Unmanned driving system of tunnel transportation battery car

Abstract

The invention discloses a tunnel transportation battery car unmanned driving system, which relates to the technical field of tunnel transportation and comprises an environment sensing module, a positioning navigation module, a central control module, an execution control module and a communication module. The environment sensing module is used for accurately acquiring obstacle and road condition information; the positioning navigation module determines a real-time position and generates an optimal path; the central control module generates a control instruction; the execution control module realizes vehicle driving control; and the communication module ensures stable transmission of data. According to the invention, the problems of easy fatigue of a driver, large environmental harm, low transportation efficiency and the like of a traditional manned battery car are solved, and the defects of insufficient positioning precision, incomplete environmental perception, inflexible decision control and the like of an existing unmanned technology in a tunnel environment are overcome; safe, efficient and intelligent unmanned driving of the tunnel transportation battery car is achieved, the safety and efficiency of tunnel engineering construction are improved, and transportation resource waste and construction period delay are reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of tunnel transportation, and more particularly to an unmanned driving system for a tunnel transportation battery vehicle. Background Art

[0002] Currently, electric vehicles are commonly used to transport materials such as slag and construction materials during tunnel construction. Traditional manned electric vehicles present numerous challenges, including driver fatigue after prolonged work in tunnels, which increases the risk of operational errors. Tunnel environments are harsh, with dust, humidity, and dim lighting, which pose a threat to driver health and safety. Manual driving makes it difficult to precisely control transportation efficiency, leading to wasted resources and delays. With the advancement of automation technology, achieving unmanned tunnel transport electric vehicles has become a key approach to improving the safety and efficiency of tunnel construction. However, existing unmanned driving technologies, in complex tunnel environments, suffer from insufficient positioning accuracy, incomplete environmental perception, and inflexible decision-making and control, making them difficult to meet the actual needs of tunnel transportation.

[0003] Therefore, how to provide an unmanned driving system for tunnel transport electric vehicles is an urgent problem that needs to be solved by those skilled in the art. Summary of the Invention

[0004] In view of this, the present invention provides an unmanned driving system for tunnel transport electric vehicles to solve the problems of large safety hazards, low transportation efficiency and poor adaptability of existing unmanned driving technology in tunnel environments in the existing technology of manned tunnel transport electric vehicles, and realize safe, efficient and intelligent unmanned driving of tunnel transport electric vehicles.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions:

[0006] An unmanned driving system for a tunnel transport battery vehicle, comprising: an environment perception module, a positioning and navigation module, a central control module, an execution control module, and a communication module;

[0007] The environmental perception module is used to obtain environmental information around the battery vehicle, and transmit the collected information to the positioning navigation module and the central control module after pre-processing;

[0008] The positioning and navigation module is used to determine the real-time position of the battery vehicle in the tunnel and plan the driving path;

[0009] The central control module is used to receive information transmitted by the positioning and navigation module and the environmental perception module, analyze and process it, generate control instructions according to the preset control strategy, and send the instructions to the execution control module;

[0010] The execution control module is used to receive control instructions from the central control module and control the driving of the battery vehicle;

[0011] The communication module is used to realize data transmission between modules and between the electric vehicle and the remote monitoring center; the staff of the remote monitoring center monitors the operating status of the electric vehicle in real time through the communication module, and remotely intervenes in the operation of the vehicle when an abnormality occurs.

[0012] Optionally, the positioning and navigation module adopts multi-sensor fusion positioning technology, combining the Beidou satellite positioning system, the inertial navigation system and the UWB positioning base station deployed in the tunnel; the Beidou satellite positioning system can provide rough location information at the tunnel entrance and some open areas; the inertial navigation system uses accelerometers and gyroscopes to measure the acceleration and angular velocity of the electric vehicle in real time, and calculates the vehicle's posture and position changes; UWB positioning base stations are arranged at fixed distances along the tunnel wall, and the electric vehicle is positioned by transmitting and receiving UWB signals; in terms of path planning, the positioning and navigation module uses an algorithm that combines the ant colony algorithm and the dynamic particle swarm algorithm to generate the optimal driving path according to the construction layout, real-time traffic conditions and transportation task requirements in the tunnel, and transmits the path information to the central control module in real time.

