Navigation system and navigation method of metro vehicle inspection robot

Abstract

The invention belongs to the technical field of metro inspection, and relates to a navigation system and a navigation method of a metro vehicle inspection robot, and the navigation system comprises a multi-sensor fusion positioning module, an intelligent path planning module, a map management and updating module, a communication and control module and a power management module. According to the navigation system and the navigation method of the metro vehicle inspection robot, the multi-sensor fusion positioning module integrates the advantages of various sensors and combines an advanced data fusion algorithm, so that the positioning precision of the robot in a complex environment is remarkably improved, and the influence of train parking deviation and environmental interference on positioning is effectively overcome; the map management and update module constructs a semantic map, adopts a flexible update strategy and an efficient storage management mechanism, timely and accurately reflects environment changes, provides a reliable basis for path planning, helps the robot to better adapt to different region and environment changes, realizes seamless navigation, and improves the flexibility and adaptability of inspection work.

Description

Technical Field

[0001] The invention belongs to the technical field of subway inspection, and in particular relates to a navigation system and a navigation method of a subway vehicle inspection robot. Background Art

[0002] In the urban rail transit system, the subway undertakes an important passenger transportation task. The structure of subway vehicles is complex and contains many key components. In order to ensure its safe and stable operation, a comprehensive and detailed inspection is required after the end of daily operation. The traditional manual inspection method mainly relies on inspectors holding simple tools such as flashlights, rulers, hammers, etc., and judging the condition of vehicle components by visual observation. This method is not only inefficient, but also significantly affected by the subjective factors of inspectors. The differences in technical levels and working conditions of different personnel can easily lead to problems such as missed inspections and false inspections, making it difficult to ensure the accuracy and reliability of the test results.

[0003] With the advancement of science and technology, robot inspection technology has gradually been applied to the field of subway vehicle inspection. However, the existing robot inspection technology still has many limitations. Some robots use the RGV method and need to travel along a dedicated track in the trench, which greatly limits the robot's range of activities, making it impossible to flexibly switch between different trenches and difficult to conduct comprehensive inspections of vehicles on multiple inspection tracks. At the same time, there is a deviation in the position of each train stop, making it difficult for the robot to accurately detect the corresponding components at the preset points, resulting in a significant reduction in the accuracy of the inspection. In addition, in terms of autonomous navigation and transportation inside and outside the train inspection channel, existing robots also have shortcomings and cannot efficiently and stably complete cross-regional inspection tasks, affecting the overall efficiency of subway vehicle inspection work. Summary of the invention

[0004] The present invention provides a navigation system and a navigation method for a subway vehicle inspection robot to solve the problems raised in the background technology.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] A navigation system for a subway vehicle inspection robot, comprising:

[0007] The multi-sensor fusion positioning module integrates laser sensors, magnetic sensors, IMUs and wheel odometers. It uses an algorithm combining extended Kalman filtering and particle filtering to fuse data and has sensor fault diagnosis and fault tolerance mechanisms.

[0008] Intelligent path planning module, which adopts hierarchical path planning strategy, combines A-star algorithm, Dijkstra algorithm, dynamic window method or fast exploration random tree algorithm, considers robot kinematics, dynamics, energy consumption and time constraints for path planning and optimization, and has dynamic adjustment and coordination mechanism;

[0009] The map management and update module uses two-dimensional laser radar combined with deep learning algorithms to build semantic maps, adopts an update strategy that combines event-driven and time-driven, and has an incremental update algorithm and efficient data storage and management mechanism;

[0010] The communication and control module uses 5G, Wi-Fi and Bluetooth communication technologies to build a redundant network, uses model predictive control algorithms combined with reinforcement learning algorithms for precise control, and has remote monitoring and fault diagnosis functions;

[0011] The power management module uses a high-precision battery monitoring chip to monitor battery parameters, and has intelligent charging strategies and energy recovery functions, as well as a power system thermal management system.

[0012] Furthermore, in the multi-sensor fusion positioning module, the laser sensor is installed on the top of the robot, the magnetic sensors are symmetrically distributed on both sides of the bottom of the robot, the IMU is installed at the center of gravity of the robot, and the wheel odometer is tightly connected to the driving wheels of the robot.

[0013] Furthermore, in the map management and update module, a multi-resolution map construction technology is adopted to dynamically adjust the map resolution according to the robot's motion range and accuracy requirements.

[0014] Furthermore, in the communication and control module, data encryption technology is used to encrypt the transmission data, and the communication link monitoring mechanism monitors indicators such as communication signal strength, data transmission rate and bit error rate in real time.

