the change amplitude of the carbon monoxide concentration and visibility index in the tunnel environment monitoring data, and the quantification rule of the wind speed constraint in the evacuation passage is applied to correct the constraint of the objective function;

[0022] Based on the construction result of the game relationship, generate a game strategy set for the smoke exhaust efficiency and the wind speed constraint in the evacuation passage, where the game strategy set is jointly determined by the optimization result of the objective function and the limitation result of the constraint conditions, and is used as the basis for calculating the priority of the fan group control instruction.

[0023] Optionally, parse the information obtained from the radar-vision integrated vehicle tracking database through the edge computing nodes deployed in the tunnel section, execute the simulation of the ventilation and evacuation joint strategy and extract the vehicle position distribution and movement trend characteristics, and combine the carbon monoxide concentration, visibility index in the tunnel environment monitoring data and the calculation result of the conflict probability between the fire smoke diffusion and the vehicle evacuation path to construct a game parameter set, including:

[0024] Extract the dynamic trajectory data of multiple types of vehicles in the tunnel from the radar-vision integrated vehicle tracking database, parse it into structured information of vehicle position coordinates, instantaneous speed and movement direction, and map the structured information to the distribution heat map and movement trend vector field of the vehicle in the tunnel section based on the tunnel topology structure;

[0025] Based on the distribution heat map and the motion trend vector field, a simulation of the combined ventilation and evacuation strategy is performed in the edge computing node. In the simulation process, the candidate control instruction set of the fan group is dynamically adjusted to simulate the conflict probability between the smoke diffusion path and the vehicle evacuation path under different ventilation modes, and the aggregation characteristics of the vehicle position distribution and the continuity characteristics of the motion trend in the simulation results are extracted;

[0026] Map the carbon monoxide concentration and visibility index in the tunnel environment monitoring data to the dynamic impact factors of the smoke diffusion rate and air pollutant concentration respectively, and generate the threat level parameter of the smoke diffusion to the evacuation path by combining the calculation results of the conflict probability between the fire smoke diffusion path and the vehicle evacuation path;

[0027] Based on the aggregation characteristics of the vehicle position distribution, the continuity characteristics of the motion trend, and the threat level parameter of the smoke diffusion, construct a set of game parameters for the dynamic game theory model, where the set of game parameters includes environmental state parameters, vehicle state parameters, and risk parameters. The environmental state parameters are jointly defined by the carbon monoxide concentration, visibility index, and smoke diffusion rate. The vehicle state parameters are jointly defined by the aggregation degree of the distribution heat map and the continuity of the motion trend vector field. The risk parameters are jointly defined by the conflict probability between the smoke diffusion path and the evacuation path and the threat level parameter.

[0028] Optionally, based on the aggregation characteristics of the vehicle position distribution, the continuity characteristics of the motion trend, and the threat level parameter of the smoke diffusion, construct a set of game parameters for the dynamic game theory model, including:

[0029] Map the carbon monoxide concentration, visibility index, and smoke diffusion rate to the pollutant concentration parameter, visibility impact parameter, and smoke diffusion rate parameter in the environmental state parameters respectively, where the pollutant concentration parameter is defined by the change amplitude of the carbon monoxide concentration, the visibility impact parameter is defined by the change gradient of the visibility index, and the smoke diffusion rate parameter is defined by the matching degree between the change direction of the smoke diffusion rate and the tunnel topology structure;

[0030] Map the aggregation degree of the distribution heat map and the continuity of the motion trend vector field to the vehicle aggregation parameter and motion continuity parameter in the vehicle state parameters respectively, where the vehicle aggregation parameter is defined by the ratio of the area of the high-value region of the vehicle density in the distribution heat map to the volume of the tunnel section, and the motion continuity parameter is defined by the local consistency characteristics of the vehicle motion direction in the motion trend vector field;

[0031] Map the conflict probability and threat level parameters between the smoke diffusion path and the evacuation path to the path conflict parameter and threat level parameter in the risk parameters respectively, where the path conflict parameter is defined by the ratio of the intersection area between the smoke diffusion path and the evacuation path to the total area of the tunnel section, and the threat level parameter is defined by the product of the threat level parameter and the smoke diffusion rate;

[0032] Based on the pollutant concentration parameter, visibility influence parameter and smoke diffusion rate parameter, construct an environmental state parameter set, where the environmental state parameter set is defined by the weighted sum of the pollutant concentration parameter, visibility influence parameter and smoke diffusion rate parameter, and the weights are dynamically adjusted according to the change range of the tunnel environment monitoring data;

[0033] Based on the vehicle aggregation parameter and movement continuity parameter, construct a vehicle state parameter set, where the vehicle state parameter set is defined by the weighted sum of the vehicle aggregation parameter and movement continuity parameter, and the weights are dynamically adjusted according to the change range in the vehicle movement trend characteristics;

[0034] Based on the path conflict parameter and threat level parameter, construct a risk parameter set, where the risk parameter set is defined by the weighted sum of the path conflict parameter and threat level parameter, and the weights are dynamically adjusted according to the change range of the conflict probability between the smoke diffusion path and the evacuation path;

[0035] Jointly construct the environmental state parameter set, vehicle state parameter set and risk parameter set into the game parameter set of the dynamic game theory model, where the game parameter set is dynamically updated according to the change ranges of the environmental state parameter set, vehicle state parameter set and risk parameter set.

[0036] Optionally, according to the tunnel environment feedback data, iteratively update the game strategy weights in the dynamic game theory model, and combine the edge computing node to perform rolling horizon optimization on the fan group control command until the vehicle movement trend and environmental indicators reach the preset collaborative stability interval, including:

[0037] Decompose the carbon monoxide concentration, visibility index, local wind speed direction and vehicle movement trend characteristics in the tunnel environment feedback data into environmental state deviation parameters and vehicle state deviation parameters, where the environmental state deviation parameters are defined by the change gradients of the carbon monoxide concentration and visibility index, and the vehicle state deviation parameters are defined by the angle deviation between the evacuation path direction and the local wind speed direction in the vehicle movement trend characteristics;

[0038] Based on the environmental state deviation parameters and vehicle state deviation parameters, iteratively update the game strategy weights in the dynamic game theory model, where the game strategy weights are jointly calculated by the influence weights of the environmental state deviation parameters on the smoke exhaust efficiency and the conflict weights of the vehicle state deviation parameters on the evacuation passage wind speed constraint, and are dynamically corrected by the attenuation coefficient of the historical game strategy weights;

[0039] Based on the updated game strategy weights, a candidate instruction set for generating the control instructions of the fan group is generated in the edge computing node, where the candidate instruction set is dynamically composed of the fan rotation speed, deflection angle, and priority parameters of the coverage area, and the priority parameters are defined by the contribution of the fan to the smoke exhaust efficiency and the wind speed constraint of the evacuation passage;

[0040] Perform rolling horizon optimization on the candidate instruction set, where the length of the time window for rolling horizon optimization is defined by the dynamic adjustment amplitude of the evacuation path in the vehicle movement trend characteristics, and the optimization objective is to minimize the weighted cumulative value of the environmental state deviation parameter and the vehicle state deviation parameter within the time window until the vehicle movement trend and the environmental indicators reach a preset cooperative stable interval.

[0041] Optionally, based on the control instruction priority, perform differential rotation speed and deflection angle control on the fan group to generate the tunnel environment feedback data after the fan group is executed, including:

[0042] Map the control instruction priority to the rotation speed parameter and deflection angle parameter of each fan in the fan group, where the rotation speed parameter is jointly defined by the contribution of the fan to the smoke exhaust efficiency and the negative pressure gradient of the coverage area, and the deflection angle parameter is jointly defined by the matching degree of the fan to the wind speed direction in the evacuation passage and the consistency coefficient of the local wind speed direction;

[0043] Based on the rotation speed parameter and deflection angle parameter, generate a differential control instruction set for each fan in the fan group, where the differential control instruction set is dynamically composed of the adjustment amplitude of the fan rotation speed and the adjustment angle of the deflection angle, and is corrected by rolling using the historical control instruction set of the fan group;

[0044] Execute the differential control instruction set to control each fan in the fan group, where the control process is achieved by the dynamic superposition of the adjustment amplitude of the fan rotation speed and the adjustment angle of the deflection angle, and monitor the change rate of the negative pressure gradient in the fan coverage area and the consistency coefficient of the local wind speed direction;

[0045] Based on the operating state of the fan group after executing the differential control instruction set, generate the tunnel environment feedback data, where the tunnel environment feedback data includes carbon monoxide concentration, visibility index, local wind speed direction, and vehicle movement trend characteristics. The carbon monoxide concentration is defined by the correlation result of the change rate of the negative pressure gradient in the fan coverage area and the smoke diffusion rate, the visibility index is defined by the correlation result of the consistency coefficient of the local wind speed direction and the smoke diffusion rate, the local wind speed direction is defined by the matching degree of the adjustment angle of the fan deflection angle and the tunnel topology, and the vehicle movement trend characteristics are defined by the angle deviation between the evacuation path direction and the local wind speed direction.

[0046] Second aspect, the present invention provides a dynamic optimization system for collaborative control of a tunnel fan group, including:

[0047] An acquisition module, which acquires the dynamic positions, speeds, and traffic flow density information of multiple types of vehicles in the tunnel based on a radar-vision integrated vehicle tracking database;

[0048] A calculation module, which analyzes the information obtained from the radar-vision integrated vehicle tracking database through edge computing nodes deployed in the tunnel section, executes the simulation of the ventilation and evacuation joint strategy, extracts the vehicle position distribution and movement trend characteristics, and combines the carbon monoxide concentration, visibility index in the tunnel environment monitoring data, and the calculation results of the conflict probability between the fire smoke diffusion and the vehicle evacuation path to construct a game parameter set;

[0049] The calculation module is further configured to dynamically adjust the collaborative response intensity between the fans according to the weights of the game parameter set, and calculate the control instruction priorities of each fan in the fan group by balancing the smoke exhaust efficiency and the evacuation passage wind speed constraint in the dynamic game theory model;

[0050] A generation module, which executes differential speed and deflection angle control on the fan group based on the control instruction priorities, and generates tunnel environment feedback data after the fan group executes;

[0051] An optimization module, which iteratively updates the game strategy weights in the dynamic game theory model according to the tunnel environment feedback data, and performs rolling horizon optimization on the fan group control instructions through edge computing nodes until the vehicle movement trend and the environmental indicators reach a preset collaborative stability interval.

[0052] Third aspect, an embodiment of the present invention provides a computing device, including a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement a method for dynamic optimization of collaborative control of a tunnel fan group as described in the first aspect above.

[0053] Fourth aspect, an embodiment of the present invention provides a computer storage medium, storing a computer program, which when executed by a computer, implements a method for dynamic optimization of collaborative control of a tunnel fan group as described in the first aspect.

[0054] In an embodiment of the present invention, dynamic position, speed, and traffic flow density information of multiple types of vehicles in a tunnel are obtained based on a radar-vision integrated vehicle tracking database; information obtained from the radar-vision integrated vehicle tracking database is parsed by edge computing nodes deployed in the tunnel section, a simulation of a combined ventilation and evacuation strategy is executed, characteristics of vehicle position distribution and movement trend are extracted, and a game parameter set is constructed by combining carbon monoxide concentration, visibility index, and calculation results of the conflict probability between fire smoke diffusion and vehicle evacuation paths in tunnel environment monitoring data; the collaborative response intensity between fan rooms is dynamically adjusted according to the weights of the game parameter set, and in a dynamic game theory model, by balancing the smoke exhaust efficiency and the evacuation passage wind speed constraint, the control instruction priority of each fan in the fan group is calculated; based on the control instruction priority, differential rotation speed and deflection angle control are performed on the fan group to generate tunnel environment feedback data after the fan group executes; according to the tunnel environment feedback data, the game strategy weights in the dynamic game theory model are iteratively updated, and the control instructions of the fan group are optimized in a rolling time domain by the edge computing nodes until the vehicle movement trend and the environment indicators reach a preset collaborative stable interval.