[0013] Optionally, the specific steps of the algorithm combining the ant colony algorithm and the dynamic particle swarm algorithm to generate the optimal driving path are as follows:

[0014] S1. Utilizing fixed monitoring equipment and vehicle-mounted laser detection devices within the tunnel, data on the tunnel's internal environment is collected. Using rasterization technology, the three-dimensional tunnel space is converted into a two-dimensional grid model. Safety buffer zones are defined based on the dimensional parameters of the electric vehicle's profile and its travel speed. Given that tunnel roads are mostly straight or gently curved, the traditional grid expansion model is simplified to a four-neighborhood model, guiding electric vehicles to prioritize straight paths.

[0015] S2. Using "number of path turns x total path length" as the optimization objective function, the ant colony optimization algorithm is used to perform global path planning on the tunnel's pre-stored map to generate an initial obstacle avoidance route. While driving along the global path, the electric vehicle uses a laser detection device to obtain real-time information about the surrounding environment, and combines this with data from wheel speed sensors and positioning devices to determine the state of obstacles. If the obstacle is stationary or its trajectory does not intersect with the vehicle's path, the original route is maintained. If a collision risk is predicted, the vehicle first decelerates and brakes. If a collision is unavoidable, the vehicle fine-tunes its driving direction toward the open side of the tunnel to avoid the obstacle. If local path replanning is required, the process proceeds to S3.

[0016] S3. When a local path needs to be replanned, the ant particles in the ant colony algorithm are given attribute characteristics, learning parameters and inertia weights are set, a large number of candidate paths are generated and the optimal paths are screened. Pheromones are marked on these paths, multiple virtual ants are distributed on the path nodes, and the initial positions of sensitive particles are set.

[0017] S4. Calculate the objective function value based on the current position as the fitness evaluation index, set the current fitness value as the individual optimal value, and find the global optimal value; at the same time, calculate the fitness value of the sensitive particle and include the initial position of each ant in the current solution set;

[0018] S5. When the environment changes, that is, when the fitness value of the sensitive particle changes by more than the set threshold, the ant parameters are reinitialized proportionally and the ant paths are cross-mutated. The objective function value is calculated based on the new position. If a better solution is obtained, it is adopted; otherwise, it is discarded. The length of the path generated by each ant is calculated, the current optimal solution is recorded, and the pheromone concentration of the path is adjusted according to the update rule.

[0019] Optionally, the environmental perception module is used to obtain environmental information around the electric vehicle, including obstacle information and road condition information; the environmental perception module includes multiple laser radars, which detect obstacle information by emitting laser beams and receiving reflected signals.

[0020] Optionally, the environment perception module obtains obstacle information specifically as follows:

[0021] Collect multi-source lidar point cloud data, including left front, right front, and top lidar data;

[0022] Generate global point cloud data based on point cloud fusion algorithm;

[0023] Introducing a priori tunnel map containing information about impassable areas;

[0024] Based on the prior map, static background point clouds are filtered out from the global point cloud to extract dynamic target point clouds.

[0025] Obstacle classification and state estimation are completed through target recognition algorithm.

[0026] Optionally, the multi-source lidar point cloud data includes: the left front field of view point cloud obtained by the left lidar, the right front field of view point cloud obtained by the right lidar, and the full-view field point cloud obtained by the top lidar.

[0027] Optionally, obtaining multi-source lidar point cloud data includes:

[0028] Get the left front raw field of view point cloud from the left lidar, get the right front raw field of view point cloud from the right lidar, and get the full circumference raw field of view point cloud from the top lidar;

[0029] The left front original field of view point cloud, the right front original field of view point cloud and the full circumference original field of view point cloud are transformed from the lidar coordinate system to the vehicle coordinate system in the Z-axis direction to obtain the Z-axis left vehicle field of view point cloud data, the Z-axis right vehicle field of view point cloud data and the Z-axis top vehicle field of view point cloud data;

[0030] Denoising the Z-axis left vehicle field of view point cloud data, the Z-axis right vehicle field of view point cloud data, and the Z-axis top vehicle field of view point cloud data;

[0031] The denoised Z-axis left vehicle field of view point cloud data, the Z-axis right vehicle field of view point cloud data, and the Z-axis top vehicle field of view point cloud data are transformed from the lidar coordinate system to the vehicle body coordinate system in the X-axis and Y-axis directions to obtain the left vehicle field of view point cloud data, the right vehicle field of view point cloud data, and the top vehicle field of view point cloud data;

[0032] The left vehicle field of view point cloud data, the right vehicle field of view point cloud data and the top vehicle field of view point cloud data are transformed from the vehicle body coordinate system to the map coordinate system to obtain the left front field of view point cloud, the right front field of view point cloud and the full-circle field of view point cloud.

