Multi-machine collaborative operation cabin cleaning method

Abstract

According to the innovative port cabin cleaning scheme, multi-machine cooperative operation of the unmanned excavator and the ship unloader is utilized, and the operation safety and efficiency are remarkably improved. The system integrates multi-sensor mapping, path planning, intelligent collision avoidance and real-time position tracking technologies, and elevation grid map modeling and vehicle positioning in a cabin are realized through a mapping sensing module. The task scheduling module intelligently selects an operation point and plans a path according to the coal pile layout, and the cooperative control module monitors cooperative work of the unmanned excavator and the ship unloader in real time to prevent the collision risk. The work control module is responsible for dynamically adjusting the equipment state and ensuring safe and efficient operation. The scheme flexibly adapts to various ship types and cargo hold layouts, not only improves the port operation efficiency, but also reduces the risk of manual operation errors, and is an important step for modern port automation and intelligent development.

Description

Technical Field

[0001] The present invention relates to the field of intelligent vehicles and port ship operation machinery, and in particular to a method for cleaning cargo holds by collaborative operation of multiple machines. The method combines the joint operation of an unmanned excavator and a ship unloader, improves the safety and efficiency of cargo hold cleaning operations through intelligent path planning, real-time position tracking and collision avoidance algorithms, and belongs to the category of port automation technology. Background Art

[0002] In the process of coal loading and unloading in modern ports, tank cleaning is a key task. Traditional tank cleaning methods mainly rely on manual operation, which is inefficient and has high safety hazards. Due to the narrow space in the cabin and the complex distribution of coal, mechanical equipment often faces inconvenient operation and potential collision risks during tank cleaning. In recent years, the trend of port automation has accelerated, but the existing automated tank cleaning technology is usually unable to operate effectively in complex environments, especially when multiple mechanical equipment are operating at the same time, the collision avoidance between different equipment needs to be considered, and coordination and cooperation are difficult. Traditional multi-machine collaborative algorithms such as centralized collaborative algorithms and algorithms based on swarm intelligence are widely used in robots, drones, self-driving cars and other fields, but in the specific scenario of port tank cleaning, the application of these algorithms still faces some challenges. First, the environment of coal tank cleaning is complex and changeable, the space in the cabin is narrow and the coal is piled irregularly, which increases the difficulty of equipment perception and path planning. Secondly, the physical properties of coal will limit sensors, especially in high dust environments, traditional visual sensors may fail. Finally, the excavator has a large operating range and turning radius, and the yaw angle changes significantly. When the excavator and the ship unloader operate at the same time in the same small space, it is a difficult problem to avoid collision and ensure efficient task allocation.

[0003] Therefore, the existing multi-machine collaborative algorithms need to be improved according to the needs of actual port operations. For example, in complex dynamic environments, real-time perception and collaborative algorithms may be more appropriate, allowing each device to make real-time decisions based on its own perception and share information with other devices through low-latency communication networks. This can greatly improve the flexibility and robustness of the system. Summary of the invention

[0004] The purpose of the present invention is to overcome the shortcomings of the prior art and propose a method for clearing the cargo hold based on the coordinated operation of multiple machines. Through the coordination of unmanned excavators and ship unloaders, the operation efficiency and safety are significantly improved. The solution aims to solve the problems of complex operating environment and difficulty in cooperation between equipment during coal cargo hold clearing. It adopts intelligent path planning, real-time position tracking and intelligent collision avoidance algorithm to ensure efficient cooperation between equipment, reduce human intervention, and reduce the risk of collision and accidents. Its main designs include:

[0005] Path planning and operation point selection: The present invention designs an intelligent path planning algorithm, and presets multiple operation points according to the ship loading situation, cargo hold layout and coal distribution. The system automatically selects the current optimal operation point through real-time data analysis to ensure efficient operation of the equipment and reduce unnecessary movement paths. The selection of each operation point takes into account the operation range, operation priority and integrity of cargo hold cleaning of the ship unloader and unmanned excavator.