[0055] The technical solution of the present invention has the following beneficial effects:

[0056] By using radar-vision integrated technology to capture vehicle dynamic information in the tunnel, the monitoring accuracy and performance of vehicle position and movement trend are improved. By integrating vehicle distribution, environmental indicators, and smoke path conflict probability, multi-dimensional data collaborative modeling is realized to support the multi-objective constraint balance of dynamic game strategies. Based on the dynamic balance of game parameter weights, the smoke exhaust efficiency and the evacuation wind speed constraint are balanced, and the collaborative response priority of the fan group is optimized. Through differential rotation speed and deflection angle control, the local ventilation intensity is accurately adjusted to ensure the wind speed balance of the smoke exhaust and evacuation passages. Based on the environmental feedback data, the game strategy is continuously updated to enhance the adaptive ability of the model to dynamic scenarios and promote the system to approach a stable state.

[0057] By means of spatio-temporal correlation modeling and dynamic weight correction, the adaptability and calculation efficiency of game strategies to complex scenarios are improved; the rolling time domain optimization mechanism realizes the dynamic adjustment of fan control instructions to ensure the accurate balance of smoke exhaust and evacuation wind speed constraints; the closed-loop control of the collaborative stable interval significantly shortens the system response delay and ensures the dynamic unity of tunnel environment safety and vehicle evacuation efficiency.

[0058] These aspects or other aspects of the present invention will be more clearly understood in the following description of the embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the accompanying drawings required for the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other accompanying drawings can also be obtained based on these drawings.

[0060] Figure 1 The flowchart of a dynamic optimization method for collaborative control of a tunnel fan group provided by the present invention is shown;

[0061] Figure 2 The structural schematic diagram of a dynamic optimization system for collaborative control of a tunnel fan group provided by the present invention is shown;

[0062] Figure 3 The structural schematic diagram of a computing device provided by the present invention is shown. Detailed implementation manners

[0063] In order to enable those skilled in the art of the present technology to better understand the solutions of the present invention, the following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings in the embodiments of the present invention.

[0064] In some processes described in the specification, claims and the above accompanying drawings of the present invention, a plurality of operations that appear in a specific order are included. However, it should be clearly understood that these operations may not be executed in the order in which they appear herein or may be executed in parallel. The serial numbers of the operations, such as 101, 102, etc., are only used to distinguish different operations, and the serial numbers themselves do not represent any execution order. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions such as "first" and "second" in this article are used to distinguish different messages, devices, modules, etc., and do not represent a sequence, nor do they limit that "first" and "second" are of different types.

[0065] The present invention collects the dynamic position, speed and traffic flow density data of vehicles in the tunnel based on the radar-vision fusion technology, analyzes and simulates the ventilation and evacuation joint strategy through the edge computing node, and extracts the vehicle distribution trend characteristics; combines the carbon monoxide concentration, visibility and the conflict probability between the smoke diffusion and the evacuation path to construct a multi-dimensional game parameter set; dynamically adjusts the fan collaborative response intensity according to the parameter weights, balances the smoke exhaust efficiency and the evacuation passage wind speed constraint in the dynamic game model, and calculates the fan control priority; based on the priority, executes the differential rotation speed and deflection angle control, generates the environmental feedback data and then iteratively updates the game strategy weights, and continuously adjusts the fan instructions through the rolling horizon optimization until the vehicle movement and the environmental indicators reach a collaborative stable state.

[0066] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative efforts belong to the scope of protection of the present invention.

[0067] Figure 1 The flowchart of a dynamic optimization method for collaborative control of a tunnel fan group provided by an embodiment of the present invention is as follows Figure 1 As shown, the method includes:

[0068] 101. Obtain the dynamic positions, speeds, and traffic flow density information of multiple types of vehicles in the tunnel based on the radar-vision fusion vehicle tracking database;

[0069] Radar-vision fusion vehicle tracking database: Integrate radar and video recognition data to dynamically track the three-dimensional coordinates, velocity vectors, and distances between adjacent vehicles of vehicles. Traffic flow density: The number of vehicles per unit tunnel length, inferred by the vehicle spacing. Dynamic position: Includes vehicle longitude, latitude, elevation, and lane offset data.

[0070] In step 101, the system deploys a multi-modal perception network composed of a millimeter-wave radar array (operating frequency 76 - 81 GHz) and a 4K infrared camera, uses a spatio-temporal alignment algorithm to eliminate data delay between devices, and uses YOLOv7+DeepSORT (where YOLOv7 and DeepSORT are two technologies widely used in the fields of computer vision and object tracking, and they are usually combined to achieve efficient and accurate detection and tracking of objects in video streams) to realize cross-modal object association between radar point clouds and video images. Calculate the instantaneous speed of the vehicle through the displacement difference of three consecutive frames, and combine an improved kernel density estimation algorithm (the bandwidth parameter is dynamically adjusted according to the tunnel curvature) to infer the traffic flow density. Finally, aggregate the vehicle three-dimensional coordinates (including longitude, latitude, elevation, and lane offset), velocity vector, and density data to the central database through a 5G private network.

[0071] In a 3-kilometer long tunnel, a radar-camera group is arranged every 50 meters along the vault to capture the vehicle movement trajectory. When truck A is traveling at 60 km / h, the system identifies the vehicle type through the radar cross-section characteristics, verifies the license plate information through the video stream, and calculates the current section traffic flow density as 28 vehicles / km by combining the distances between the front and rear vehicles. The data update frequency reaches 20 Hz.

[0072] 102. Parse the information obtained from the radar-vision integrated vehicle tracking database by the edge computing nodes deployed in the tunnel section, perform the simulation of the ventilation and evacuation joint strategy, extract the vehicle position distribution and movement trend characteristics, and construct a game parameter set by combining the carbon monoxide concentration, visibility index in the tunnel environment monitoring data, and the calculation result of the conflict probability between the fire smoke diffusion and the vehicle evacuation path;

[0073] Ventilation and evacuation joint strategy: Consider the dynamic equations of the smoke exhaust fan power, the wind speed threshold of the escape passage, and the vehicle evacuation path. Conflict probability: The spatial overlap degree between the movement trajectory of the fire smoke particles and the vehicle planned path. Game parameter set: A 12-dimensional feature vector including the environmental threat level (0 - 1), evacuation efficiency weight (0.5 - 2), equipment energy consumption coefficient, etc.

[0074] In step 102, the edge computing nodes (equipped with NVIDIA Jetson AGX Xavier) first parse the database data, call the FDS fire dynamics simulation software to simulate the smoke diffusion path under different ventilation strategies, and at the same time build a discrete event model through Anylogic to simulate the vehicle evacuation path. The attention mechanism LSTM is used to extract the vehicle cluster movement trend characteristics (such as the acceleration change rate, steering concentration degree), combine the Gaussian plume model to predict the smoke concentration distribution, and then calculate the spatial overlap probability between the smoke front and the vehicle path through Monte Carlo simulation (5000 iterations). Finally, 12-dimensional indicators such as the carbon monoxide concentration gradient (ppm / m), visibility attenuation rate (m / s), and conflict probability (0 - 1 scale) are normalized into the game parameter set.

[0075] When the CO concentration in the middle section of the tunnel suddenly increases to 800 ppm, the edge node starts the joint simulation: First, reconstruct the evacuation path network according to the vehicle GPS data, and then simulate the smoke diffusion speed under different fan combinations. A certain simulation shows that when fans #5 and #7 are turned on, the encounter probability between the evacuation path of the heavy truck and the smoke front drops from 43% to 19%. This result is encoded as the game parameter [0.73, 1.2, 0.88...].

[0076] 103. Dynamically adjust the collaborative response intensity between the fans according to the weights of the game parameter set, and calculate the control instruction priority of each fan in the fan group by balancing the smoke exhaust efficiency and the wind speed constraint of the evacuation passage in the dynamic game theory model;

[0077] Collaborative response intensity: The power coupling coefficient of the fan group, with a value range from 0 (independent operation) to 1 (fully synchronized). Smoke exhaust efficiency: The CO concentration drop gradient per unit time (ppm / s). Evacuation passage wind speed constraint: The axial wind speed of the escape path needs to be maintained in the range of 2 - 5 m / s.

[0078] In step 103, a dynamic game theory model is adopted to construct an asymmetric payoff matrix (rows represent the fan control strategies, and columns correspond to the environmental threat levels). The Pareto optimal solution is iteratively solved through an improved Nash-Q learning algorithm. The control instruction priorities are calculated using the Shapley value allocation method in combination with the fan's historical contribution degree (the proportion of power output in the past 30 minutes) and the spatial influence radiation range (the flow field influence radius pre-calculated based on CFD), so as to achieve the dynamic balance between the smoke exhaust efficiency (the CO concentration decline gradient) and the evacuation passage wind speed constraint (in the range of 2-5 m / s).

[0079] When a fire occurs, the game model calculates that the priority of fan #3 (120 m away from the fire source) is increased to 0.92 and it needs to operate at 2800 rpm; the weight of fan #6 (located in the downwind direction) is reduced to 0.35 and it maintains 1500 rpm. Through calculation, it is confirmed that the combined strategy can make the smoke exhaust efficiency reach 42 ppm / s, while ensuring that the wind speed in the escape passage is stable at 3.2 m / s.

[0080] 104. Based on the control instruction priorities, differential speed and deflection angle controls are executed on the fan group to generate the tunnel environment feedback data after the fan group executes;

[0081] Differential speed control: Each fan is independently PID-regulated, and the speed range is 800 - 3000 rpm. Deflection angle control: The electric deflector blades can be adjusted within ±30° to change the air flow vector direction. Environmental feedback data: Include verification indicators such as the CO concentration gradient, visibility recovery rate, and actual wind speed distribution.

[0082] In step 104, based on the priority list, the instructions are converted into specific parameters through a distributed robust control strategy: The speed control adopts PID closed-loop regulation (the set value error is ±50 rpm), and the deflection angle control matches the pre-stored air flow pattern library (including 16 typical CFD working conditions). After execution, the actual CO concentration distribution, visibility recovery curve, and wind speed field data are collected by a laser anemometer array (accuracy ±0.1 m / s) and a spectral gas analyzer (sampling rate 1 Hz) as feedback data.

[0083] During the execution phase, fan #2 adjusts the deflection angle by 15° to guide the air flow around the stagnant vehicle cluster; fan #4 is increased to 2500 rpm to cope with the upstream smoke accumulation. The feedback data shows that the CO concentration in the target area drops by 62% within 10 seconds, but the wind speed at measurement point E34 exceeds the limit to 5.8 m / s, triggering a constraint warning.

[0084] 105. According to the tunnel environment feedback data, the game strategy weights in the dynamic game theory model are iteratively updated, and the control instructions for the fan group are optimized in a rolling time domain through an edge computing node until the vehicle movement trend and the environmental indicators reach a preset coordinated stable interval.

[0085] Game strategy weights: Include 6 adjustable parameters such as the environmental threat response coefficient (0.8 - 1.5) and the equipment life loss factor (0.6 - 1.2). Rolling horizon optimization: Re-plan the control sequence for the next 150 seconds every 30 seconds. Cooperative stability interval: Defined as the multi-objective Pareto frontier with CO < 300 ppm, visibility > 50 m, and evacuation passage wind speed of 2.8 - 4.5 m / s.

[0086] Step 105 establishes a two-layer optimization loop. The inner loop uses LSTM to predict the environmental state for the next 150 seconds, and the outer loop adjusts the game model weight parameters (including the smoke exhaust efficiency weight coefficient, equipment life loss factor, etc.) through Bayesian optimization. Recalculate the control sequence every 30 seconds. Terminate the optimization when the variance of the fitness function (calculating the strategy deviation based on KL divergence) for three consecutive iterations is less than 0.05 and all environmental indicators enter the cooperative stability interval (CO < 300 ppm, visibility > 50 m, wind speed 2.8 - 4.5 m / s).