[0033] Optionally, the central control module uses PLC as the core controller. After receiving the path information from the positioning and navigation module and the environmental information from the environmental perception module, it first analyzes the environmental information through a deep learning algorithm to identify different types of obstacles and traffic conditions. Then, it combines the path planning information and uses the PID control algorithm to calculate the control parameters of the electric vehicle and generate corresponding control instructions.

[0034] Optionally, the execution control module includes a motor drive unit, a steering control unit and a braking control unit; the motor drive unit adjusts the speed and torque of the electric vehicle drive motor according to the control instructions to achieve acceleration, deceleration and constant speed driving of the vehicle; the steering control unit adjusts the driving direction of the vehicle by controlling the rotation angle of the steering motor; the braking control unit controls the braking system according to the control instructions to achieve emergency braking or normal deceleration of the vehicle.

[0035] Through the above technical solutions, it can be seen that compared with the existing technology, the present invention discloses an unmanned driving system for tunnel transport battery vehicles, including an environmental perception module, a positioning and navigation module, a central control module, an execution control module and a communication module. The environmental perception module collects and processes data through multi-source laser radar, and can accurately obtain information on obstacles and road conditions; the positioning and navigation module adopts a path planning algorithm that combines multi-sensor fusion positioning technology with ant colony algorithm and dynamic particle swarm algorithm to determine the real-time position and generate the optimal path; the central control module uses PLC as the core, combines deep learning and PID control algorithm to generate control instructions; the execution control module realizes vehicle driving control; the communication module ensures data transmission between each module and the remote monitoring center. The present invention solves the problems of traditional manned battery vehicles, such as easy fatigue of drivers, great environmental damage, and low transportation efficiency. At the same time, it overcomes the defects of existing unmanned driving technology in tunnel environments, such as insufficient positioning accuracy, incomplete environmental perception, and inflexible decision-making and control, and realizes safe, efficient and intelligent unmanned driving of tunnel transport battery vehicles, improves the safety and efficiency of tunnel engineering construction, and reduces waste of transportation resources and construction delays. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are merely embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0037] Figure 1 This is a schematic diagram of the system structure provided by the present invention. DETAILED DESCRIPTION

[0038] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0039] The embodiment of the present invention discloses an unmanned driving system for a tunnel transport battery vehicle, such as Figure 1 As shown, it includes: an environmental perception module, a positioning and navigation module, a central control module, an execution control module and a communication module;

[0040] The environmental perception module is used to obtain environmental information around the battery vehicle, and transmit the collected information to the positioning navigation module and the central control module after pre-processing;

[0041] The positioning and navigation module is used to determine the real-time position of the battery vehicle in the tunnel and plan the driving path;

[0042] The central control module is used to receive information transmitted by the positioning and navigation module and the environmental perception module, analyze and process it, generate control instructions according to the preset control strategy, and send the instructions to the execution control module;

[0043] The execution control module is used to receive control instructions from the central control module and control the driving of the battery vehicle;

[0044] The communication module is used to realize data transmission between modules and between the electric vehicle and the remote monitoring center; the staff of the remote monitoring center monitors the operating status of the electric vehicle in real time through the communication module, and remotely intervenes in the operation of the vehicle when an abnormality occurs.

[0045] In a specific embodiment, the positioning and navigation module adopts multi-sensor fusion positioning technology, combining the Beidou satellite positioning system, the inertial navigation system and the UWB positioning base station deployed in the tunnel; the Beidou satellite positioning system can provide rough location information at the tunnel entrance and some open areas; the inertial navigation system uses accelerometers and gyroscopes to measure the acceleration and angular velocity of the electric vehicle in real time, and calculates the vehicle's posture and position changes; the UWB positioning base station is deployed at fixed intervals along the tunnel wall, and realizes the positioning of the electric vehicle by transmitting and receiving UWB signals; in terms of path planning, the positioning and navigation module uses an algorithm that combines the ant colony algorithm and the dynamic particle swarm algorithm to generate the optimal driving path based on the construction layout, real-time traffic conditions and transportation task requirements in the tunnel, and transmits the path information to the central control module in real time.