[0015] Furthermore, in the power management module, a battery balancing technology is used to balance the charging and discharging of each battery in the battery pack.

[0016] A navigation method of a navigation system, comprising the following steps:

[0017] S1: Initial positioning and calibration: After the robot is started, the multi-sensor fusion positioning module obtains the initial position information, and locates through laser sensor scanning combined with map data matching, and calibrates and corrects the magnetic sensor, IMU and wheel odometer; when the laser sensor is affected by harsh environment, the pre-set landmarks with coded information are used for visual positioning auxiliary calibration;

[0018] S2: Real-time positioning and path tracking: During the inspection process, the multi-sensor fusion positioning module continuously collects data to update the location information, the intelligent path planning module plans the path according to the real-time position and target point, and the communication and control module controls the robot movement and tracks the path; when multiple robots are working at the same time, the global path coordination mechanism is used to avoid path conflicts; in case of sudden emergency, the robot immediately brakes and replans a safe path;

[0019] S3: Map update and path replanning: When the robot detects changes in the environment, the map management and update module updates the map; if the update is not timely or inaccurate, the robot uses sensors to monitor the environment. If there is a large difference, the task is suspended, the map is partially rebuilt, and the path is replanned;

[0020] S4: Cross-regional navigation and map switching: When the robot switches inside or outside the train inspection channel, the map management and update module automatically switches the map. During cross-regional navigation, when the robot is driving on special terrain, the adaptive suspension system and anti-skid tires ensure stable driving. The map switching moment uses the IMU and wheel odometer data to transition and maintain positioning continuity.

[0021] S5: Charging management and task scheduling: The power management module monitors the power level. When the power is low, the communication and control module controls the robot to suspend the task and go to charge. When multiple robots need to be charged at the same time, a charging pile allocation strategy combining priority and first-come-first-served is adopted. Task scheduling dynamically adjusts the task priority according to factors such as the urgency of the inspection task, the remaining power and location of the robot.

[0022] Compared with the prior art, the present invention has the following beneficial effects:

[0023] 1. The navigation system and navigation method of the subway vehicle inspection robot, through the multi-sensor fusion positioning module, integrates the advantages of multiple sensors and combines advanced data fusion algorithms, which significantly improves the positioning accuracy of the robot in complex environments, effectively overcomes the influence of train stop deviation and environmental interference on positioning, ensures that the robot arrives at the inspection point accurately, improves the accuracy and reliability of detection, and lays a solid foundation for the safe operation of subway vehicles; the intelligent path planning module adopts a hierarchical strategy and multiple algorithms, which can quickly plan and adjust the path according to the real-time environment and task requirements, consider multiple constraints, optimize energy consumption, extend the robot's battery life, and the path coordination mechanism during multi-robot operation to avoid conflicts, improve the overall inspection efficiency, and ensure the safe and efficient operation of the robot; the map management and update module constructs a semantic map, adopts a flexible update strategy and an efficient storage management mechanism, timely and accurately reflects environmental changes, provides a reliable basis for path planning, helps robots better adapt to different regions and environmental changes, realizes seamless navigation, and improves the flexibility and adaptability of inspection work.

[0024] 2. The navigation system and navigation method of the subway vehicle inspection robot build a redundant communication network through the communication and control module, and adopt advanced control algorithms to ensure stable communication and precise control. The remote monitoring and fault diagnosis functions facilitate timely detection and handling of robot faults, improve the reliability and stability of the robot's work, and the operator can grasp the robot's operating status at any time through remote monitoring, and can intervene quickly in case of emergencies, thereby reducing the robot's downtime and improving the continuity of the inspection work; the power management module improves the battery's efficiency and life, enhances the robot's endurance, and reduces The number of task interruptions caused by insufficient power enables the robot to complete more inspection tasks after a single charge, reducing operating costs and improving the efficiency and reliability of inspection work; the navigation method covers all key links from initial positioning to task completion, and each step is closely coordinated, with intelligent charging management and task scheduling. It dynamically adjusts task priorities based on factors such as task urgency, robot power and location, reasonably allocates resources, ensures that urgent tasks are executed first, balances the robot's workload, further improves inspection efficiency and quality, and realizes efficient and autonomous navigation and transfer of robots inside and outside the train inspection channel, reducing human intervention and improving the automation level and overall effectiveness of subway vehicle inspection. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0026] Figure 1 This is a schematic diagram of the structure of the local subway vehicle inspection robot navigation system; DETAILED DESCRIPTION

[0027] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. 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 creative work are within the scope of protection of the present invention.