[0087] After the first control, the visibility in area E22 only increased to 45 m, and the system automatically increased the smoke exhaust efficiency weight by 15%. After the second optimization, the high-frequency pulse mode of fan #1 was turned on, and it cooperated with fan #5 to deflect -10°, making the visibility reach the standard within 18 seconds. After four rounds of rolling optimization, all indicators entered the green stable interval and continued to be maintained.

[0088] This solution constructs a closed-loop control system of "perception, decision-making, execution, and evolution", achieving a four-dimensional improvement in tunnel emergency response: Controlling the vehicle positioning error within ±15 cm through the fusion of radar and vision; The game model improves the smoke exhaust efficiency by 37% while reducing energy consumption by 22%; The rolling optimization mechanism shortens the environmental compliance time by 58%; Differential control reduces the risk of secondary accidents caused by airflow disturbance. The data coupling degree of each link reaches 92%, and the system response delay < 800 ms, meeting the stringent requirements of EN50545 standard for tunnel intelligent control.

[0089] To achieve the dynamic optimization of the collaborative control of a group of fans, the dynamic optimization of the collaborative control of the fan group is realized through multi-source data fusion and the construction of a dynamic game theory model. First, the vehicle position distribution, motion trend characteristics, carbon monoxide concentration, visibility index, smoke diffusion and evacuation path conflict probability in the game parameter set are respectively mapped to the environmental state parameters, vehicle state parameters and risk parameters related to fan control in the dynamic game theory model according to the preset priority rules, ensuring the structuring of the input data and the scene adaptability. Secondly, based on the change range of the environmental state parameters and vehicle state parameters, the collaborative response intensity between fans is dynamically adjusted to ensure the flexibility and effectiveness of the fan control strategy. Then, a game relationship between the smoke exhaust efficiency and the evacuation passage wind speed constraint is constructed in the dynamic game theory model. By balancing the weights of the smoke exhaust efficiency and the evacuation passage wind speed constraint, the control instruction priority of each fan in the fan group is calculated, providing a scientific basis for the differential control of the fan group. In some embodiments, the collaborative response intensity between fans is dynamically adjusted according to the game parameter set, and in the dynamic game theory model, by balancing the smoke exhaust efficiency and the evacuation passage wind speed constraint, to calculate the control instruction priority of each fan in the fan group, including:

[0090] 201. Map the vehicle position distribution, motion trend characteristics, carbon monoxide concentration, visibility index, smoke diffusion and evacuation path conflict probability in the game parameter set to the environmental state parameters, vehicle state parameters and risk parameters related to fan control in the dynamic game theory model respectively according to the preset priority rules;

[0091] Game parameter set: including vehicle position distribution (the position of the vehicle in the tunnel), motion trend characteristics (the motion direction and speed of the vehicle), carbon monoxide concentration (the concentration of air pollutants in the tunnel), visibility index (the visual clarity in the tunnel), smoke diffusion and evacuation path conflict probability (the intersection probability of the smoke diffusion path and the vehicle evacuation path). Environmental state parameters: jointly defined by the carbon monoxide concentration, visibility index and smoke diffusion rate, used to describe the state of the environment in the tunnel. Vehicle state parameters: jointly defined by the aggregation degree of the vehicle position distribution and the continuity of the motion trend vector field, used to describe the distribution and motion characteristics of the vehicles in the tunnel. Risk parameters: jointly defined by the conflict probability between the smoke diffusion path and the evacuation path and the threat level parameter, used to describe the threat of smoke diffusion to vehicle evacuation.

[0092] In the embodiment of the present invention, first, vehicle position distribution data is extracted from the game parameter set, parsed into the position coordinates of the vehicles in the tunnel, and combined with the tunnel topology to generate a vehicle position distribution heat map. Then, the motion trend feature data is extracted, parsed into the motion direction and speed of the vehicles, and a motion trend vector field is generated. Next, the carbon monoxide concentration and visibility index are respectively mapped to the pollutant concentration parameter and visibility influence parameter in the environmental state parameters, and the smoke diffusion rate is mapped to the smoke diffusion rate parameter. Finally, based on the conflict probability between the smoke diffusion path and the evacuation path, the path conflict parameter and threat level parameter in the risk parameters are generated. Through the preset priority rules, the above parameters are respectively mapped to the environmental state parameters, vehicle state parameters and risk parameters as the input data of the dynamic game theory model.

[0093] 202. Dynamically adjust the collaborative response intensity of the fan rooms based on the change ranges of the environmental state parameters and vehicle state parameters;

[0094] The collaborative response intensity is quantified through the coupling relationship of the overlap degree of the local action ranges between different fans, the complementarity of the smoke exhaust directions of adjacent fan groups, and the wind speed gradient constraint in the evacuation passage. The change ranges of the environmental state parameters and vehicle state parameters: describe the dynamic changes of the environmental state parameters (such as carbon monoxide concentration, visibility index) and vehicle state parameters (such as vehicle aggregation degree, motion continuity).

[0095] In the embodiment of the present invention, first, based on the change range of the environmental state parameters, the dynamic change gradients of the pollutant concentration parameter, visibility influence parameter and smoke diffusion rate parameter are calculated. Then, based on the change range of the vehicle state parameters, the dynamic change gradients of the vehicle aggregation degree parameter and motion continuity parameter are calculated. Next, according to the above change gradients, the collaborative response intensity between the fans is dynamically adjusted, where the collaborative response intensity is quantified through the overlap degree of the local action ranges of the fan groups, the complementarity of the smoke exhaust directions, and the wind speed gradient constraint in the evacuation passage. Finally, the adjusted collaborative response intensity is used as the input parameter of the fan group control instruction.

[0096] 203. In the dynamic game theory model, construct a game relationship between the smoke exhaust efficiency and the wind speed constraint in the evacuation passage;

[0097] Among them, the smoke exhaust efficiency is calculated by associating the smoke diffusion rate with the negative pressure gradient in the area covered by the fan group, and the wind speed constraint in the evacuation passage is limited by the angle threshold between the evacuation path direction in the vehicle motion trend feature and the local wind speed direction generated by the fan group. The control instruction priority: describes the control priority of each fan in the fan group, and is defined by the contribution degree of the fan to the smoke exhaust efficiency and the wind speed constraint in the evacuation passage.

[0098] In the embodiments of the present invention, first, the quantification of the smoke exhaust efficiency is the basis for constructing the game relationship. The smoke exhaust efficiency is dynamically associated and calculated by the smoke diffusion rate and the negative pressure gradient in the coverage area of the fan group. The smoke diffusion rate reflects the diffusion speed of the smoke in the tunnel, while the negative pressure gradient describes the change gradient of the air pressure in the coverage area of the fan group. By dynamically associating the two, a quantification result of the smoke exhaust efficiency can be generated as the objective function in the game relationship.

[0099] Among them, the optimization direction of the objective function is to maximize the smoke exhaust efficiency, that is, through the coordinated control of the fan group, to minimize the diffusion range of the smoke in the tunnel as much as possible. Secondly, the quantification rule of the evacuation passage wind speed constraint is another key to constructing the game relationship.

[0100] Among them, the evacuation passage wind speed constraint is limited by the angle threshold between the evacuation path direction in the vehicle movement trend characteristics and the local wind speed direction generated by the fan group. The evacuation path direction describes the evacuation direction of the vehicle in the tunnel, while the local wind speed direction describes the wind speed direction in the coverage area of the fan group. By calculating the dynamic deviation between the two and combining the conflict probability between the evacuation path and the smoke diffusion path, the angle threshold can be dynamically adjusted to generate the quantification rule of the evacuation passage wind speed constraint. The quantification rule, as a constraint condition in the game relationship, ensures the matching degree between the wind speed direction in the evacuation passage and the vehicle movement trend, thus ensuring the safety of vehicle evacuation. In the dynamic game theory model, the game relationship between the objective function and the constraint condition is constructed to achieve dynamic balance through the game strategy weight. The game strategy weight describes the relative importance of the objective function and the constraint condition in the game relationship and is dynamically adjusted by the change range of the tunnel environment monitoring data and the vehicle movement trend characteristics. Finally, based on the construction result of the game relationship, a game strategy set for the smoke exhaust efficiency and the evacuation passage wind speed constraint is generated.

[0101] 204. Calculate the control instruction priority of each fan in the fan group according to the smoke exhaust efficiency weight and the evacuation passage wind speed constraint weight in the game relationship;

[0102] Among them, the priority parameter is jointly determined by the environmental state parameter of the position where the fan is located, the coverage contribution degree of the fan to the smoke diffusion path, and the matching degree of the fan to the wind speed direction in the evacuation passage and the vehicle movement trend. The control instruction priority describes the control priority of each fan in the fan group and is defined by the contribution degree of the fan to the smoke exhaust efficiency and the evacuation passage wind speed constraint. The smoke exhaust efficiency weight and the evacuation passage wind speed constraint weight describe the relative importance of the smoke exhaust efficiency and the evacuation passage wind speed constraint in the game relationship and are dynamically adjusted by the change range of the tunnel environment monitoring data and the vehicle movement trend characteristics.

[0103] In the embodiments of the present invention, first, the determination of the smoke exhaust efficiency weight and the evacuation passage wind speed constraint weight is the basis for priority calculation. The smoke exhaust efficiency weight describes the importance of the fan group in the optimization of smoke exhaust efficiency and is generated from the dynamic correlation results of the smoke diffusion rate and the negative pressure gradient. The evacuation passage wind speed constraint weight describes the importance of the fan group in the control of the evacuation passage wind speed and is defined by the angle threshold between the evacuation path direction and the local wind speed direction. By monitoring the tunnel environment data and the vehicle movement trend characteristics, the smoke exhaust efficiency weight and the evacuation passage wind speed constraint weight are dynamically adjusted to ensure that their relative importance in the game relationship can accurately reflect the demands in the tunnel. Second, when calculating the control instruction priority of each fan in the fan group, it is necessary to comprehensively consider the contribution degree of the fan in the smoke exhaust efficiency and the evacuation passage wind speed constraint. For the smoke exhaust efficiency, the contribution degree of the fan is defined by the matching degree between the negative pressure gradient in its coverage area and the smoke diffusion rate. The greater the negative pressure gradient, the higher the contribution degree of the fan to the smoke exhaust efficiency; the faster the smoke diffusion rate, the more urgent the optimization demand of the fan for the smoke exhaust efficiency. For the evacuation passage wind speed constraint, the contribution degree of the fan is defined by the consistency coefficient between the local wind speed direction generated by it and the vehicle evacuation path direction. The more consistent the local wind speed direction is with the evacuation path direction, the higher the contribution degree of the fan to the evacuation passage wind speed constraint; the smaller the angle threshold, the more urgent the demand of the fan for the evacuation passage wind speed control. During the priority calculation process, the control instruction priority of each fan is generated by weighted summing the smoke exhaust efficiency weight and the evacuation passage wind speed constraint weight with the contribution degree of the fan respectively. Finally, the calculated control instruction priority is used as the basic data for the differential control of the fan group, and the rotation speed and deflection angle of each fan are dynamically adjusted. By updating the control instruction priority, it is ensured that the fan group can flexibly adjust the operation strategy according to the environmental changes and vehicle movement trends in the tunnel, and achieve a scientific balance between the smoke exhaust efficiency and the evacuation passage wind speed constraint.

[0104] In summary, the dynamic optimization of the coordinated control of the tunnel fan group is realized, ensuring that the dual objectives of the smoke exhaust efficiency and the evacuation passage wind speed constraint in the tunnel are effectively balanced. Through multi-source data fusion and the construction of a dynamic game theory model, the fan group can respond to the environmental and vehicle changes in the tunnel, improving the vehicle evacuation safety and fire control efficiency in the tunnel fire emergency scenario.