[0046] In a specific embodiment, the specific steps of generating the optimal driving path by combining the ant colony algorithm with the dynamic particle swarm algorithm are as follows:

[0047] S1. Utilizing fixed monitoring equipment and vehicle-mounted laser detection devices within the tunnel, data on the tunnel's internal environment is collected. Using rasterization technology, the three-dimensional tunnel space is converted into a two-dimensional grid model. Safety buffer zones are defined based on the dimensional parameters of the electric vehicle's profile and its travel speed. Given that tunnel roads are mostly straight or gently curved, the traditional grid expansion model is simplified to a four-neighborhood model, guiding electric vehicles to prioritize straight paths.

[0048] S2. Using "number of path turns x total path length" as the optimization objective function, the ant colony optimization algorithm is used to perform global path planning on the tunnel's pre-stored map to generate an initial obstacle avoidance route. While driving along the global path, the electric vehicle uses a laser detection device to obtain real-time information about the surrounding environment, and combines this with data from wheel speed sensors and positioning devices to determine the state of obstacles. If the obstacle is stationary or its trajectory does not intersect with the vehicle's path, the original route is maintained. If a collision risk is predicted, the vehicle first decelerates and brakes. If a collision is unavoidable, the vehicle fine-tunes its driving direction toward the open side of the tunnel to avoid the obstacle. If local path replanning is required, the process proceeds to S3.

[0049] S3. When a local path needs to be replanned, the ant particles in the ant colony algorithm are given attribute characteristics, learning parameters and inertia weights are set, a large number of candidate paths are generated and the optimal paths are screened. Pheromones are marked on these paths, multiple virtual ants are distributed on the path nodes, and the initial positions of sensitive particles are set.

[0050] S4. Calculate the objective function value based on the current position as the fitness evaluation index, set the current fitness value as the individual optimal value, and find the global optimal value; at the same time, calculate the fitness value of the sensitive particle and include the initial position of each ant in the current solution set;

[0051] S5. When the environment changes, that is, when the fitness value of the sensitive particle changes by more than the set threshold, the ant parameters are reinitialized proportionally and the ant paths are cross-mutated. The objective function value is calculated based on the new position. If a better solution is obtained, it is adopted; otherwise, it is discarded. The length of the path generated by each ant is calculated, the current optimal solution is recorded, and the pheromone concentration of the path is adjusted according to the update rule.

[0052] In a specific embodiment, the environmental perception module is used to obtain environmental information around the electric vehicle, including obstacle information and road condition information; the environmental perception module includes multiple laser radars, which detect obstacle information by emitting laser beams and receiving reflected signals.

[0053] In a specific embodiment, the environment perception module obtains obstacle information by:

[0054] Collect multi-source lidar point cloud data, including left front, right front, and top lidar data;

[0055] Generate global point cloud data based on point cloud fusion algorithm;

[0056] Introducing a priori tunnel map containing information about impassable areas;

[0057] Based on the prior map, static background point clouds are filtered out from the global point cloud to extract dynamic target point clouds.

[0058] Obstacle classification and state estimation are completed through target recognition algorithm.

[0059] In a specific embodiment, the multi-source lidar point cloud data includes: a left front field of view point cloud obtained by a left lidar, a right front field of view point cloud obtained by a right lidar, and a full-view field of view point cloud obtained by a top lidar.

