Research And Test Of Anti-collision System For Intelligent Forklift

As a big country in the world,with the continuous development of social economy,China’s various types of material storage capacity and throughput are getting bigger and bigger.Under the background of the global intelligent era,China proposes the"China Manufacturing 2025"national development strategy plan,all of which creates a good environment for the research on intelligent vehicles and is the best opportunity to promote intelligent and efficient warehouse management.At any time,however,we can not ignore safety,safety as the primary security of production and development,especially should be paid attention to in the process of the application of the intelligent vehicle,so this paper studies the anti-collision system with two key technology of obstacle detection and obstacle avoidance path planning in warehouse,and achieved some valuable results.In the first subject of this research,the focus is on the problem of anti-collision detection for short-distance obstacles.This paper proposed a dynamic detection method of obstacles based on speed and steering angle for intelligent forklift,by using the vehicle-mounted laser sensor to obtain real-time information of the body posture and surrounding information,and combining with the motion geometric models of forklift,it was formed that a dynamic detection method of obstacles based on the double-plane fusion of the scanning planes of the horizontal and tilted laser ranging sensor,which made the detection region of the intelligent forklift changed dynamically with the speed and steering angle.To verify the effectiveness and superiority of the proposed method,the detection methods based on horizontal scan ranging sensor and tilted scan ranging sensor were tested respectively,and trial tested the speed and steering angle respectively on the effects of obstacle detection,at the same time,2 traditional methods,the sector method and the rectangle method,were selected for contrast tests.Through actual driving,this proposed method was in line with the actual situation,while the traditional methods had false detection.With the horizontal scan ranging sensor as the main one and the tilted scan as the auxiliary one,it could detect obstacles about 31 mm minimum height and effectively solve the problem of false obstacle detection of intelligent forklift in the warehouse,and was more suitable for warehousing and transportation than the traditional detection method,thus improving the mobility and safety of intelligent forklift in the warehouse.In the second subject,the core of this research focuses on the obstacle avoidance path planning for longer-distance obstacles.In order to avoid the forklift’s frequent parking affecting work efficiency and prevent false alarm when the autonomous obstacle avoidance,this paper combined the characteristics of the forklift operating environment,and used laser navigation sensor to obtain the forklift pose information and laser ranging sensor to detect the driving environment,a dynamic identification zone that the detection range changed along with the speed and the steering angle was set up to obtain information of longer-distance obstacles,and 9 control points were selected according to the information.Taking the feature control point as the segmentation point and combining the properties of the B-spline curve,the B-spline path through some data points was generated by combining the front and rear segments of quartic five-order quasi-uniform B-spline curves,so as to ensure the continuity of the whole path and eliminate the step of inverting the control points.At the same time,because of the flat ground,the global path was composed of straight road sections and turning road sections in the structured warehouse,so the obstacles emergent situation could be divided into two types that obstacles encountered in the straight road or in the turning.Therefore,this research carried out the obstacle avoidance path planning tests on the straight road section and the turning in the warehouse,the results showed that the obstacle avoidance path curvature was not greater than 1.06×10^-3 mm^-1,steering angle was not greater than 60°,angular velocity of equivalent steering wheel was not greater than1.05 rad/s,which satisfied the forklift obstacle avoidance constraint,the minimum turning radius constraint,the curvature continuous constraint,the maximum steering angle and the maximum angular velocity of steering wheel constraint,and the starting and ending point pose constraint.The feasibility of the proposed algorithm was verified.The research in this paper can provide reference for obstacle detection methods and obstacle avoidance path planning methods for large-sized warehouse intelligent transportation vehicles,which is conducive to promoting the intelligent development of forklifts and improving freight efficiency.