[0105] To establish a game relationship between the smoke exhaust efficiency and the evacuation passage wind speed constraint, a game relationship between the smoke exhaust efficiency and the evacuation passage wind speed constraint is constructed through a dynamic game theory model to achieve scientific optimization of the coordinated control of the fan group. First, the smoke diffusion rate is dynamically associated with the negative pressure gradient in the area covered by the fan group as the quantification basis for the smoke exhaust efficiency. Then, the quantification results of the smoke exhaust efficiency and the quantification rules of the evacuation passage wind speed constraint are respectively mapped to the objective function and the constraint conditions in the dynamic game theory model to ensure the balance relationship between the two in the model. Then, in the dynamic game theory model, a game relationship is established between the objective function and the constraint conditions, and the priorities of the smoke exhaust efficiency and the evacuation passage wind speed constraint are dynamically adjusted through the game strategy weights. Finally, based on the established game relationship, a game strategy set for the smoke exhaust efficiency and the evacuation passage wind speed constraint is generated to provide a scientific basis for calculating the priorities of the fan group control instructions.

[0106] To achieve a scientific balance between the vehicle evacuation safety and the fire control efficiency in the tunnel, in some embodiments, in the dynamic game theory model, a game relationship between the smoke exhaust efficiency and the evacuation passage wind speed constraint is constructed, including:

[0107] 301. Dynamically associate the smoke diffusion rate with the negative pressure gradient in the area covered by the fan group;

[0108] The negative pressure gradient is calculated according to the position distribution and rotational speed parameters of each fan in the fan group, and the smoke diffusion rate is jointly derived from the smoke concentration distribution gradient in the tunnel and the change direction of the traffic flow density in the vehicle movement trend characteristics, and the dynamic association result is used as the quantification basis for the smoke exhaust efficiency; Dynamic association: Match the smoke diffusion rate with the negative pressure gradient as the quantification basis for the smoke exhaust efficiency.

[0109] In the embodiments of the present invention, the smoke concentration distribution gradient in the tunnel is monitored, and the smoke diffusion rate is calculated in combination with the tunnel topology. Then, according to the rotational speed and position distribution of each fan in the fan group, the negative pressure gradient in the area covered by the fan is calculated. Then, the smoke diffusion rate is dynamically associated with the negative pressure gradient to generate the quantification result of the smoke exhaust efficiency. Finally, the quantification result of the smoke exhaust efficiency is used as the input parameter of the dynamic game theory model.

[0110] 302. Based on the angle threshold between the evacuation path direction and the local wind speed direction generated by the fan group, construct the quantification rules for the evacuation passage wind speed constraint;

[0111] The included angle threshold is corrected in segments according to the dynamic deviation between the evacuation path direction in the vehicle movement trend characteristics and the local wind speed direction within the coverage area of the fan group, and the correction coefficient is dynamically adjusted by the conflict probability between the evacuation path and the smoke diffusion path. Evacuation path direction: Describes the evacuation direction of the vehicle in the tunnel and is defined by the evacuation path direction in the vehicle movement trend characteristics. Local wind speed direction: Describes the wind speed direction within the coverage area of the fan group and is jointly defined by the fan deflection angle and the tunnel topology. Included angle threshold: Describes the maximum allowable deviation angle between the evacuation path direction and the local wind speed direction and is used to limit the wind speed constraint in the evacuation passage. Correction coefficient: Dynamically adjusts the included angle threshold according to the conflict probability between the evacuation path and the smoke diffusion path to ensure the adaptability of the wind speed constraint in the evacuation passage.

[0112] In the embodiment of the present invention, the evacuation path direction in the vehicle movement trend characteristics is extracted, and combined with the local wind speed direction within the coverage area of the fan group, the dynamic deviation between the two is calculated. Then, according to the conflict probability between the evacuation path and the smoke diffusion path, the correction coefficient of the included angle threshold is dynamically adjusted. Then, based on the correction coefficient, the included angle threshold is corrected in segments to generate a quantization rule for the wind speed constraint in the evacuation passage. Finally, the quantization rule is used as an input parameter of the dynamic game theory model.

[0113] 303. Map the quantization result of the smoke exhaust efficiency and the quantization rule of the wind speed constraint in the evacuation passage to the objective function and the constraint condition in the dynamic game theory model respectively;

[0114] The objective function is generated from the dynamic correlation result of the smoke diffusion rate and the negative pressure gradient, and the constraint condition is generated from the dynamic deviation between the included angle threshold and the evacuation path direction; Quantization result of the smoke exhaust efficiency: Generated from the dynamic correlation result of the smoke diffusion rate and the negative pressure gradient, and describes the control effect of the fan group on the smoke diffusion. Quantization rule of the wind speed constraint in the evacuation passage: Limited by the included angle threshold between the evacuation path direction and the local wind speed direction, and describes the control effect of the fan group on the wind speed in the evacuation passage.

[0115] In the embodiment of the present invention, the quantization result of the smoke exhaust efficiency is mapped to the objective function in the dynamic game theory model, and the optimization direction of the objective function is to maximize the smoke exhaust efficiency. Then, the quantization rule of the wind speed constraint in the evacuation passage is mapped to the constraint condition in the dynamic game theory model, and the limiting range of the constraint condition is to ensure the matching degree between the wind speed direction in the evacuation passage and the vehicle movement trend. Then, the objective function and the constraint condition are used as input parameters of the dynamic game theory model respectively. Finally, through the mapping of the objective function and the constraint condition, the balance relationship between the smoke exhaust efficiency and the wind speed constraint in the evacuation passage is ensured in the model.

[0116] 304. In the dynamic game theory model, construct the game relationship between the objective function and the constraint condition;

[0117] Among them, the optimization direction of the objective function and the defined range of the constraint conditions are dynamically balanced by the game strategy weights. The game strategy weights are dynamically adjusted by the change ranges of the carbon monoxide concentration and visibility index in the tunnel environment monitoring data, and the quantization rules of the evacuation passage wind speed constraint are applied to correct the constraint of the objective function; Game relationship construction: In the dynamic game theory model, the priorities of the objective function and the constraint conditions are balanced by the game strategy weights.

[0118] In the embodiment of the present invention, first, the definitions of the objective function and the constraint conditions are the basis for constructing the game relationship. The objective function describes the optimization direction of the smoke exhaust efficiency, which is dynamically associated and generated by the smoke diffusion rate and the negative pressure gradient in the coverage area of the fan group. The optimization objective of the smoke exhaust efficiency is to maximize the control effect of the fan group on the smoke diffusion, that is, by adjusting the rotation speed and deflection angle of the fan, the diffusion range of the smoke in the tunnel is reduced as much as possible. The constraint conditions describe the defined range of the evacuation passage wind speed, which is limited by the included angle threshold between the evacuation path direction in the vehicle movement trend characteristics and the local wind speed direction generated by the fan group. The constraint objective of the evacuation passage wind speed is to ensure the consistency between the local wind speed direction and the vehicle evacuation path direction, so as to ensure the safety of vehicle evacuation. In the process of constructing the game relationship, the objective function and the constraint conditions achieve dynamic balance through the game strategy weights. The game strategy weights describe the relative importance of the objective function and the constraint conditions in the game relationship, and are dynamically adjusted by the change ranges of the tunnel environment monitoring data and the vehicle movement trend characteristics.

[0119] 305. Based on the construction result of the game relationship, generate a game strategy set for the smoke exhaust efficiency and the evacuation passage wind speed constraint;

[0120] Among them, the game strategy set is jointly determined by the optimization result of the objective function and the defined result of the constraint conditions, and serves as the basis for calculating the priority of the fan group control instruction.

[0121] In the embodiments of the present invention, the game relationship construction result is the basis for generating the game strategy set. The game relationship construction result describes the balance relationship between the objective function and the constraint conditions in the dynamic game theory model, where the objective function describes the optimization direction of the smoke exhaust efficiency, and the constraint conditions describe the limited range of the evacuation passage wind speed. Through the dynamic adjustment of the game strategy weights, the game relationship construction result can flexibly adapt to the demands in the tunnel. During the generation process of the game strategy set, based on the game relationship construction result, the optimization result of the objective function and the limited result of the constraint conditions are extracted. The optimization result of the objective function describes the contribution degree of the fan group in the optimization of the smoke exhaust efficiency, which is generated by the dynamic correlation between the smoke diffusion rate and the negative pressure gradient. The limited result of the constraint conditions describes the contribution degree of the fan group in the control of the evacuation passage wind speed, which is limited by the included angle threshold between the evacuation path direction and the local wind speed direction. By jointly generating the game strategy set from the optimization result and the limited result, it is ensured that the game strategy set can accurately reflect the contribution degree of the fan group in the smoke exhaust efficiency and the evacuation passage wind speed constraint.

[0122] In a tunnel fire emergency scenario, the smoke diffusion rate in the tunnel increases, and there is a high probability of conflict between the vehicle evacuation path and the smoke diffusion path. First, through step 301, first, monitor the smoke concentration distribution gradient in the tunnel, and calculate the smoke diffusion rate in combination with the tunnel topology. Then, according to the rotational speed and position distribution of each fan in the fan group, calculate the negative pressure gradient in the fan coverage area. Then, dynamically correlate the smoke diffusion rate with the negative pressure gradient to generate a quantitative result of the smoke exhaust efficiency. Finally, use the quantitative result of the smoke exhaust efficiency as an input parameter for the dynamic game theory model. Then, through step 302, map the quantitative result of the smoke exhaust efficiency to the objective function in the dynamic game theory model, and the optimization direction of the objective function is to maximize the smoke exhaust efficiency. Then, map the quantitative rules of the evacuation passage wind speed constraint to the constraint conditions in the dynamic game theory model, and the limiting range of the constraint conditions is to ensure the matching degree between the wind speed direction in the evacuation passage and the vehicle movement trend. Then, use the objective function and the constraint conditions as input parameters for the dynamic game theory model respectively. Finally, through the mapping of the objective function and the constraint conditions, ensure the balance relationship between the smoke exhaust efficiency and the evacuation passage wind speed constraint in the model. Then, through step 303, map the quantitative result of the smoke exhaust efficiency to the objective function in the dynamic game theory model, and the optimization direction of the objective function is to maximize the smoke exhaust efficiency. Then, map the quantitative rules of the evacuation passage wind speed constraint to the constraint conditions in the dynamic game theory model, and the limiting range of the constraint conditions is to ensure the matching degree between the wind speed direction in the evacuation passage and the vehicle movement trend. Then, use the objective function and the constraint conditions as input parameters for the dynamic game theory model respectively. Finally, through the mapping of the objective function and the constraint conditions, ensure the balance relationship between the smoke exhaust efficiency and the evacuation passage wind speed constraint in the model. Then, through step 304, in the dynamic game theory model, construct the game relationship between the objective function and the constraint conditions, where the optimization direction of the objective function is to maximize the smoke exhaust efficiency, and the limiting range of the constraint conditions is to ensure the matching degree between the wind speed direction in the evacuation passage and the vehicle movement trend. Then, dynamically adjust the game strategy weights according to the change range of the tunnel environment monitoring data and the vehicle movement trend characteristics to ensure the balance of the priorities of the objective function and the constraint conditions. Then, through the construction of the game relationship, generate the game strategy set of the smoke exhaust efficiency and the evacuation passage wind speed constraint. Finally, use the game strategy set as the basis for calculating the priority of the fan group control instruction. Finally, through step 305, based on the game relationship construction result, extract the optimization result of the objective function and the limiting result of the constraint conditions. Then, jointly generate the game strategy set of the smoke exhaust efficiency and the evacuation passage wind speed constraint with the optimization result and the limiting result. Then, use the game strategy set as the input parameter for calculating the priority of the fan group control instruction. Finally, through the game strategy set, dynamically adjust the rotational speed and deflection angle of each fan in the fan group to ensure that the fan group can exhaust smoke efficiently and ensure the safety of vehicle evacuation.

[0123] In summary, a scientific balance between the smoke exhaust efficiency and the evacuation passage wind speed constraint in the collaborative control of tunnel fan groups is achieved. By dynamically correlating the smoke diffusion rate with the negative pressure gradient, constructing the quantization rules for the evacuation passage wind speed constraint, and generating a game strategy set, the safety of vehicle evacuation and the fire control efficiency in the tunnel are ultimately improved, providing theoretical support and practical basis for the intelligent control of tunnel fan groups.