[0006] Intelligent collision avoidance and dynamic path adjustment: In order to ensure the safety between devices, the present invention designs an intelligent collision avoidance algorithm based on multi-sensor fusion and path planning optimization. First, the system models the environment inside the cabin by integrating multiple sensors such as LiDAR, ultrasonic sensors and IMU, obtains the internal boundary of the cabin, and tracks the precise position, speed and direction information of the equipment in real time based on multiple sensors, so as to perceive the surrounding environment in all directions. Secondly, based on the kinematic model of the equipment, the system predicts the motion trajectory of the two devices and evaluates the potential collision risk. When it is detected that the equipment may collide, the system starts the emergency collision avoidance mechanism and adjusts the speed of the equipment through the PID controller to ensure smooth avoidance. In order to avoid the impact of collision avoidance on the progress of the operation, the system adopts the A* path planning algorithm combined with the B-spline curve to achieve smooth path planning, so that the equipment can stably turn and go straight in the cabin environment, and can also quickly resume efficient operation after avoidance. In addition, the control processing module is used to analyze the position information and path planning results of the cleaning machine to ensure the smooth progress of collaborative collision avoidance and operation. The intelligent collision avoidance algorithm combines real-time position tracking and dynamic path adjustment, which can effectively avoid collision risks in complex environments and ensure the continuity and safety of the cleaning task.

[0007] Multi-device collaborative operation mechanism: The present invention proposes an innovative operation mechanism in terms of multi-machine collaboration, ensuring that the unmanned excavator and the ship unloader can cooperate efficiently and form a mode of clear division of labor and collaborative operation. Specifically, the ship unloader is responsible for handling the handling and large-scale cleaning of the main coal in the cabin, and quickly unloading the large pieces of coal in the cargo hold of the ship. The working area of ​​the ship unloader is mainly in the middle part of the cabin; while the unmanned excavator is responsible for cleaning the scattered coal left after the ship unloader operation, especially those areas that are difficult to reach in the bottom of the cabin or corners. The system dynamically monitors the operation progress of each device, the distribution of coal, and the changes in the cabin environment through a task allocation algorithm based on real-time perception and intelligent analysis, and adjusts the work tasks of the unmanned excavator. Specifically, the system can assign cleaning tasks to the unmanned excavator in real time according to the cleaning progress of the ship unloader, the distribution density of coal, and the operation priority, ensuring that it enters the corresponding area for operation at the right time to avoid idling or interference of the equipment. Through this dynamic task allocation mechanism, the system can flexibly cope with the complex operating environment in the cargo hold, maximize equipment utilization, and significantly improve overall operating efficiency.

[0008] System monitoring and dynamic adjustment: During the entire cleaning process, the system continuously monitors the equipment's operating status, task progress, and external environmental information, and adjusts the operation plan in real time. When a task at a certain work point is completed, the system will detect changes in the cabin environment and update the map, and reallocate the task points, thereby ensuring that the automatic cleaning machine can automatically optimize the operation process and ensure that the operation task is completed safely and efficiently. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] Figure 1 It is a module diagram of the multi-machine collaborative tank cleaning method of the present invention;

[0010] Figure 2 It is a work flow chart of the multi-machine collaborative tank cleaning method of the present invention;

[0011] Figure 3 It is an obstacle avoidance and coordination module diagram of the multi-machine coordinated cabin clearing method of the present invention; DETAILED DESCRIPTION

[0012] In order to make the purpose, technical solution and advantages of the present invention clearer, the present invention is further described in detail below with reference to the accompanying drawings. It should be understood that the examples described this time are only used to explain the present invention, not all the embodiments, and are not used to limit the present invention. The embodiments provided are only to disclose the present invention in detail and completely. Based on the embodiments of the present invention, all the premise embodiments obtained by ordinary technicians in this field without making creative work belong to the scope of protection of the present invention.