[0060] In a specific embodiment, obtaining multi-source laser radar point cloud data includes:

[0061] Get the left front raw field of view point cloud from the left lidar, get the right front raw field of view point cloud from the right lidar, and get the full circumference raw field of view point cloud from the top lidar;

[0062] The left front original field of view point cloud, the right front original field of view point cloud and the full circumference original field of view point cloud are transformed from the lidar coordinate system to the vehicle coordinate system in the Z-axis direction to obtain the Z-axis left vehicle field of view point cloud data, the Z-axis right vehicle field of view point cloud data and the Z-axis top vehicle field of view point cloud data;

[0063] De-noising the Z-axis left vehicle field of view point cloud data, the Z-axis right vehicle field of view point cloud data, and the Z-axis top vehicle field of view point cloud data;

[0064] The denoised Z-axis left vehicle field of view point cloud data, the Z-axis right vehicle field of view point cloud data, and the Z-axis top vehicle field of view point cloud data are transformed from the lidar coordinate system to the vehicle body coordinate system in the X-axis and Y-axis directions to obtain the left vehicle field of view point cloud data, the right vehicle field of view point cloud data, and the top vehicle field of view point cloud data;

[0065] The left vehicle field of view point cloud data, the right vehicle field of view point cloud data and the top vehicle field of view point cloud data are transformed from the vehicle body coordinate system to the map coordinate system to obtain the left front field of view point cloud, the right front field of view point cloud and the full-circle field of view point cloud.

[0066] In a specific embodiment, the central control module uses PLC as the core controller. After receiving the path information from the positioning and navigation module and the environmental information from the environmental perception module, it first analyzes the environmental information through a deep learning algorithm to identify different types of obstacles and traffic conditions. Then, combined with the path planning information, the PID control algorithm is used to calculate the control parameters of the electric vehicle and generate corresponding control instructions.

[0067] In a specific embodiment, the execution control module includes a motor drive unit, a steering control unit and a braking control unit; the motor drive unit adjusts the speed and torque of the electric vehicle drive motor according to the control instructions to achieve acceleration, deceleration and constant speed driving of the vehicle; the steering control unit adjusts the vehicle's driving direction by controlling the rotation angle of the steering motor; the braking control unit controls the braking system according to the control instructions to achieve emergency braking or normal deceleration of the vehicle.

[0068] A specific embodiment is introduced below to further illustrate the method of using the system of the present invention.

[0069] Taking the slag transportation scenario of a tunnel project as an example, the implementation method and workflow of each module of the system are explained in detail:

[0070] 1. System Composition and Hardware Deployment

[0071] 1. Environmental Perception Module

[0072] Hardware configuration: A 16-line lidar, such as the Velodyne VLP-16, is installed on the top of the electric scooter for full-field point cloud acquisition. An 8-line lidar, such as the Ouster OS0-128, is installed on the left and right front sides, covering 120° on the left and right front, respectively.

[0073] Data collection frequency: The lidar emits laser beams at a frequency of 10 Hz to collect obstacle distance, direction, and reflection intensity data.

[0074] 2. Positioning and navigation module

[0075] Multi-sensor deployment: The Beidou satellite positioning module is installed on the vehicle roof, providing meter-level positioning at tunnel entrances and open areas. The inertial navigation system (which can be the ADI ADIS16505) is integrated into the vehicle controller, using an accelerometer (range ±50g) and a gyroscope (range ±2000° / s) to measure the vehicle's motion status in real time. UWB positioning base stations (which can be the DecaWave DW1000) are deployed every 20 meters along the tunnel wall. The on-board UWB tags receive base station signals to achieve centimeter-level positioning.

[0076] 3. Central control module

[0077] Core controller: A Siemens S7-1200 PLC with expansion modules (AI / AO modules for signal conversion and high-speed counter modules for encoder data processing). Edge computing unit: An integrated NVIDIA Jetson XavierNX for running deep learning algorithms.

[0078] 4.Execution control module

[0079] Motor drive unit: permanent magnet synchronous motor (power 30kW) with vector control drive;

[0080] Steering control unit: servo motor (rated torque 5N·m) drives the rack and pinion steering mechanism, with an angle accuracy of ±0.5°;

[0081] Braking control unit: air brake system (brake pressure 0.6-0.8MPa) combined with motor braking, emergency braking distance ≤5m (at vehicle speed 20km / h).

[0082] 5. Communication module

[0083] A 5G private network (frequency band 26GHz) is deployed in the tunnel, and vehicle-mounted terminals (such as Huawei ME909s-821) support a transmission rate of 1.2Gbps; the remote monitoring center accesses the tunnel network through a VPN gateway to achieve real-time data interaction.