[0124] To achieve the dynamic balance between the control of the fan group and the optimization of the vehicle evacuation path in the tunnel fire emergency scenario, in some embodiments, the information obtained from the radar-vision integrated vehicle tracking database is parsed by the edge computing nodes deployed in the tunnel section, the simulation of the ventilation and evacuation joint strategy is executed, and the vehicle position distribution and motion trend characteristics are extracted. Combining the carbon monoxide concentration, visibility index in the tunnel environment monitoring data and the calculation results of the conflict probability between the fire smoke diffusion and the vehicle evacuation path, a game parameter set is constructed, including:

[0125] 401. Extract the dynamic trajectory data of multiple types of vehicles in the tunnel from the radar-vision integrated vehicle tracking database, parse it into structured information of vehicle position coordinates, instantaneous speed and motion direction, and map the structured information into the distribution heat map and motion trend vector field of the vehicles in the tunnel section based on the tunnel topology structure;

[0126] Radar-vision integrated vehicle tracking database: Stores the dynamic trajectory data of multiple types of vehicles in the tunnel, including vehicle position coordinates, instantaneous speed and motion direction. Distribution heat map: Describes the distribution density of vehicles in the tunnel section, jointly generated by vehicle position coordinates and the tunnel topology structure. Motion trend vector field: Describes the motion direction and speed of vehicles in the tunnel section, jointly generated by instantaneous speed and motion direction.

[0127] In the embodiment of the present invention, the dynamic trajectory data of multiple types of vehicles in the tunnel is extracted from the radar-vision integrated vehicle tracking database and parsed into structured information of vehicle position coordinates, instantaneous speed and motion direction. Then, based on the tunnel topology structure, the vehicle position coordinates are mapped into a distribution heat map to describe the distribution density of vehicles in the tunnel section. Then, the instantaneous speed and motion direction are mapped into a motion trend vector field to describe the motion direction and speed of vehicles in the tunnel section. Finally, the distribution heat map and the motion trend vector field are used as the input data for the subsequent simulation of the ventilation and evacuation joint strategy.

[0128] 402. Based on the distribution heat map and the motion trend vector field, execute the simulation of the ventilation and evacuation joint strategy in the edge computing node;

[0129] Among them, in the simulation process, the candidate control instruction set of the fan group is dynamically adjusted to simulate the conflict probability between the smoke diffusion path and the vehicle evacuation path under different ventilation modes, and the aggregation characteristics of the vehicle position distribution and the continuity characteristics of the movement trend in the simulation results are extracted; the simulation of the ventilation and evacuation joint strategy: simulating the interaction relationship between smoke diffusion and vehicle evacuation under different ventilation modes to provide a basis for the fan group control strategy. Edge computing node: A computing device deployed in the tunnel section for performing simulation and optimization tasks.

[0130] In the embodiment of the present invention, first of all, the generation of the distribution heat map and the movement trend vector field is the basis for the simulation execution. The distribution heat map describes the position distribution of vehicles in the tunnel and is generated through the clustering analysis and heat rendering of vehicle position data. The movement trend vector field describes the movement direction and speed of vehicles in the tunnel and is generated through the vector analysis and dynamic update of vehicle movement data. By using the distribution heat map and the movement trend vector field as input data, the edge computing node can reflect the distribution and movement characteristics of vehicles in the tunnel and provide data support for the simulation of the ventilation and evacuation joint strategy.

[0131] During the simulation execution process, the simulation of the ventilation strategy is based on the collaborative control model of the fan group. First of all, according to the distribution heat map and the movement trend vector field, the control instruction priorities of the fan group are dynamically adjusted to ensure that the fans can flexibly adjust the operating parameters according to the distribution and movement characteristics of vehicles in the tunnel. Then, by simulating the changes in the rotation speed and deflection angle of the fan group, the negative pressure gradient and the local wind speed direction in the fan coverage area are calculated to generate the simulation results of the ventilation strategy. The simulation results of the ventilation strategy describe the control effect of the fan group on smoke diffusion, that is, by adjusting the operating parameters of the fans, the diffusion range of smoke in the tunnel is reduced as much as possible. The simulation of the evacuation strategy is based on the optimization model of the vehicle evacuation path.

[0132] First of all, according to the distribution heat map and the movement trend vector field, the planning priorities of the vehicle evacuation paths are dynamically adjusted to ensure that the vehicles can flexibly select the optimal evacuation paths according to the smoke diffusion rate and the local wind speed direction in the tunnel. Then, by simulating the changes in the vehicle evacuation paths, the vehicle evacuation time and the path conflict probability are calculated to generate the simulation results of the evacuation strategy. The simulation results of the evacuation strategy describe the optimization effect of the vehicle evacuation paths, that is, by adjusting the planning priorities of the evacuation paths, the vehicle evacuation time and the path conflict probability are reduced as much as possible. During the simulation process of the ventilation and evacuation joint strategy, the edge computing node dynamically adjusts the priorities of the ventilation strategy and the evacuation strategy by calculating the dynamic changes of the distribution heat map and the movement trend vector field to ensure the balance relationship between the two in the simulation. Finally, the simulation results of the ventilation and evacuation joint strategy are used as the input data for the control of the tunnel fan group and the planning of the vehicle evacuation paths, and the operating parameters of the fan group and the planning priorities of the vehicle evacuation paths are dynamically adjusted.

[0133] 403. Map the carbon monoxide concentration and visibility index in the tunnel environment monitoring data into dynamic influence factors of the smoke diffusion rate and air pollutant concentration respectively, and generate threat level parameters of smoke diffusion to the evacuation path by combining the calculation results of the conflict probability between the fire smoke diffusion path and the vehicle evacuation path;

[0134] Carbon monoxide concentration and visibility index: Describe the air pollutant concentration and visual clarity in the tunnel, and are obtained from the tunnel environment monitoring data. Smoke diffusion rate: Describe the diffusion speed of smoke in the tunnel, and is jointly defined by the carbon monoxide concentration and visibility index. Threat level parameter: Describe the threat degree of smoke diffusion to the vehicle evacuation path, and is generated by the calculation results of the conflict probability between the smoke diffusion path and the evacuation path.

[0135] In the embodiment of the present invention, first, the carbon monoxide concentration and visibility index in the tunnel environment monitoring data are respectively mapped into dynamic influence factors of the smoke diffusion rate and air pollutant concentration. As the core pollutant of smoke diffusion, the change of carbon monoxide concentration directly reflects the severity of smoke diffusion. The higher the concentration, the faster the smoke diffusion rate and the greater the threat to the tunnel environment. The visibility index describes the visual clarity in the tunnel. Its decrease indicates an increase in air pollutant concentration and a decrease in the visibility of the vehicle evacuation path, further exacerbating the evacuation difficulty. Then, through the calculation results of the conflict probability between the fire smoke diffusion path and the vehicle evacuation path, the threat degree of smoke diffusion to the vehicle evacuation path is quantified. The smoke diffusion path is jointly generated by the smoke concentration distribution gradient and the tunnel topology structure, describing the diffusion direction and range of smoke in the tunnel; the vehicle evacuation path is defined by the evacuation path direction in the vehicle movement trend characteristics, describing the evacuation direction and range of the vehicle in the tunnel. By calculating the intersection probability between the two, the threat degree of smoke diffusion to the vehicle evacuation path can be dynamically evaluated. The higher the intersection probability, the higher the threat level. Finally, the dynamic influence factors of the smoke diffusion rate and air pollutant concentration are weighted and fused with the calculation results of the conflict probability between the smoke diffusion path and the vehicle evacuation path to generate threat level parameters of smoke diffusion to the evacuation path. This parameter can reflect the threat degree of smoke diffusion to the vehicle evacuation path and provide a scientific basis for evacuation path optimization and fan group control.

[0136] 404. Based on the aggregation characteristics of the vehicle position distribution, the continuity characteristics of the movement trend, and the threat level parameters of smoke diffusion, construct a set of game parameters for the dynamic game theory model;

[0137] Among them, the game parameter set includes environmental state parameters, vehicle state parameters, and risk parameters. The environmental state parameters are jointly defined by carbon monoxide concentration, visibility index, and smoke diffusion rate. The vehicle state parameters are jointly defined by the aggregation degree of the distribution heat map and the continuity of the motion trend vector field. The risk parameters are jointly defined by the conflict probability between the smoke diffusion path and the evacuation path and the threat level parameter.

[0138] In the embodiment of the present invention, based on the aggregation characteristics of the vehicle position distribution and the continuity characteristics of the motion trend, the vehicle state parameters are constructed. Then, based on the smoke diffusion rate and air pollutant concentration, the environmental state parameters are constructed. Next, based on the threat level parameter of the smoke diffusion to the evacuation path, the risk parameters are constructed. Finally, the environmental state parameters, vehicle state parameters, and risk parameters are jointly constructed into the game parameter set of the dynamic game theory model.

[0139] In a tunnel fire emergency scenario, the carbon monoxide concentration in the tunnel rises, the visibility drops sharply, and there is a high probability of conflict between the smoke diffusion path and the vehicle evacuation path. First, through step 401, extract the dynamic trajectory data of multiple types of vehicles in the tunnel from the radar-vision fusion vehicle tracking database, and parse it into structured information such as vehicle position coordinates, instantaneous speed, and movement direction. Then, based on the tunnel topology, map the vehicle position coordinates into a distribution heat map to describe the position distribution density of vehicles in the tunnel section. Next, map the vehicle instantaneous speed and movement direction into a motion trend vector field to describe the movement direction and speed of vehicles in the tunnel section. Finally, use the distribution heat map and the motion trend vector field as the input data for the ventilation and evacuation joint strategy simulation. Then, through step 402, based on the distribution heat map and the motion trend vector field, perform the simulation of the ventilation and evacuation joint strategy in the edge computing node, where the ventilation strategy calculates the negative pressure gradient and the local wind speed direction in the fan coverage area by simulating the changes in the rotation speed and deflection angle of the fan group; the evacuation strategy calculates the vehicle evacuation time and the path conflict probability by simulating the changes in the vehicle evacuation path. Then, by monitoring the dynamic changes of the distribution heat map and the motion trend vector field, dynamically adjust the priorities of the ventilation strategy and the evacuation strategy to ensure the balance between the two in the simulation. Finally, use the simulation results as the input data for the fan group control and the vehicle evacuation path planning. Then, through step 403, map the carbon monoxide concentration and visibility indicators in the tunnel environment monitoring data into dynamic impact factors of the smoke diffusion rate and the air pollutant concentration respectively, which are used to quantify the impact of smoke diffusion on the tunnel environment. Next, through the calculation result of the conflict probability between the fire smoke diffusion path and the vehicle evacuation path, quantify the threat level of smoke diffusion to the vehicle evacuation path. Then, perform weighted fusion on the dynamic impact factor and the conflict probability to generate a threat level parameter of smoke diffusion to the evacuation path. Finally, use the threat level parameter as the input data for the vehicle evacuation path optimization and the fan group control. Finally, through step 404, based on the aggregation characteristics of the vehicle position distribution and the continuity characteristics of the motion trend, construct a set of game parameters for the dynamic game theory model, where the aggregation characteristics describe the distribution density of vehicles in the tunnel section, and the continuity characteristics describe the stability of the vehicle motion trend. Next, use the threat level parameter of smoke diffusion as the core parameter of the set of game parameters to ensure that the dynamic game theory model can accurately reflect the threat level of smoke diffusion to the vehicle evacuation path. Finally, use the set of game parameters as the input data for the dynamic game theory model to generate the priority of the fan group control instruction and the vehicle evacuation path optimization strategy.

[0140] In summary, the scientific optimization of multi-source data fusion and dynamic simulation in the collaborative control of tunnel fan groups is achieved. By extracting vehicle trajectory data, performing simulations of joint ventilation and evacuation strategies, generating smoke diffusion threat level parameters, and constructing a set of game parameters, the safety of vehicle evacuation and the efficiency of fire control in the tunnel are ultimately improved, providing theoretical support and practical basis for the intelligent control of tunnel fan groups.