[0013] The embodiment of the present invention provides a method for the coordinated operation of an unmanned excavator and a ship unloader in a cabin environment. The overall system module diagram of the system is shown in FIG. Figure 1 As shown, the system can be divided into three main modules, namely sensor module, environment perception and positioning module, and task execution module. The following is the definition and description of each module:

[0014] Sensor module 101. It mainly includes IMU, LiDAR, and ultrasonic radar. IMU estimates the posture and movement of the device by measuring the acceleration and angular velocity of the device. It can help the device remain stable in a dynamic environment and provide preliminary location information, especially in indoor environments without GPS. LiDAR can build a three-dimensional or two-dimensional point cloud map of the surrounding environment by emitting laser beams and measuring the return time to help the device perform environmental perception and obstacle detection. It is a key tool for building maps and vehicle positioning. Ultrasonic radar: Ultrasonic radar uses sound waves to detect nearby objects and is suitable for short-range obstacle detection. Due to the long wavelength of ultrasonic waves, it is suitable for detecting dusty environments and provides close-range perception capabilities, especially obstacle avoidance when close to obstacles.

[0015] Environmental perception and positioning module 102. This module mainly includes the functions of realizing the construction of the elevation map of the interior of the cabin and positioning according to the elevation map. At the same time, the working range of the ship unloader is marked on the elevation map for the obstacle avoidance function of the unmanned excavator. The mapping step builds a map of the cabin or working environment by fusing the data of sensors (such as lidar and IMU). This is the basis for subsequent navigation and path planning. The accuracy and completeness of the map directly affect the positioning of the equipment and the performance of the task. Vehicle positioning integrates the data of IMU, lidar and other sensors, so that the equipment can determine its exact position in the working area. This is crucial for the path planning and collaborative work of the equipment to ensure that the equipment moves along the predetermined trajectory. The obstacle perception function is realized by real-time detection of obstacles in the working environment through sensors (lidar, ultrasonic radar). The system will adjust the path planning according to the perceived obstacle information to avoid collisions and accidents.

[0016] Task execution module 103. Collaborative work is that the unmanned excavator and the ship unloader work together in the same environment. They can share environmental information, assign tasks, and perform synchronous operations according to the predetermined work area. Collaborative work requires good communication and division of labor to avoid interference between devices. Path planning is based on the previously constructed map and vehicle positioning information, and the system will generate the best path for each device. These paths take into account the working area of ​​the equipment, task requirements, and the requirements to avoid obstacles to ensure that the equipment can complete the task efficiently. The real-time obstacle avoidance function ensures that the equipment can cope with emergencies in a dynamic environment when performing tasks, such as the sudden appearance of new obstacles. This function relies on real-time data input from sensors to ensure that the equipment can safely avoid obstacles without interfering with the overall task execution.

[0017] Figure 2 This is a flowchart of how each module is executed in an embodiment of the present invention. The process includes multiple steps from environmental scanning, positioning calibration to actual tank cleaning operations. The detailed description of each step in the flowchart is as follows:

[0018] In step S201, the entire tank cleaning operation process is started, and an operator or a system control program issues an instruction to start the entire tank cleaning process.

[0019] In step S202, the laser radar will scan the cabin environment and build a global map. The system scans the cabin environment through sensors (such as laser radar and IMU). The laser radar is mainly used to scan the surrounding depth information and the coal pile information attached to the inner wall of the cabin, and the IMU will record the displacement and offset of the scanning process, thereby creating a global map of the current cabin, providing a basis for subsequent path planning and operations.

[0020] In step S203, since the ground inside the cabin is not necessarily completely level, there will be interference when forming the elevation map, because it is necessary to use the linear acceleration data of the IMU in the initial state to correct the map to help the system calibrate its own position and direction to ensure map accuracy.

[0021] In step S204, since the cabin environment is relatively complex and there is dust or other interference, it is necessary to fuse the millimeter wave radar data to further optimize the environmental map. Millimeter wave radar can penetrate dust or other interference and provide more accurate distance measurement, which is particularly suitable for complex environments such as cabins.

[0022] In step S205, the point cloud data is converted into an elevation grid map after voxel filtering and rasterization, and the space in the cabin is divided into navigable grid areas. At the same time, the point cloud of the current frame is matched with the point cloud of the global map for accurate positioning. This step lays the foundation for subsequent path planning.

[0023] In step S206, the accuracy of the positioning is determined according to the pre-set cabin size specifications. If the positioning is inaccurate, the environmental scanning and map generation steps may need to be repeated. If the positioning is accurate, the next step is entered to ensure the safety of subsequent tasks.

