AGV as a logistics intelligent robot has a history of more than 60 years and has been widely used. The AGV is equipped with an electromagnetic or optical automatic guiding device, which can travel along a designated guiding path to meet the requirements of transportation and production automation, reduce labor costs, and improve production efficiency. However, during the driving process, the car may encounter obstacles in front, which may affect the driving and even cause a collision. Therefore, providing an automatic obstacle avoidance method can improve production efficiency and reduce the staying time due to obstacles. At present, a multi-sensor-based outdoor automatic obstacle avoidance AGV navigation method is disclosed, which includes the following steps: calculating a shorter path according to the local path plan and the target starting point and ending point; the lidar module is used to detect the surrounding environment and avoid obstacles; The correct driving direction of the road is compared with the direction angle of the current heading obtained by the electronic compass, and the car driving direction correction angle 1 is obtained; the camera module is used to identify the road marking line, and the car driving direction correction angle 2 is obtained through analysis. Process 1 and 2 to get ideal angles in different environments. The industrial computer processes the relevant parameters, makes the trolley move forward through the wireless module and the drive module, and at the same time coordinates the planning and detection through the coordinator.

Aiming at the shortcomings of automatic guided vehicles encountering obstacles in front during driving, technicians provide an AGV that can accurately avoid obstacles and bypass obstacles when there are obstacles. Methods. In this method, the laser sensor is set up, combined with the short-term memory method, and the obstacle distribution of the scene around the AGV is collected in real time through the laser sensor. The obstacle avoidance algorithm is designed and the obstacle avoidance strategy is formulated according to the collected obstacle distribution, and the AGV is controlled to find a feasible channel and bypass Go over obstacles to reach the target location to achieve autonomous obstacle avoidance. The laser sensor emits a laser beam to obtain the distribution information of surrounding obstacles. Taking a fan-shaped area with a radius of 5m and 180 degrees in front of the laser sensor as the detection range, the fan-shaped area is simplified into a rectangular area to reduce the amount of calculation. The data collected by the laser sensor is constantly changing, so that the rectangular area can be expanded. When updating a frame of data, the grid data is selectively deleted according to the current coordinate offset, and new data is added to the grid, and the grid data is constantly updated during the movement to obtain the obstacle distribution around the AGV. The laser sensor obtains the polar coordinates of the obstacle distribution in the surrounding scene, and obtains the Cartesian coordinates (x, y) according to the polar coordinates, and then calculates the obstacle distribution in the rectangular area. If there is no obstacle, write 0; if there is an obstacle, write 1. Designed obstacle avoidance algorithm and formulated obstacle avoidance strategy. First determine the boundary scene, input a proportional field route map to ensure that the AGV travels within the boundary wall; then update the grid according to the navigation sensor and obstacle avoidance laser. The navigation laser sensor obtains the current AGV coordinates through the reflector and the principle of triangulation, performs coordinate transformation, updates the grid, and adds new obstacles to the grid; to find alternative channels, consider the body radius R1 and the smaller obstacles and the body The distance d1 extends the size of the obstacle to R1 d1, divides the grid within 360 degrees into 360/sectors, uses a fixed angle as the unit sector, traverses all sectors, and all sectors blocked by obstacles are Alternative channel: Find a reasonable channel, remove the infeasible channel from the candidate channels, and construct a cost function in the remaining feasible channels to select the better channel.

The outstanding effect of the above-mentioned AGV autonomous obstacle avoidance method is: by combining the short-term memory method, the laser sensor collects the obstacle distribution in the scene around the AGV in real time, and designs the obstacle avoidance algorithm and formulates the obstacle avoidance strategy according to the collected obstacle distribution. Control the AGV to autonomously find a feasible channel, bypass obstacles to reach the target location, so as to achieve autonomous obstacle avoidance. This method for AGV autonomous obstacle avoidance can automatically avoid obstacles and operate without obstacles in a complex environment, reducing the possibility of accidents, and greatly improving the work efficiency of AGVs by selecting reasonable channels.