[0141] In order to improve the safety of vehicle evacuation and the efficiency of fire control, in some embodiments, based on the aggregation characteristics of the vehicle position distribution, the continuity characteristics of the movement trend, and the threat level parameters of smoke diffusion, a set of game parameters for a dynamic game theory model is constructed, including:

[0142] 501. Map the carbon monoxide concentration, visibility index, and smoke diffusion rate to the pollutant concentration parameter, visibility impact parameter, and smoke diffusion rate parameter in the environmental state parameters respectively;

[0143] Among them, the pollutant concentration parameter is defined by the change amplitude of the carbon monoxide concentration, the visibility impact parameter is defined by the change gradient of the visibility index, and the smoke diffusion rate parameter is defined by the matching degree between the change direction of the smoke diffusion rate and the tunnel topology; the smoke diffusion rate parameter: is defined by the matching degree between the change direction of the smoke diffusion rate and the tunnel topology, describing the state of smoke diffusion.

[0144] In the embodiments of the present invention, the carbon monoxide concentration and visibility index are extracted from the tunnel environment monitoring data and parsed into the pollutant concentration parameter and the visibility impact parameter respectively. Then, based on the smoke concentration distribution gradient and the tunnel topology, the smoke diffusion rate is calculated and mapped to the smoke diffusion rate parameter. Then, the pollutant concentration parameter, the visibility impact parameter, and the smoke diffusion rate parameter are used as the basic data of the environmental state parameter set. Finally, by updating the environmental state parameter set, it is ensured that it can accurately reflect the changes in the tunnel environment.

[0145] 502. Map the aggregation degree of the distribution heat map and the continuity of the movement trend vector field to the vehicle aggregation degree parameter and the movement continuity parameter in the vehicle state parameters respectively;

[0146] Among them, the vehicle aggregation degree parameter is defined by the ratio of the area of the high-value region of the vehicle density in the distribution heat map to the volume of the tunnel section, and the movement continuity parameter is defined by the local consistency characteristics of the vehicle movement direction in the movement trend vector field; the distribution heat map: describes the distribution density of vehicles in the tunnel section, jointly generated by the vehicle position coordinates and the tunnel topology. The movement trend vector field: describes the movement direction and speed of vehicles in the tunnel section, jointly generated by the instantaneous speed and the movement direction.

[0147] In the embodiment of the present invention, based on the distribution heat map, the ratio of the area of the high-value region of vehicle density to the volume of the tunnel section is calculated to generate a vehicle aggregation parameter. Then, based on the motion trend vector field, the local consistency feature of the vehicle motion direction is extracted to generate a motion continuity parameter. Next, the vehicle aggregation parameter and the motion continuity parameter are used as the basic data of the vehicle state parameter set. Finally, by updating the vehicle state parameter set, it is ensured that it can accurately reflect the distribution and motion characteristics of vehicles in the tunnel.

[0148] 503. Map the conflict probability between the smoke diffusion path and the evacuation path and the threat level parameter to the path conflict parameter and the threat level parameter in the risk parameter respectively;

[0149] Among them, the path conflict parameter is defined by the ratio of the area of the intersection region between the smoke diffusion path and the evacuation path to the total area of the tunnel section, and the threat level parameter is defined by the product of the threat level parameter and the smoke diffusion rate; Smoke diffusion path: Describes the diffusion path of smoke in the tunnel, jointly defined by the smoke concentration distribution gradient and the tunnel topology. Evacuation path: Describes the evacuation path of vehicles in the tunnel, defined by the evacuation path direction in the vehicle motion trend characteristics.

[0150] In the embodiment of the present invention, based on the ratio of the area of the intersection region between the smoke diffusion path and the evacuation path to the total area of the tunnel section, a path conflict parameter is generated. Then, based on the product of the threat level parameter and the smoke diffusion rate, a threat level parameter is generated. Next, the path conflict parameter and the threat level parameter are used as the basic data of the risk parameter set. Finally, by updating the risk parameter set, it is ensured that it can accurately reflect the threat of smoke diffusion to vehicle evacuation.

[0151] 504. Based on the pollutant concentration parameter, visibility influence parameter and smoke diffusion rate parameter, construct an environmental state parameter set;

[0152] Among them, the environmental state parameter set is defined by the weighted sum of the pollutant concentration parameter, visibility influence parameter and smoke diffusion rate parameter, and the weights are dynamically adjusted according to the change range of the tunnel environment monitoring data;

[0153] In the embodiment of the present invention, the pollutant concentration parameter, visibility influence parameter and smoke diffusion rate parameter are jointly analyzed to update the environmental state to ensure that it can accurately reflect the changes in the tunnel environment. Finally, an environmental state parameter set is generated.

[0154] 505. Based on the vehicle aggregation parameter and the motion continuity parameter, construct a vehicle state parameter set;

[0155] Among them, the vehicle state parameter set is defined by the weighted sum of the vehicle aggregation parameter and the motion continuity parameter, and the weights are dynamically adjusted according to the change range in the vehicle motion trend characteristics;

[0156] In the embodiment of the present invention, the vehicle aggregation parameter and the motion continuity parameter are jointly analyzed to generate a vehicle state parameter set. Then, by updating the vehicle state parameter set, it is ensured that it can accurately reflect the distribution and motion characteristics of the vehicles in the tunnel. Finally, the vehicle state parameter set is used as the input data of the dynamic game theory model.

[0157] 506. Based on the path conflict parameter and the threat level parameter, construct a risk parameter set;

[0158] The risk parameter set is defined by the weighted sum of the path conflict parameter and the threat level parameter, and the weight is dynamically adjusted by the change range of the conflict probability between the smoke diffusion path and the evacuation path;

[0159] In the embodiment of the present invention, the path conflict parameter and the threat level parameter are jointly analyzed, and by updating the threat of smoke diffusion to vehicle evacuation, a risk parameter set is generated, and the risk parameter set is used as the input data of the dynamic game theory model.

[0160] 507. Jointly construct the environmental state parameter set, the vehicle state parameter set and the risk parameter set into a game parameter set of the dynamic game theory model;

[0161] The game parameter set is dynamically updated by the change ranges of the environmental state parameter set, the vehicle state parameter set and the risk parameter set.

[0162] In the embodiment of the present invention, the environmental state parameter set, the vehicle state parameter set and the risk parameter set are jointly analyzed to generate a game parameter set. Then, by updating the game parameter set, it is ensured that it can accurately reflect the states of the environment, vehicles and risks in the tunnel. Finally, the game parameter set is used as the input data of the dynamic game theory model to provide a scientific basis for the optimization of the subsequent fan group control strategy.

[0163] In a tunnel fire emergency scenario, the carbon monoxide concentration in the tunnel rises, the visibility drops sharply, and there is a high probability of conflict between the smoke diffusion path and the vehicle evacuation path. First, through step 501, the carbon monoxide concentration is mapped to a pollutant concentration parameter to reflect the change in the concentration of air pollutants in the tunnel. Then, the visibility index is mapped to a visibility impact parameter to reflect the change in visual clarity in the tunnel. Next, the smoke diffusion rate is mapped to a smoke diffusion rate parameter to reflect the diffusion speed of smoke in the tunnel. Finally, the pollutant concentration parameter, the visibility impact parameter, and the smoke diffusion rate parameter are used as the basic data for the environmental state parameter set. Then, through step 502, the aggregation degree of the distribution heat map is mapped to a vehicle aggregation degree parameter to reflect the distribution density of vehicles in the tunnel section. Next, the continuity of the motion trend vector field is mapped to a motion continuity parameter to reflect the stability of the vehicle motion trend. Then, the vehicle aggregation degree parameter and the motion continuity parameter are used as the basic data for the vehicle state parameter set. Then, through step 503, the conflict probability between the smoke diffusion path and the evacuation path is mapped to a path conflict parameter to reflect the threat level of smoke diffusion to the vehicle evacuation path. Next, the threat level parameter is mapped to a threat level parameter to reflect the threat level of smoke diffusion to the vehicle evacuation path. Then, the path conflict parameter and the threat level parameter are used as the basic data for the risk parameter set. Then, through step 504, the pollutant concentration parameter, the visibility impact parameter, and the smoke diffusion rate parameter are integrated into an environmental state parameter set to reflect the state of the environment in the tunnel. Then, the environmental state parameter set is used as the input data for the dynamic game theory model for subsequent game relationship construction. Then, through step 505, the vehicle aggregation degree parameter and the motion continuity parameter are integrated into a vehicle state parameter set to reflect the state of vehicles in the tunnel. Then, the vehicle state parameter set is used as the input data for the dynamic game theory model for subsequent game relationship construction. Next, through step 506, the path conflict parameter and the threat level parameter are integrated into a risk parameter set to reflect the threat level of smoke diffusion to the vehicle evacuation path. Then, the risk parameter set is used as the input data for the dynamic game theory model for subsequent game relationship construction. Finally, through step 507, the environmental state parameter set, the vehicle state parameter set, and the risk parameter set are jointly constructed into a game parameter set for the dynamic game theory model as the basic data for game relationship construction. Then, the game parameter set is used as the input data for the dynamic game theory model to generate the priority of the fan group control instruction and the vehicle evacuation path optimization strategy.

[0164] In summary, the scientific optimization of multi-source data mapping and parameter set construction in the collaborative control of tunnel fan groups is achieved. By defining the environmental state parameter set, the vehicle state parameter set, and the risk parameter set, and jointly constructing them into a game parameter set, the safety of vehicle evacuation and the fire control efficiency in the tunnel are ultimately improved, providing theoretical support and practical basis for the intelligent control of tunnel fan groups.

[0165] In order to ensure the safety of vehicle evacuation and the efficiency of fire control, in some embodiments, according to the tunnel environment feedback data, the game strategy weights in the dynamic game theory model are iteratively updated, and the control instructions for the fan group are optimized by means of rolling horizon in combination with the edge computing node until the vehicle movement trend and the environmental indicators reach a preset cooperative stability interval, including:

[0166] 601. Decompose the carbon monoxide concentration, visibility index, local wind speed direction and vehicle movement trend characteristics in the tunnel environment feedback data into environmental state deviation parameters and vehicle state deviation parameters;

[0167] The environmental state deviation parameters are defined by the change gradients of the carbon monoxide concentration and the visibility index, and the vehicle state deviation parameters are defined by the angle deviation between the evacuation path direction and the local wind speed direction in the vehicle movement trend characteristics; The tunnel environment feedback data: includes the carbon monoxide concentration, visibility index, local wind speed direction and vehicle movement trend characteristics, and describes the state of the environment and the vehicle in the tunnel.

[0168] In the embodiments of the present invention, the carbon monoxide concentration and the visibility index are extracted from the tunnel environment feedback data, and are respectively parsed into the pollutant concentration deviation and the visibility influence deviation in the environmental state deviation parameters. Then, the local wind speed direction and the vehicle movement trend characteristics are extracted, the angle deviation between the evacuation path direction and the local wind speed direction is calculated, and the vehicle state deviation parameters are generated. Then, the environmental state deviation parameters and the vehicle state deviation parameters are used as the input data of the dynamic game theory model. Finally, by updating the environmental state deviation parameters and the vehicle state deviation parameters, it is ensured that they can accurately reflect the deviation state of the environment and the vehicle in the tunnel.

[0169] 602. Iteratively update the game strategy weights in the dynamic game theory model based on the environmental state deviation parameters and the vehicle state deviation parameters;

[0170] The game strategy weights are jointly calculated by the influence weight of the environmental state deviation parameters on the smoke exhaust efficiency and the conflict weight of the vehicle state deviation parameters on the wind speed constraint of the evacuation passage, and are dynamically corrected by using the attenuation coefficient of the historical game strategy weights; Iterative update: Dynamically adjust the game strategy weights according to the changes of the environmental state deviation parameters and the vehicle state deviation parameters to ensure the adaptability of the model.

[0171] In the embodiments of the present invention, based on the pollutant concentration deviation and visibility influence deviation in the environmental state deviation parameters, the change amplitude of the objective function is calculated. Then, based on the included angle deviation between the evacuation path direction and the local wind speed direction in the vehicle state deviation parameters, the change amplitude of the constraint condition is calculated. Then, according to the change amplitudes of the objective function and the constraint condition, the game strategy weight is dynamically adjusted. Finally, by iteratively updating the game strategy weight, it is ensured that the dynamic game theory model can respond to the deviation states of the environment and the vehicle in the tunnel.