[0024] In step S207, the work area is generated according to the elevation grid map. The system divides the tank cleaning work area according to the converted elevation grid map to ensure that the tank cleaning work can cover the entire cabin space. At the same time, the grid elevation map is uploaded to the host computer for display through the LZ4 compression algorithm. Different work areas will be assigned to the tank cleaning equipment for sequential execution.

[0025] In step S208, on the generated elevation grid map, the A* algorithm uses heuristic search to find the optimal path for the equipment from the starting point to the target point. This path is mainly based on the principle of shortest distance, while taking into account various obstacles inside the cabin (such as the inner wall of the cabin, ship unloader, etc.) to ensure that the cleaning equipment can pass safely. When planning, the A* algorithm will combine the actual size of the equipment to avoid collisions between the equipment and the inner wall of the cabin or the ship unloader. The initial A* path consists of a series of discrete points, and unnatural sharp turns may occur at corners. In order to improve the smoothness of the path, the system will use B-spline curves to post-process the path and generate a smooth curve by interpolation to make the equipment more stable when turning. Collision detection will be performed on each node in the path to ensure the safe movement of the equipment.

[0026] In step S209, collaborative cabin clearing will be achieved. After the path planning is completed, the unmanned excavator and the ship unloader work together according to the planned path and task allocation to gradually clear the cargo in the cabin. Multiple devices need to work synchronously to avoid collisions or interference. Specifically, the unmanned excavator's clearance work is mainly aimed at coal piles at a certain height. First, the front laser radar of the unmanned excavator will scan the coal pile in the front working area. After the straight-through filtering at the height and the downsampling of the point cloud, the working point where excavation needs to be performed is obtained. Then, through the point cloud coordinate conversion, the relative position of the working base point of the unmanned excavator's mechanical arm is obtained. Finally, the mechanical arm arm and forearm are modeled to realize open-loop control.

[0027] In step S210, real-time obstacle avoidance will be achieved. In the actual cleaning work, the unmanned equipment uses sensors to perceive the environment in real time. If new obstacles are detected (such as the unloader grabbing materials or coal pile changes), the path will be dynamically adjusted to avoid collision. Specifically, the system will specify a safe working radius R safe The laser radar installed on the top of the unmanned excavator can scan the environment around the equipment. By filtering the point cloud of the excavator's mechanical arm, it can judge the surrounding obstacles. safe When there are obstacles nearby, the system will stop working and plan the path by specifying a new working point.

[0028] In step S211, the sub-map and the working area are updated. During the cleaning process, the system will update the status of the working area in real time. As the cleaning task progresses, the system will adjust the remaining working area based on the cleaned area. Specifically, when the working task of a certain working area is completed, the relative coordinates of the equipment relative to the cabin at this time will be recorded, and the sub-map of this working area in the global map will be updated.

[0029] In step S212, when the tank cleaning task is completed, the system ends the current operation and returns to the initial state. Equipment initialization and operation point preset.

[0030] Figure 3 The implementation process of the collaborative obstacle avoidance function of the unmanned excavator and the ship unloader was demonstrated. The entire system ensures that the excavator and the ship unloader can effectively avoid obstacles and work efficiently in complex working environments through data processing and collaborative calculations in multiple links.

[0031] The overall process starts with the realization of the obstacle avoidance coordination function, which is mainly divided into autonomous obstacle avoidance of unmanned excavators and obstacle avoidance under the global reference system. The main goal of the system is to ensure that the excavators and ship unloaders can perceive obstacles in the environment in real time and avoid them during operation to improve safety.

[0032] The system first processes the autonomous obstacle avoidance of the unmanned excavator. The unmanned excavator collects point cloud data of the working environment in real time through its own sensors, especially the obstacle data within the operating radius of the excavator. The system uses this data to analyze the position and distance of the obstacle relative to the excavator and calculate the path within the safe range so that the excavator can avoid obstacles and perform normal operations.

[0033] The unmanned excavator obtains point cloud data within the safety radius through LiDAR or other sensors. This data contains the spatial position, shape and distance information of objects in the environment relative to the excavator, providing accurate environmental perception input. By analyzing this data, the system can identify obstacles that may affect the excavation operation and generate safe path planning based on this. The system calculates the relative distance between each obstacle and the excavator based on the collected point cloud data. Through this distance calculation, the system can evaluate the possible collision risk and make appropriate avoidance decisions to ensure that the excavator remains safe during the operation.