[0084] 2. Workflow Implementation Steps (1) Environmental Perception and Data Processing

[0085] LiDAR point cloud acquisition: The top LiDAR obtains a full 360° point cloud (point cloud density is approximately 1 point / ㎡), and the left and right front LiDARs obtain corresponding field of view point clouds respectively. The raw data is transmitted to the edge computing unit via the on-board Ethernet (1Gbps).

[0086] Point cloud fusion and obstacle recognition:

[0087] 1. De-noise the original point cloud using voxel grid filtering (voxel size 0.2m × 0.2m × 0.2m) and unify the point cloud to the map coordinate system through coordinate transformation (pre-calibration of the LiDAR and vehicle coordinate system conversion matrices based on the on-board calibration plate);

[0088] 2. Introducing the tunnel BIM model as a priori map, filtering out static point clouds such as tunnel walls and supports, and extracting dynamic point clouds (construction personnel and transport vehicles);

[0089] 3. Based on the PointNet++ algorithm, dynamic point clouds are classified to identify obstacle types (people, vehicles, and stacked objects), and their motion trajectories are predicted using Kalman filtering (prediction error ≤ 0.3m).

[0090] (2) Positioning and path planning

[0091] Multi-sensor fusion positioning:

[0092] 1. In the tunnel entrance area (0-500m from the entrance), Beidou positioning data (accuracy 1-3m) and inertial navigation data are fused through Kalman filtering to output the preliminary position;

[0093] 2. Inside the tunnel, the UWB base station and inertial navigation data are integrated: the UWB positioning base station periodically transmits pulse signals (at intervals of 100ms). After receiving the signals, the on-board tag calculates the TOF (time of flight) and determines the position through triangulation (accuracy of ±5cm). The sliding window is optimized based on the inertial navigation data (sampling frequency of 100Hz) to output the real-time position (error ≤10cm).

[0094] Path planning algorithm execution:

[0095] 1. Global Path Planning (Ant Colony Algorithm): A two-dimensional grid map (grid size 0.5m × 0.5m) was generated based on the tunnel BIM model. A safety buffer zone (extended by 0.8m) was set based on the outline of the electric vehicle (length 5m × width 2.5m). Using "total path length × number of turns × 1.5" as the objective function, the number of ants was set to 50, the pheromone volatility coefficient was 0.1, and 50 iterations were performed to generate a global path from the muck storage area to the tunnel entrance unloading area (e.g., starting point → point A → point B → end point, total path length 1.2km, number of turns 2).

[0096] 2. Local path replanning (dynamic particle swarm algorithm): When the lidar detects an obstacle 30 meters ahead (such as temporarily stacked rebar), local replanning is triggered: the particle swarm size is set to 30 particles, the inertia weight is linearly reduced from 0.9 to 0.4, and the learning factor c1 = c2 = 2. The fitness function is "obstacle avoidance time × path deviation from the global path". A detour path is generated 20 times (for example, deviating 1.5 meters to the right to bypass the obstacle and then returning to the global path). The replanning time is ≤ 200ms.

[0097] (3) Decision-making, control and execution

[0098] The central control module processes:

[0099] 1. The edge computing unit transmits obstacle identification results (e.g., "There is a moving construction vehicle 15 meters ahead, speeding 5 km / h") and path planning information to the PLC.

[0100] 2.PLC calculates control parameters based on PID control algorithm:

[0101] Speed ​​control: Target speed is 20 km / h. If there is an obstacle ahead, the motor speed is adjusted by proportional coefficient Kp = 0.8, integral coefficient Ki = 0.2, and differential coefficient Kd = 0.1 to decelerate to 10 km / h.

[0102] Steering control: Based on the path deviation (e.g., the current position deviates by 0.5 m from the planned path), the steering motor angle is calculated using Kp = 1.2, Ki = 0.3, and Kd = 0.5 to adjust the driving direction.

[0103] Execute control module actions:

[0104] 1. The motor drive unit receives the speed command and adjusts the motor speed through the PWM signal (frequency 10kHz). The torque output range is 0-300N·m;

[0105] 2. The steering control unit drives the servo motor to rotate and provides real-time feedback of the steering angle to the PLC (feedback accuracy ±0.1°);

[0106] 3. In an emergency (e.g., predicted collision time ≤ 1.5s), the brake control unit activates the air brake and the motor brake simultaneously to ensure the vehicle stops safely.