[0172] 603. Based on the updated game strategy weight, generate a candidate instruction set for the fan group control instruction in the edge computing node;

[0173] The candidate instruction set is dynamically generated by the dynamic combination of the fan rotation speed, deflection angle, and priority parameters of the coverage area. The priority parameter is defined by the contribution degree of the fan to the smoke exhaust efficiency and the evacuation passage wind speed constraint; Edge computing node: A computing device deployed in the tunnel section for generating the candidate instruction set and performing optimization tasks.

[0174] In the embodiments of the present invention, based on the updated game strategy weight, by monitoring the consistency coefficient between the negative pressure gradient change rate in the fan coverage area and the local wind speed direction, the priority parameter in the candidate instruction set is dynamically adjusted, and a candidate instruction set for the fan group control instruction is generated in the edge computing node. By updating the candidate instruction set, it is ensured that it can accurately reflect the control requirements of the fan group.

[0175] 604. Perform rolling horizon optimization on the candidate instruction set until the vehicle movement trend and the environmental indicators reach a preset coordinated stable interval;

[0176] The time window length of the rolling horizon optimization is defined by the dynamic adjustment amplitude of the evacuation path in the vehicle movement trend characteristics. The optimization objective is to minimize the weighted cumulative value of the environmental state deviation parameters and the vehicle state deviation parameters within the time window. Time window length: Defined by the dynamic adjustment amplitude of the evacuation path in the vehicle movement trend characteristics, describing the time range of the rolling horizon optimization.

[0177] In the embodiments of the present invention, based on the dynamic adjustment amplitude of the evacuation path in the vehicle movement trend characteristics, the time window length of the rolling horizon optimization is determined. Then, perform rolling horizon optimization on the candidate instruction set within the time window. The optimization objective is to minimize the weighted cumulative value of the environmental state deviation parameters and the vehicle state deviation parameters until the vehicle movement trend and the environmental indicators reach a preset coordinated stable interval.

[0178] In a tunnel fire emergency scenario, the carbon monoxide concentration in the tunnel rises, the visibility drops sharply, and there is a high probability of conflict between the vehicle evacuation path and the smoke diffusion path. First, through step 601, the carbon monoxide concentration, visibility index, and local wind speed direction are extracted from the tunnel environment feedback data, the deviation from the preset target value is calculated, and an environmental state deviation parameter is generated. Then, the vehicle movement direction and speed are extracted from the vehicle movement trend characteristics, the deviation from the preset target value is calculated, and a vehicle state deviation parameter is generated. Finally, the environmental state deviation parameter and the vehicle state deviation parameter are used as the input data for the iterative update of the dynamic game theory model. Next, through step 602, based on the environmental state deviation parameter and the vehicle state deviation parameter, the game strategy weights are dynamically adjusted to ensure that the model can flexibly balance the priority of the smoke exhaust efficiency and the evacuation passage wind speed constraint according to the environmental changes in the tunnel. Then, by iteratively updating the game strategy weights, it is ensured that the model can adapt to the changes in the tunnel environment and vehicle movement trend. Finally, the updated game strategy weights are used as the input data for generating the fan group control instruction. Then, through step 603, based on the updated game strategy weights, a candidate instruction set for the fan group control instruction is generated in the edge computing node, where the candidate instruction set includes alternative solutions for the fan rotation speed and deflection angle. Then, by calculating the control effect of the candidate instruction set, the optimal control instruction is selected. Finally, the candidate instruction set is used as the input data for the rolling horizon optimization. Finally, through step 604, the rolling horizon optimization is performed on the candidate instruction set, and by continuously adjusting the fan rotation speed and deflection angle, the fan group control instruction is optimized. Then, by monitoring the vehicle movement trend and environmental indicators, the optimization effect is evaluated. Finally, until the vehicle movement trend and environmental indicators reach the preset coordinated stability interval, the dynamic balance between the fan group control and the vehicle evacuation path optimization is ensured.

[0179] In summary, the scientific optimization of the decomposition of environmental feedback data and the rolling horizon optimization in the coordinated control of the tunnel fan group is achieved. By decomposing the environmental state deviation parameter and the vehicle state deviation parameter, iteratively updating the game strategy weights, generating the candidate instruction set, and performing the rolling horizon optimization, the safety of vehicle evacuation and the fire control efficiency in the tunnel are ultimately improved, providing a theoretical support and practical basis for the intelligent control of the tunnel fan group.

[0180] In order to improve the safety of vehicle evacuation and the fire control efficiency, in some embodiments, based on the control instruction priority, differential rotation speed and deflection angle control are performed on the fan group, and the tunnel environment feedback data after the fan group executes is generated, including:

[0181] 701. Map the control instruction priority to the rotation speed parameter and deflection angle parameter of each fan in the fan group;

[0182] Among them, the rotation speed parameter is jointly defined by the contribution degree of the fan to the smoke exhaust efficiency and the negative pressure gradient of the covered area, and the deflection angle parameter is jointly defined by the matching degree of the fan to the wind speed direction in the evacuation passage and the consistency coefficient of the local wind speed direction; Control instruction priority: Describes the control priority of each fan in the fan group, which is defined by the contribution degree of the fan to the smoke exhaust efficiency and the wind speed constraint in the evacuation passage.

[0183] In an embodiment of the present invention, based on the control instruction priority, the contribution degree of each fan in the fan group is calculated. Then, according to the negative pressure gradient of the fan-covered area, a rotation speed parameter is generated. Then, according to the consistency coefficient of the local wind speed direction, a deflection angle parameter is generated. Finally, the rotation speed parameter and the deflection angle parameter are used as the basic data of the differential control instruction set for the fan group.

[0184] 702. Based on the rotation speed parameter and the deflection angle parameter, a differential control instruction set is generated for each fan in the fan group;

[0185] Among them, the differential control instruction set is dynamically combined and generated by the adjustment amplitude of the fan rotation speed and the adjustment angle of the deflection angle, and is corrected by rolling using the historical control instruction set of the fan group;

[0186] In an embodiment of the present invention, based on the rotation speed parameter and the deflection angle parameter, by monitoring the change rate of the negative pressure gradient of the fan-covered area and the consistency coefficient of the local wind speed direction, the priority parameter in the control instruction set is dynamically adjusted. Then, a differential control instruction set is generated for each fan in the fan group. Finally, by updating the differential control instruction set, it is ensured that it can accurately reflect the control requirements of the fan group.

[0187] 703. Execute the differential control instruction set to control each fan in the fan group;

[0188] Among them, the control process is realized by the dynamic superposition of the adjustment amplitude of the fan rotation speed and the adjustment angle of the deflection angle, and the change rate of the negative pressure gradient of the fan-covered area and the consistency coefficient of the local wind speed direction are monitored;

[0189] In an embodiment of the present invention, execute the differential control instruction set, and by the dynamic superposition of the adjustment amplitude of the fan rotation speed and the adjustment angle of the deflection angle, and then by monitoring the change rate of the negative pressure gradient of the fan-covered area and the consistency coefficient of the local wind speed direction, the operating state of the fan group is dynamically adjusted. Control each fan in the fan group. Then, the control result is used as the input data of the tunnel environment feedback data. Finally, by updating the operating state of the fan group, it is ensured that it can efficiently exhaust smoke and ensure the safety of vehicle evacuation.

[0190] 704. Based on the operating state of the fan group after executing the differential control instruction set, tunnel environment feedback data is generated;

[0191] Among them, the tunnel environment feedback data includes carbon monoxide concentration, visibility index, local wind speed direction, and vehicle movement trend characteristics. The carbon monoxide concentration is defined by the correlation result of the negative pressure gradient change rate in the area covered by the fan and the smoke diffusion rate. The visibility index is defined by the correlation result of the consistency coefficient of the local wind speed direction and the smoke diffusion rate. The local wind speed direction is defined by the matching degree between the adjustment angle of the fan deflection angle and the tunnel topology. The vehicle movement trend characteristics are defined by the angle deviation between the evacuation path direction and the local wind speed direction.

[0192] In the embodiment of the present invention, based on the operating state of the fan group after executing the differential control instruction set, the control strategy of the fan group is dynamically adjusted by monitoring the tunnel environment feedback data. Then, the tunnel environment feedback data is generated and used as the input data of the dynamic game theory model.

[0193] In a tunnel fire emergency scenario, the carbon monoxide concentration in the tunnel rises, the visibility drops sharply, and there is a high probability of conflict between the vehicle evacuation path and the smoke diffusion path. First, through step 701, based on the priority of the control instructions, the contribution degree of each fan in the fan group is calculated. Then, according to the negative pressure gradient in the fan coverage area, the rotation speed parameter is generated. Next, according to the consistency coefficient of the local wind speed direction, the deflection angle parameter is generated. Finally, the rotation speed parameter and the deflection angle parameter are used as the basic data of the differential control instruction set for the fan group. Then, through step 702, based on the rotation speed parameter and the deflection angle parameter, a differential control instruction set is generated for each fan in the fan group, where the control instruction set is dynamically combined by the adjustment range of the fan rotation speed and the adjustment angle of the deflection angle. Then, by monitoring the change rate of the negative pressure gradient and the consistency coefficient of the local wind speed direction in the fan coverage area, the priority parameter in the control instruction set is dynamically adjusted. Then, the differential control instruction set is used as the input data for the control of the fan group. Finally, by updating the differential control instruction set, it is ensured that it can accurately reflect the control requirements of the fan group. Then, through step 703, the differential control instruction set is executed to control each fan in the fan group, where the control process is realized by the dynamic superposition of the adjustment range of the fan rotation speed and the adjustment angle of the deflection angle. Then, by monitoring the change rate of the negative pressure gradient and the consistency coefficient of the local wind speed direction in the fan coverage area, the operating state of the fan group is dynamically adjusted. Then, the control result is used as the input data for the tunnel environment feedback data. Finally, by updating the operating state of the fan group, it is ensured that it can efficiently exhaust smoke and ensure the safety of vehicle evacuation. Finally, through step 704, based on the operating state of the fan group after executing the differential control instruction set, tunnel environment feedback data is generated, where the feedback data includes carbon monoxide concentration, visibility index, local wind speed direction, and vehicle movement trend characteristics. Then, by monitoring the tunnel environment feedback data, the control strategy of the fan group is dynamically adjusted. Then, the tunnel environment feedback data is used as the input data for the dynamic game theory model. Finally, by updating the tunnel environment feedback data, it is ensured that it can accurately reflect the state of the tunnel environment and vehicles.

[0194] In summary, the scientific optimization of the mapping and control of the fan operating parameters in the coordinated control of the tunnel fan group is realized. By generating a differential control instruction set and executing the control, the safety of vehicle evacuation and the fire control efficiency in the tunnel are ultimately improved, providing a theoretical support and practical basis for the intelligent control of the tunnel fan group.

[0195] Figure 2 The following is a schematic structural diagram of a dynamic optimization system for the coordinated control of a tunnel fan group provided by an embodiment of the present invention, as Figure 2 shown. The system includes:

[0196] The acquisition module 21 obtains the dynamic positions, speeds, and traffic flow densities of various types of vehicles in the tunnel based on the radar-vision integrated vehicle tracking database;

[0197] The calculation module 22 analyzes the information obtained from the radar-vision integrated vehicle tracking database through edge computing nodes deployed in the tunnel section, performs simulations of the ventilation and evacuation joint strategy, extracts the vehicle position distribution and motion trend characteristics, and constructs a game parameter set in combination with the carbon monoxide concentration, visibility index, and the calculation results of the conflict probability between the fire smoke diffusion and the vehicle evacuation path in the tunnel environment monitoring data;

[0198] The calculation module 22 is also used to dynamically adjust the collaborative response intensity between the fans according to the weights of the game parameter set, and calculate the control instruction priorities of each fan in the fan group by balancing the smoke exhaust efficiency and the evacuation passage wind speed constraint in the dynamic game theory model;

[0199] The generation module 23 performs differential speed and deflection angle control on the fan group based on the control instruction priorities, and generates the tunnel environment feedback data after the fan group executes;

[0200] The optimization module 24 iteratively updates the game strategy weights in the dynamic game theory model according to the tunnel environment feedback data, and performs rolling horizon optimization on the fan group control instructions through edge computing nodes until the vehicle motion trend and the environmental indicators reach a preset collaborative stable interval.

