Research On Behavior Decision Method Of Composite Robot

In recent years,due to the advances in the fields of computer,communication,and machinery.The shape,function and application level of robot have been greatly developed and improved.The composite robot combines a robotic arm with a moving chassis for good handling and mobility.It can perform a variety of tasks at the same time,is widely used in industrial production,family services,logistics and transportation and anti-terrorist explosive ejection and other fields.The decision making of composite behavior method is the main research field of composite robot.It includes behavior planning,path planning and trajectory planning of manipulator.The precision,stability and safety of the composite robot can be improved by good planning.The composite robot studied in this paper is composed of a seven-dof manipulator and an omnidirectional moving chassis with four Mecanum wheels.It is a hyper-redundant robotic system with 10 degrees of freedom.The characteristic of hyperredundancy makes the decision of its behavior method difficult.Therefore,the main research contents of this paper are as follows:(1)The physical structure of the composite robot was designed and the kinematics model of the robot was established,including the forward and inverse kinematics models of the redundant manipulator and the moving chassis.Aiming at the problem of solving the inverse kinematics of the seven-degree-of-freedom manipulator,an approach based on parametric method is proposed.However,this method has the problem that complex solutions are easy to find with the specified parameter .Therefore,a method based on the arm Angle is proposed to realize the complete control of arm Angle.(2)According to the characteristics of asynchronous,coupling and restriction of robot actions,the Petri net behavior model of composite robot with forbidden arc and inert transition was established by taking the tasks of grasping and obstacle avoidance performed by composite robot as an example,and the important modules were analyzed and explained to realize the ability of autonomous behavior planning of robot.(3)The path planning of composite robot was studied,and the path planning of robot was realized by using A* algorithm and RRT algorithm.For RRT algorithm extended slowly,is not sensitive to small area and produce a disadvantage of low path availability,in this paper,the RRT algorithm is improved,and the map is divided into multiple regions,according to the concentration area node sampling probability distribution region,improve the excessive exploration of local space RRT algorithm,obviously improves the search speed of the algorithm.The method of node culling is used to shorten the path distance and improve the availability of the path.(4)The classical trajectory planning of mechanical arm in cartesian space and joint space is expounded,and the quadratic b-spline curve is applied to the cartesian space trajectory planning.The random search algorithm is used to optimize the trajectory of the seven-dof manipulator.In this paper,the characteristics of the population change of differential evolution algorithm in the iterative process are analyzed,and the factor k in the DE/current-tobest/1/bin strategy in the differential evolution algorithm is optimized by using the Beetle Antennae Search Algorithm principle,and the hybrid algorithm is formed by combining with the DE/rand/1/bin strategy,which has the characteristics of high solution accuracy,little evolutionary algebra and strong robustness.The improved algorithm is applied to the trajectory optimization of the seven-degree-of-freedom manipulator,and the singleobjective optimization of the redundant manipulator trajectory is realized.For seven degrees of freedom mechanical arm Lagrange mechanics model is more complex,result in optimization model of the problem of large amount of calculation,using the simulation of curved surface of joint torque load torque model and the NSGA-? algorithm quickly,seven degrees of freedom mechanical arm joint trajectory for multi-objective optimization,the ideal results have been achieved in the simulation test.