[0107] (4) Remote monitoring and communication

[0108] The remote monitoring center displays vehicle location, speed, obstacle distribution and other data in real time through a web-based visual interface (refresh rate 10Hz);

[0109] When the system detects an abnormality (such as lidar failure or positioning error exceeding the threshold), the communication module automatically sends an alarm message to the monitoring center. The staff can switch to manual control mode through the remote intervention function and control the vehicle's driving with the joystick.

[0110] It should be further explained that the algorithms used in the present invention that are not specifically described in the above description are all existing algorithms.

[0111] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. Reference can be made to the common and similar parts between the various embodiments. For the devices disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple, and the relevant parts can be referred to the method description.

[0112] The above description of the disclosed embodiments is intended to enable one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be readily apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not limited to the embodiments shown herein but is intended to conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. An unmanned driving system for battery vehicles for tunnel transportation, characterized in that: include: Environmental perception module, positioning and navigation module, central control module, execution control module and communication module; The environmental perception module is used to obtain environmental information around the battery vehicle, and transmit the collected information to the positioning navigation module and the central control module after pre-processing; The positioning and navigation module is used to determine the real-time position of the battery vehicle in the tunnel and plan the driving path; The central control module is used to receive information transmitted by the positioning and navigation module and the environmental perception module, analyze and process it, generate control instructions according to the preset control strategy, and send the instructions to the execution control module; The execution control module is used to receive control instructions from the central control module and control the driving of the battery vehicle; The communication module is used to realize data transmission between modules and between the electric vehicle and the remote monitoring center; the staff of the remote monitoring center monitors the operating status of the electric vehicle in real time through the communication module, and remotely intervenes in the operation of the vehicle when an abnormality occurs.

2. The unmanned driving system for tunnel transportation battery vehicles according to claim 1 is characterized in that: The positioning and navigation module adopts multi-sensor fusion positioning technology, combining the Beidou satellite positioning system, the inertial navigation system and the UWB positioning base station deployed in the tunnel; the Beidou satellite positioning system provides rough location information at the tunnel entrance and open areas; the inertial navigation system uses accelerometers and gyroscopes to measure the acceleration and angular velocity of the electric vehicle in real time, and calculates the vehicle's posture and position changes; UWB positioning base stations are deployed at fixed intervals along the tunnel wall, and realize the positioning of the electric vehicle by transmitting and receiving UWB signals; in terms of path planning, the positioning and navigation module uses an algorithm that combines ant colony algorithm and dynamic particle swarm algorithm to generate the optimal driving path according to the construction layout, real-time traffic conditions and transportation task requirements in the tunnel, and transmits the path information to the central control module in real time.

3. The unmanned driving system for tunnel transportation battery vehicles according to claim 2, characterized in that: The specific steps of the algorithm combining the ant colony algorithm and the dynamic particle swarm algorithm to generate the optimal driving path are as follows: S1. Utilizing fixed monitoring equipment and vehicle-mounted laser detection devices within the tunnel, data on the tunnel's internal environment is collected. Using rasterization technology, the three-dimensional tunnel space is converted into a two-dimensional grid model. Safety buffer zones are defined based on the dimensional parameters of the electric vehicle's profile and its travel speed. Given that tunnel roads are mostly straight or gently curved, the traditional grid expansion model is simplified to a four-neighborhood model, guiding electric vehicles to prioritize straight paths. S2: Using "number of path turns x total path length" as the optimization objective function, the ant colony optimization algorithm is used to perform global path planning on the tunnel's pre-stored map, generating an initial obstacle avoidance route. While driving along the global path, the electric vehicle uses a laser detection device to obtain real-time information about the surrounding environment, and combines this with data from wheel speed sensors and positioning devices to determine the state of obstacles. If the obstacle is stationary or its trajectory does not intersect with the vehicle's path, the original route is maintained. If a collision risk is predicted, the vehicle first decelerates and brakes. If a collision is unavoidable, the vehicle fine-tunes its driving direction toward the open side of the tunnel to avoid the obstacle. If local path replanning is required, the process proceeds to S3. S3. When a local path needs to be replanned, the ant particles in the ant colony algorithm are given attribute characteristics, learning parameters and inertia weights are set, a large number of candidate paths are generated and the optimal paths are screened. Pheromones are marked on these paths, multiple virtual ants are distributed on the path nodes, and the initial positions of sensitive particles are set. S4. Calculate the objective function value based on the current position as the fitness evaluation index, set the current fitness value as the individual optimal value, and find the global optimal value; at the same time, calculate the fitness value of the sensitive particle and include the initial position of each ant in the current solution set; S5. When the environment changes, that is, when the fitness value of the sensitive particle changes by more than the set threshold, the ant parameters are reinitialized proportionally and the ant paths are cross-mutated. The objective function value is calculated based on the new position. If a better solution is obtained, it is adopted; otherwise, it is discarded. The length of the path generated by each ant is calculated, the current optimal solution is recorded, and the pheromone concentration of the path is adjusted according to the update rule.

