Research On Error Compensation Method Of The Salvage Robot

The salvage robot studied in this paper is mainly used to grab high-value sleeping floats.Due to its own structure,dynamic characteristics,environment and other factors,the accuracy of its automated grabbing is difficult to guarantee.Therefore,this article focuses on the impact on its grabbing accuracy The main factor is to conduct research to improve its grasping accuracy.This article takes the 8-degree-of-freedom manipulator of the salvage robot system as the research object,and aims at improving the pose accuracy.The error model and parameter identification and compensation methods are studied,including its forward kinematics model and its static structural parameters.Error modeling is based on the establishment of an elastic deformation error model based on the flexibility of the boom,the establishment of a dynamic model and its modal analysis,parameter identification method and its comprehensive error compensation method.This article analyzes the source of the error of the salvage robot,studies the kinematics of the salvage robot and its kinematics error model,based on the forward kinematics theory and micro-motion theory of the salvage robot,and improves the deficiencies described by the traditional DH parameters,The MDH model is introduced and the MCPC model is referenced to establish a complete static error model from the base to the terminal operator.Then according to the expression of the error model and the Jacobian matrix theory,the error model is generalized to establish its generalized error model.The flexibility of the salvage robot is established as the main influencing factor,and the flexible error model of the salvage robot under external force and its own gravity is established.Based on the flexibility of the boom,the Hamilton principle is used to establish the rigid-flexible coupling dynamic model of the salvage robot.Through this model,the dynamic model of the salvage robot is obtained,and the modal analysis of the salvage robot is carried out,and the modal of the main boom that affects the grabbing process is analyzed.Because the salvage robot has too many degrees of freedom,it is difficult to directly solve the error model.The particle swarm algorithm is used to study the parameter identification of the salvage robot.In the research process,the optimization problem is prone to multiple solutions.The error parameters of the robot are grouped.By simplifying an optimization problem with a large number of parameters into several optimization problems with fewer parameters,the accuracy,speed,and stability of the identification are improved.For the other influencing factors of the salvage robot,to avoid establishing more models,and the excessive model,calculation is complex,it is difficult to implement compensation,the RBF algorithm is used to study the comprehensive error compensation,using the theoretical pose as input and actual The pose is used as the output to train the RBF neural network,and a network that can predict the actual pose through the theoretical pose is obtained,and then the error compensation theory is used to perform comprehensive error compensation.