[0028] Example

[0029] See also Figure 1 , the present invention provides the following technical solutions:

[0030] The navigation system of the subway vehicle inspection robot includes:

[0031] Multi-sensor fusion positioning module, this module integrates laser sensors, magnetic sensors, IMU and wheel odometers. Each sensor is optimized on the robot body. The laser sensor is installed on the top of the robot, has a wide scanning field of view, can scan the surrounding environment in real time, obtain accurate distance information to build a local map; the magnetic sensor is symmetrically distributed on both sides of the bottom, reads the magnetic stripe information to assist in positioning and orientation, and plays a key role when the laser sensor is interfered with; the IMU is installed at the center of gravity to measure acceleration and attitude angle in real time; the wheel odometer is closely connected to the drive wheel, accurately records the driving distance and speed, and uses an algorithm combining extended Kalman filtering and particle filtering to fuse data. According to the characteristics of sensor data and environmental changes, the weights of the two algorithms are dynamically adjusted; at the same time, a complete sensor fault diagnosis system is established, and the working status of the sensor is monitored in real time through statistical analysis, threshold judgment and data correlation test. Once a fault is detected, the fault tolerance mechanism is automatically started, and data repair, switching to backup sensors or other positioning solutions are taken according to the type of fault to ensure the reliability of positioning;

[0032] The intelligent path planning module adopts a hierarchical path planning strategy: at the global level, the A-star algorithm or Dijkstra algorithm is used to plan the global path according to the preset inspection tasks and target points, and the main travel direction and key nodes are determined; at the local level, the dynamic window method or the rapid exploration random tree (RRT) algorithm is used to combine real-time obstacle information and robot kinematic constraints to locally optimize the global path and generate a safe and smooth local path. When planning the path, the robot kinematics, dynamics, energy consumption and time constraints are fully considered. By establishing the corresponding model, these constraints are integrated into the algorithm, and the path that meets multiple constraints and has low energy consumption is given priority. The inspection sequence and speed are reasonably arranged to ensure that the inspection task is completed on time; at the same time, the feasibility of the path is monitored in real time. When the path is infeasible due to environmental changes, the re-planning mechanism is triggered, and the existing planning results and environmental information are used to improve the re-planning efficiency; in the multi-robot operation scenario, a path coordination mechanism is established to exchange information through wireless communication to avoid path conflicts;

[0033] The map management and update module uses two-dimensional laser radar combined with deep learning algorithms to build semantic maps. On the basis of traditional raster maps, it uses deep learning target detection algorithms to identify and classify obstacles, train parts, landmarks, etc. in the scan data, and gives the map semantic information. It adopts multi-resolution map construction technology and dynamically adjusts the map resolution according to the robot's motion range and accuracy requirements. It adopts an update strategy that combines event-driven and time-driven. When the environment changes significantly, it immediately triggers event-driven updates and quickly updates the map through rescanning and data fusion. At the same time, it regularly checks and updates the map to ensure timeliness. During the update process, an incremental update algorithm is used to update only the changed areas to reduce data transmission and processing. An efficient storage structure is designed, and distributed databases or cloud storage are used to store map data in a classified manner. A backup and recovery mechanism is established, and strict data access rights and security policies are formulated to ensure map data security.

[0034] The communication and control module uses 5G, Wi-Fi and Bluetooth communication technologies to build a redundant network. 5G realizes long-distance high-speed communication, Wi-Fi provides local stable network support, and Bluetooth is used for short-range device communication. Data encryption technology is used to ensure data security, establish a link monitoring mechanism, monitor communication signal strength, data transmission rate, bit error rate and other indicators in real time, and automatically switch links when the signal is poor. Based on the robot kinematics and dynamics model, the model predictive control algorithm is used to accurately control the robot movement. Combined with the reinforcement learning algorithm, the robot can make intelligent decisions based on environmental changes and task requirements. A monitoring interface is developed on the remote console and the on-site control center to display various types of robot information in real time, realize remote fault diagnosis and early warning, and support remote operation and control;

[0035] The power management module uses a high-precision battery monitoring chip to monitor battery voltage, current, temperature and other parameters in real time, accurately calculate the remaining power (SOC) and health state (SOH), establish a battery performance model to predict the driving range, and use battery balancing technology to balance the charging and discharging of each battery in the battery pack to improve the overall performance and service life of the battery pack. According to the battery status and task requirements, an intelligent charging strategy is formulated. When the battery is low and the robot is idle, the fast charging mode is used first; when the battery is close to full charge or about to perform a task, the trickle charging mode is used. During the deceleration or braking of the robot, the kinetic energy is converted into electrical energy storage through the energy recovery device to improve energy utilization efficiency. A thermal management system is designed to adjust the temperature of the battery and power module through heat sinks, fans, liquid cooling, etc. to ensure that they work at an appropriate temperature.