4. The unmanned driving system for tunnel transportation battery vehicles according to claim 1, characterized in that: The environmental perception module is used to obtain environmental information around the electric vehicle, including obstacle information and road condition information; the environmental perception module includes multiple laser radars, which detect obstacle information by emitting laser beams and receiving reflected signals.

5. The unmanned driving system for tunnel transportation battery vehicles according to claim 4 is characterized in that: The environment perception module obtains obstacle information specifically as follows: Collect multi-source lidar point cloud data, including left front, right front, and top lidar data; Generate global point cloud data based on point cloud fusion algorithm; Introducing a priori tunnel map containing information about impassable areas; Based on the prior map, static background point clouds are filtered out from the global point cloud to extract dynamic target point clouds. Obstacle classification and state estimation are completed through target recognition algorithm.

6. The unmanned driving system for tunnel transportation battery vehicles according to claim 5, characterized in that: The multi-source laser radar point cloud data includes: the left front field of view point cloud obtained by the left laser radar, the right front field of view point cloud obtained by the right laser radar, and the full-circle field of view point cloud obtained by the top laser radar.

7. The unmanned driving system for battery vehicles for tunnel transportation according to claim 6, characterized in that: The obtaining of multi-source laser radar point cloud data includes: Get the left front raw field of view point cloud from the left lidar, get the right front raw field of view point cloud from the right lidar, and get the full circumference raw field of view point cloud from the top lidar; The left front original field of view point cloud, the right front original field of view point cloud and the full circumference original field of view point cloud are transformed from the lidar coordinate system to the vehicle coordinate system in the Z-axis direction to obtain the Z-axis left vehicle field of view point cloud data, the Z-axis right vehicle field of view point cloud data and the Z-axis top vehicle field of view point cloud data; Denoising the Z-axis left vehicle field of view point cloud data, the Z-axis right vehicle field of view point cloud data, and the Z-axis top vehicle field of view point cloud data; The denoised Z-axis left vehicle field of view point cloud data, the Z-axis right vehicle field of view point cloud data, and the Z-axis top vehicle field of view point cloud data are transformed from the lidar coordinate system to the vehicle body coordinate system in the X-axis and Y-axis directions to obtain the left vehicle field of view point cloud data, the right vehicle field of view point cloud data, and the top vehicle field of view point cloud data; The left vehicle field of view point cloud data, the right vehicle field of view point cloud data and the top vehicle field of view point cloud data are transformed from the vehicle body coordinate system to the map coordinate system to obtain the left front field of view point cloud, the right front field of view point cloud and the full-circle field of view point cloud.

8. The unmanned driving system for tunnel transportation battery vehicles according to claim 1, characterized in that: The central control module uses PLC as the core controller. After receiving the path information from the positioning and navigation module and the environmental information from the environmental perception module, it first analyzes the environmental information through a deep learning algorithm to identify different types of obstacles and traffic conditions. Then, combined with the path planning information, it uses the PID control algorithm to calculate the control parameters of the electric vehicle and generate corresponding control instructions.

9. The unmanned driving system for tunnel transportation battery vehicles according to claim 1, characterized in that: The execution control module includes a motor drive unit, a steering control unit and a brake control unit; the motor drive unit adjusts the speed and torque of the battery vehicle drive motor according to the control instructions to achieve vehicle acceleration, deceleration and constant speed driving; the steering control unit adjusts the vehicle's driving direction by controlling the rotation angle of the steering motor; the brake control unit controls the brake system according to the control instructions to achieve emergency braking or normal deceleration of the vehicle.

Images