[0034] In the global reference system obstacle avoidance collaborative process, the system establishes a global reference system so that the excavator and the ship unloader can operate collaboratively in a shared global coordinate system. The introduction of the global reference system is to ensure effective communication and mutual position perception between the two devices in the same space, so as to achieve more efficient obstacle avoidance planning. The system uses the registration technology of point cloud data to accurately align the point cloud data of the excavator with the data in the global reference system to obtain the accurate position of the ship unloader relative to the excavator. The registration process is based on the 5G signals, sensor data and other location information of the excavator and the ship unloader to achieve accurate positioning of the two in a unified reference system. The ship unloader communicates with the global reference system through 5G signals to realize its own positioning function. The high-precision and low-latency characteristics of 5G signals ensure that the position of the ship unloader can be updated in real time, and share position data with the excavator to form the basis for coordinated cooperation. By registering the excavator and the ship unloader in a unified reference system, the system calculates the global relative distance between the two. This global relative distance is an important parameter for collaborative obstacle avoidance, ensuring that the two can avoid each other in the working space to avoid collision or interference with their respective work processes. Finally, the system achieved the obstacle avoidance coordination function of the unmanned excavator and the ship unloader by accurately positioning the two devices, aligning the point cloud data, and calculating the global distance. The system can continuously monitor the relative position of the two in a dynamic working environment, and adjust and avoid strategies in real time to ensure safe and efficient work.

[0035] It should be emphasized that the implementation cases described in the present invention are illustrative rather than restrictive. Therefore, the present invention includes but is not limited to the embodiments described in the specific implementation plans. Any other similar implementation plans derived by those skilled in the art based on the technical solution of the present invention also fall within the scope of protection of the present invention.

Claims

1. A method for clearing a cargo hold based on multi-machine collaborative operation, including the coordinated cooperation of an unmanned excavator and a ship unloader, characterized in that: Through intelligent path planning algorithms, multiple operation points are preset according to the ship loading situation, cargo hold layout and coal distribution; The system automatically selects the current optimal operating point through real-time data analysis; Using multi-sensor fusion technology, the position, direction and speed of equipment in the cabin can be tracked in real time; Smooth path planning is achieved based on the A* algorithm and B-spline curve, and dynamic path adjustment is performed through the PID controller to avoid equipment collisions and maintain work efficiency.

2. The method according to claim 1, characterized in that: Use multiple sensors such as LiDAR, ultrasonic sensors and IMU to obtain high-precision maps of the cabin's internal environment; Environmental modeling is performed based on multi-sensor fusion, and the relative position of the device and obstacles is updated in real time to ensure that the device can avoid collisions.

3. The method according to claim 2, characterized in that: When the device detects a potential collision risk, the system activates the emergency collision avoidance mechanism; The PID controller adjusts the equipment speed and updates the path in real time to ensure that the equipment can resume operation after avoiding obstacles.

4. A multi-machine collaborative operation mechanism, characterized in that: The ship unloader is responsible for unloading large amounts of coal from the ship's hold, while the unmanned excavator is responsible for cleaning up the coal left in the corners or at the bottom of the hold. The system dynamically allocates the operating tasks of the unmanned excavator according to the operating progress of the ship unloader, coal distribution and cabin cleaning priority.

5. The collaborative operation mechanism according to claim 4, characterized in that: The system monitors the operating status of each device in real time and adjusts the operating plan according to the progress of the task and changes in the external environment; When a task at a certain work point is completed, the system reallocates the work point and updates the map in real time.

6. A tank cleaning system based on the method of claims 1 to 5, comprising: Sensor modules for acquiring environmental data, including IMU, lidar, and ultrasonic sensors; The environmental perception and positioning module is used to generate an elevation map of the cabin based on sensor data and realize equipment positioning by fusing multi-sensor data; The task execution module includes intelligent path planning, real-time obstacle avoidance and dynamic operation point allocation.

Images