[0036] Navigation method of the navigation system:

[0037] Initial positioning and calibration: When the robot is put into use for the first time or enters a new train inspection area, initial positioning and calibration are performed. After the robot is started, the laser sensor begins to scan the surrounding environment and matches the pre-stored map data for positioning. At the same time, the magnetic sensor reads the magnetic stripe information, the IMU measures the initial posture, and the wheel odometer records the initial motion state. The positioning results are calibrated and corrected. In harsh environments such as dim light and dust, multiple landmarks with coded information are pre-set in the train inspection area. When the landmark enters the robot's visual range, the robot visually identifies the landmark and combines its position information to assist in the calibration of the laser sensor's positioning results to improve the accuracy of initial positioning;

[0038] Real-time positioning and path tracking: During the inspection process, the multi-sensor fusion positioning module continuously collects sensor data at a certain frequency (such as multiple times per second) to update the robot's location information. The intelligent path planning module continuously plans new paths based on the real-time position and target point. The communication and control module controls the robot's motor operation, adjusts the speed and steering angle based on the path planning results, and makes the robot drive along the planned path. During the driving process, the robot's posture and position deviation are monitored in real time through sensors. When the deviation exceeds the allowable range (such as the position deviation exceeds ±5 cm, the posture deviation exceeds ±3 degrees), the communication and control module sends an adjustment instruction to return the robot to the correct path;

[0039] In the scenario where multiple robots are working at the same time, each robot broadcasts its position, speed and path planning information through the wireless communication network at regular intervals (such as 1 second). The central server or distributed algorithm determines whether there is a path conflict based on this information. If a conflict is found, the path planning of the relevant robots is adjusted according to the preset conflict resolution rules to ensure that the robots maintain a safe distance and avoid collisions. When the robot encounters an emergency (such as an obstacle suddenly appears in front), the emergency braking system is immediately activated to stop the movement. Then, the surrounding environment is re-detected using sensors. The intelligent path planning module re-plans a safe path based on the new environmental information, and the communication and control module controls the robot to drive along the new path.

[0040] Map update and path replanning: During the robot's inspection process, laser sensors and other sensors continuously monitor environmental changes. When changes in the environment are detected (such as new obstacles, changes in train positions, etc.), the map management and update module immediately starts the map update program. If the update is not timely or inaccurate, the robot continues to monitor the environment using sensors. When the difference between the actual environment and the map reaches a certain level (such as more than a certain number of unknown obstacles are found or the deviation of the position of key landmarks exceeds the allowable range), the robot suspends the task, takes the current position as the starting point, and uses the laser radar to reconstruct the local map within a certain range (such as an area with a radius of 5 meters). The intelligent path planning module replans the path according to the new map, and the communication and control module controls the robot to continue the inspection task along the new path;

[0041] Cross-region navigation and map switching: When the robot needs to switch inside or outside the inspection channel, the map management and update module automatically switches the corresponding map (in-channel map or out-channel map) according to the robot's position and direction of movement. During cross-region navigation, when the robot approaches special terrain such as slopes, the adaptive suspension system is adjusted in advance to increase vehicle stability. At the same time, anti-skid tires are used to improve the vehicle's grip. At the moment of map switching, the IMU and wheel odometer data are used for transition to ensure the robot's positioning continuity, so that the robot can smoothly pass through transition areas such as slopes and realize autonomous navigation and transportation inside and outside the inspection channel.

[0042] Charging management and task scheduling: The power management module monitors the power status of the robot at certain time intervals (such as every minute). When the power is lower than the set threshold (such as 20%), the communication and control module controls the robot to suspend the inspection task, and the intelligent path planning module plans the path to the nearest charging pile. The communication and control module controls the robot to charge according to the planned path. When multiple robots need to charge at the same time, a charging pile allocation strategy combining priority and first-come-first-served is adopted. For example, robots performing emergency inspection tasks are given a higher charging priority; for non-emergency task robots that request charging at the same time, they are charged in the order of request time;

[0043] Task scheduling dynamically adjusts the task priority based on factors such as the urgency of the inspection task, the remaining power and location of the robot. For example, inspection tasks for trains that are about to be put into operation are set to high priority; for robots with lower remaining power, they are given priority to perform inspection tasks that are closer to the charging piles. After the robot is charged, the intelligent path planning module re-plans the inspection path according to the task instructions issued by the remote console or the on-site control center, and the robot continues to perform the inspection task.