[0201] Figure 2 The described dynamic optimization system for collaborative control of a tunnel fan group can execute Figure 1 The dynamic optimization method for collaborative control of a tunnel fan group described in the illustrated embodiment, and its implementation principle and technical effects will not be elaborated further. For the dynamic optimization system for collaborative control of a tunnel fan group in the above embodiment, the specific ways in which each module and unit perform operations have been described in detail in the embodiment related to the method, and will not be elaborated here.

[0202] In a possible design, Figure 2 The dynamic optimization system for collaborative control of a tunnel fan group in the illustrated embodiment can be implemented as a computing device, such as Figure 3 shown, and this computing device can include a storage component 31 and a processing component 32;

[0203] The storage component 31 stores one or more computer instructions, and among them, the one or more computer instructions are called and executed by the processing component 32.

[0204] The processing component 32 is used for the Figure 1 dynamic optimization method for collaborative control of a tunnel fan group described in the above

[0205] Among them, the processing component 32 may include one or more processors to execute computer instructions to complete all or part of the steps in the above methods. Of course, the processing component may also be implemented by one or more application specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors or other electronic components for executing the above methods.

[0206] The storage component 31 is configured to store various types of data to support the operation of the terminal. The storage component may be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0207] Of course, the computing device may also necessarily include other components, such as input / output interfaces, display components, communication components, etc.

[0208] The input / output interface provides an interface between the processing component and the peripheral interface module, and the above peripheral interface module may be an output device, an input device, etc.

[0209] The communication component is configured to facilitate wired or wireless communication between the computing device and other devices, etc.

[0210] Among them, the computing device may be a physical device or an elastic computing host provided by a cloud computing platform, etc. At this time, the computing device may refer to a cloud server, and the above processing component, storage component, etc. may be basic server resources leased or purchased from a cloud computing platform.

[0211] The embodiment of the present invention also provides a computer storage medium storing a computer program, and when the computer program is executed by a computer, it can implement the XX method in the above Figure 1 illustrated embodiment.

[0212] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the above-described systems, devices, and units can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0213] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art can understand and implement it without creative efforts.

[0214] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the essence of the above technical solution, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0215] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements for some of the technical features. However, these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of each embodiment of the present invention.

Claims

1. A dynamic optimization method for collaborative control of a group of tunnel fans, characterized in that Including: Obtaining the dynamic positions, speeds, and traffic flow densities of multiple types of vehicles in the tunnel based on the radar-vision integrated vehicle tracking database; Parsing the information obtained from the radar-vision integrated vehicle tracking database through edge computing nodes deployed in the tunnel section, performing simulations of the combined ventilation and evacuation strategy, extracting the vehicle position distribution and movement trend characteristics, and combining the carbon monoxide concentration, visibility index, and the calculation results of the conflict probability between fire smoke diffusion and vehicle evacuation paths in the tunnel environment monitoring data to construct a set of game parameters; Dynamically adjusting the collaborative response intensity between the fans according to the set of game parameters, and calculating the control instruction priorities of each fan in the fan group by balancing the smoke exhaust efficiency and the evacuation passage wind speed constraint in the dynamic game theory model; Based on the control instruction priorities, performing differential rotational speed and deflection angle control on the fan group to generate the tunnel environment feedback data after the fan group executes; According to the tunnel environment feedback data, iteratively updating the game strategy weights in the dynamic game theory model, and combining the edge computing nodes to perform rolling horizon optimization on the fan group control instructions until the vehicle movement trend and environmental indicators reach the preset collaborative stability interval.

2. The method according to claim 1, wherein Dynamically adjusting the collaborative response intensity between the fans according to the set of game parameters, and calculating the control instruction priorities of each fan in the fan group by balancing the smoke exhaust efficiency and the evacuation passage wind speed constraint in the dynamic game theory model, including: Mapping the vehicle position distribution, movement trend characteristics, carbon monoxide concentration, visibility index, and the conflict probability between smoke diffusion and evacuation paths in the set of game parameters to the environmental state parameters, vehicle state parameters, and risk parameters related to fan control in the dynamic game theory model respectively according to the preset priority rules; Dynamically adjusting the collaborative response intensity between the fans based on the change amplitudes of the environmental state parameters and vehicle state parameters; Constructing a game relationship between the smoke exhaust efficiency and the evacuation passage wind speed constraint in the dynamic game theory model; Calculating the control instruction priorities of each fan in the fan group according to the smoke exhaust efficiency weight and the evacuation passage wind speed constraint weight in the game relationship, where the priority parameters are jointly determined by the environmental state parameters at the position where the fan is located, the coverage contribution degree of the fan to the smoke diffusion path, and the matching degree between the wind speed direction in the evacuation passage and the vehicle movement trend.

3. The method according to claim 2, characterized in that, Constructing a game relationship between the smoke exhaust efficiency and the evacuation passage wind speed constraint in the dynamic game theory model, including: Dynamically associating the smoke diffusion rate with the negative pressure gradient in the coverage area of the fan group, and using the dynamic association result as the quantification basis for the smoke exhaust efficiency; Constructing a quantification rule for the evacuation passage wind speed constraint based on the angle threshold between the evacuation path direction and the local wind speed direction generated by the fan group; Mapping the quantification result of the smoke exhaust efficiency and the quantification rule of the evacuation passage wind speed constraint to the objective function and constraint conditions in the dynamic game theory model respectively; In the dynamic game theory model, constructing a game relationship between the objective function and the constraint conditions, and using the quantification rule of the evacuation passage wind speed constraint to correct the constraint of the objective function. Based on the results of the game relationship construction, generate a game strategy set for the smoke exhaust efficiency and the evacuation passage wind speed constraint.

4. The method according to claim 1, characterized in that, The edge computing nodes deployed in the tunnel section parse the information obtained from the radar-vision integrated vehicle tracking database, execute the simulation of the ventilation and evacuation joint strategy, extract the vehicle position distribution and movement trend characteristics, and combine the carbon monoxide concentration, visibility index in the tunnel environment monitoring data, and the calculation results of the conflict probability between the fire smoke diffusion and the vehicle evacuation path to construct a game parameter set, including: Extract the dynamic trajectory data of multiple types of vehicles in the tunnel from the radar-vision integrated vehicle tracking database, parse it into structured information of vehicle position coordinates, instantaneous speed, and movement direction, and map the structured information to the distribution heat map and movement trend vector field of the vehicle in the tunnel section based on the tunnel topology structure; Based on the distribution heat map and movement trend vector field, execute the simulation of the ventilation and evacuation joint strategy in the edge computing node, and extract the aggregation characteristics of the vehicle position distribution and the continuity characteristics of the movement trend in the simulation results; Map the carbon monoxide concentration and visibility index in the tunnel environment monitoring data to the dynamic influence factors of the smoke diffusion rate and air pollutant concentration respectively, and combine the calculation results of the conflict probability between the fire smoke diffusion path and the vehicle evacuation path to generate the threat level parameter of the smoke diffusion to the evacuation path; Based on the aggregation characteristics of the vehicle position distribution, the continuity characteristics of the movement trend, and the threat level parameter of the smoke diffusion, construct the game parameter set of the dynamic game theory model.

5. The method according to claim 4, wherein Based on the aggregation characteristics of the vehicle position distribution, the continuity characteristics of the movement trend, and the threat level parameter of the smoke diffusion, construct the game parameter set of the dynamic game theory model, including: Map the carbon monoxide concentration, visibility index, and smoke diffusion rate to the pollutant concentration parameter, visibility influence parameter, and smoke diffusion rate parameter in the environmental state parameters respectively; Map the aggregation degree of the distribution heat map and the continuity of the movement trend vector field to the vehicle aggregation degree parameter and movement continuity parameter in the vehicle state parameters respectively; Map the conflict probability and threat level parameter between the smoke diffusion path and the evacuation path to the path conflict parameter and threat level parameter in the risk parameters respectively; Based on the pollutant concentration parameter, visibility influence parameter, and smoke diffusion rate parameter, construct the environmental state parameter set; Based on the vehicle aggregation degree parameter and movement continuity parameter, construct the vehicle state parameter set; Based on the path conflict parameter and threat level parameter, construct the risk parameter set; Jointly construct the game parameter set of the dynamic game theory model with the environmental state parameter set, vehicle state parameter set, and risk parameter set.

6. The method according to claim 1, wherein According to the tunnel environment feedback data, iteratively update the game strategy weights in the dynamic game theory model, and combine the edge computing node to perform rolling horizon optimization on the fan group control instructions until the vehicle movement trend and environmental indicators reach the preset coordinated stable interval, including: Decompose the carbon monoxide concentration, visibility index, local wind speed direction, and vehicle movement trend characteristics in the tunnel environment feedback data into environmental state deviation parameters and vehicle state deviation parameters; Based on the environmental state deviation parameters and vehicle state deviation parameters, iteratively update the game strategy weights in the dynamic game theory model; Based on the updated game strategy weights, generate a candidate instruction set for the fan group control instruction in the edge computing node; Perform rolling horizon optimization on the candidate instruction set, where the time window length of the rolling horizon optimization is defined by the dynamic adjustment amplitude of the evacuation path in the vehicle movement trend characteristics, and the optimization objective is to minimize the weighted cumulative value of the environmental state deviation parameters and vehicle state deviation parameters within the time window until the vehicle movement trend and environmental indicators reach a preset collaborative stable interval.

7. The method according to claim 1, wherein Based on the control instruction priority, perform differential speed and deflection angle control on the fan group, and generate the tunnel environment feedback data after the fan group executes, including: Map the control instruction priority to the speed parameter and deflection angle parameter of each fan in the fan group; Based on the speed parameter and deflection angle parameter, generate a differential control instruction set for each fan in the fan group; Execute the differential control instruction set to control each fan in the fan group, and monitor the consistency coefficient between the negative pressure gradient change rate and the local wind speed direction in the fan coverage area; Generate the tunnel environment feedback data based on the operating state of the fan group after executing the differential control instruction set.

8. A dynamic optimization system for collaborative control of a group of tunnel fans, characterized in that It includes the following steps: An acquisition module, which acquires the dynamic position, speed, and traffic flow density information of multiple types of vehicles in the tunnel based on the radar-vision integrated vehicle tracking database; A calculation module, which analyzes the information obtained from the radar-vision integrated vehicle tracking database through the edge computing nodes deployed in the tunnel section, executes the simulation of the ventilation and evacuation joint strategy and extracts the vehicle position distribution and movement trend characteristics, and constructs a game parameter set in combination with the carbon monoxide concentration, visibility index, and the calculation results of the conflict probability between the fire smoke diffusion and the vehicle evacuation path in the tunnel environment monitoring data; The calculation module is also used to dynamically adjust the collaborative response intensity between the fans according to the game parameter set weights, and calculate the control instruction priority of each fan in the fan group by balancing the smoke exhaust efficiency and the evacuation channel wind speed constraint in the dynamic game theory model; A generation module, which performs differential speed and deflection angle control on the fan group based on the control instruction priority, and generates the tunnel environment feedback data after the fan group executes; An optimization module, which iteratively updates the game strategy weights in the dynamic game theory model according to the tunnel environment feedback data, and performs rolling horizon optimization on the fan group control instruction through the edge computing node until the vehicle movement trend and environmental indicators reach a preset collaborative stable interval.

9. A computing device, characterized in that, It includes a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement a dynamic optimization method for collaborative control of a tunnel fan group as described in any one of claims 1 to 7.

10. A computer storage medium, characterized in that, A computer program is stored, and when the computer program is executed by a computer, it implements a dynamic optimization method for collaborative control of a group of tunnel fans as described in any one of claims 1 to 7.

Images