[0044] Although the embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and are not to be construed as limitations of the present invention. A person skilled in the art may alter, modify, replace and modify the above embodiments within the scope of the present invention.

Claims

1. A navigation system for a subway vehicle inspection robot, characterized in that: include: The multi-sensor fusion positioning module integrates laser sensors, magnetic sensors, IMUs and wheel odometers. It uses an algorithm combining extended Kalman filtering and particle filtering to fuse data and has sensor fault diagnosis and fault tolerance mechanisms. Intelligent path planning module, which adopts hierarchical path planning strategy, combines A-star algorithm, Dijkstra algorithm, dynamic window method or fast exploration random tree algorithm, considers robot kinematics, dynamics, energy consumption and time constraints for path planning and optimization, and has dynamic adjustment and coordination mechanism; The map management and update module uses two-dimensional laser radar combined with deep learning algorithms to build semantic maps, adopts an update strategy that combines event-driven and time-driven, and has an incremental update algorithm and efficient data storage and management mechanism; The communication and control module uses 5G, Wi-Fi and Bluetooth communication technologies to build a redundant network, uses model predictive control algorithms combined with reinforcement learning algorithms for precise control, and has remote monitoring and fault diagnosis functions; The power management module uses a high-precision battery monitoring chip to monitor battery parameters, and has intelligent charging strategies and energy recovery functions, as well as a power system thermal management system.

2. The navigation system of the subway vehicle inspection robot according to claim 1, characterized in that: In the multi-sensor fusion positioning module, the laser sensor is installed on the top of the robot, the magnetic sensors are symmetrically distributed on both sides of the bottom of the robot, the IMU is installed at the center of gravity of the robot, and the wheel odometer is tightly connected to the driving wheel of the robot.

3. The navigation system of the subway vehicle inspection robot according to claim 1, characterized in that: In the map management and update module, a multi-resolution map construction technology is adopted to dynamically adjust the map resolution according to the robot's motion range and accuracy requirements.

4. The navigation system of the subway vehicle inspection robot according to claim 1, characterized in that: In the communication and control module, data encryption technology is used to encrypt the transmission data, and the communication link monitoring mechanism monitors indicators such as communication signal strength, data transmission rate and bit error rate in real time.

5. The navigation system of the subway vehicle inspection robot according to claim 1, characterized in that: In the power management module, a battery balancing technology is used to balance the charging and discharging of each battery in the battery pack.

6. A navigation method for a subway vehicle inspection robot based on the navigation system according to any one of claims 1 to 5, characterized in that: The following steps are involved: S1: Initial positioning and calibration: After the robot is started, the multi-sensor fusion positioning module obtains the initial position information, and locates through laser sensor scanning combined with map data matching, and calibrates and corrects the magnetic sensor, IMU and wheel odometer; when the laser sensor is affected by harsh environment, the pre-set landmarks with coded information are used for visual positioning auxiliary calibration; S2: Real-time positioning and path tracking: During the inspection process, the multi-sensor fusion positioning module continuously collects data to update the location information, the intelligent path planning module plans the path according to the real-time position and target point, and the communication and control module controls the robot movement and tracks the path; when multiple robots are working at the same time, the global path coordination mechanism is used to avoid path conflicts; in case of sudden emergency, the robot immediately brakes and replans a safe path; S3: Map update and path replanning: When the robot detects changes in the environment, the map management and update module updates the map; if the update is not timely or inaccurate, the robot uses sensors to monitor the environment. If there is a large difference, the task is suspended, the map is partially rebuilt, and the path is replanned; S4: Cross-region navigation and map switching: When the robot switches inside or outside the train inspection channel, the map management and update module automatically switches the map; During cross-region navigation, when the robot is driving on special terrain, it uses an adaptive suspension system and anti-skid tires to ensure stable driving. When the map is switched, it uses IMU and wheel odometer data to transition and maintain positioning continuity. S5: Charging management and task scheduling: The power management module monitors the power. When the power is low, the communication and control module controls the robot to suspend the task and go to charge. When multiple robots need to charge at the same time, a charging pile allocation strategy combining priority and first-come-first-served is adopted. Task scheduling dynamically adjusts task priorities based on factors such as the urgency of the inspection task, the robot's remaining battery power and location.

Images