Research On Modeling And Control Strategies For Excavation Robots

The enhanced intelligence and automation technology of excavators has resulted in a wider range of application fields.Its significance is progressively escalating,particularly in the domains of water management,infrastructure,hydropower,and emergency response and mitigation.The aim of investigating the trajectory tracking of excavator working devices is to achieve independent excavation operations in challenging conditions,resulting in decreased manual involvement,enhanced operational stability and precision,and increased work productivity.The present study focuses on the PC1012 excavator as the subject of investigation.Specifically,the research involves the development of a kinematics and hydraulic system model for the excavator’s working device.The reliability of the hydraulic system model is subsequently verified through simulation.The present study proposes an improved sparrow algorithm to address the issues and inadequacies in the trajectory tracking process of excavators.The algorithm is utilised to optimise the quintic polynomial trajectory planning.Additionally,the cuckoo search algorithm is employed to optimise the RBF neural network PID control methods.(1)The conventional approach to quintic polynomial trajectory planning solely accounts for position and velocity constraints,while disregarding dynamic constraints and temporal considerations.The present study suggests the utilisation of an enhanced sparrow algorithm for the purpose of optimising the quintic polynomial trajectory planning.The present study aims to incorporate the dynamic constraints of the excavator and optimise the angular trajectory of each joint of its working device using the particle swarm optimisation algorithm.The ultimate objective is to achieve time optimisation.A trajectory with a reduced running time is obtained in comparison to trajectory planning.(2)Due to the multi-variable and strong coupling nature of the excavator working device system,the conventional PID control strategy exhibits inadequate control accuracy and slow response speed in track tracking control.Furthermore,it lacks the ability to dynamically control during the control process.The issue pertaining to the optimisation of parameters.This paper presents a novel approach for PID control utilising a radial basis function(RBF)neural network.The proposed method employs a learning algorithm to automatically adjust the PID parameters,resulting in improved precision compared to traditional PID control.One potential drawback is the necessity for frequent parameter adjustments,which increases the risk of converging to a suboptimal solution.Consequently,the development of a cuckoo search algorithm has been undertaken with the aim of optimising the performance of a radial basis function neural network proportional-integral-derivative controller.The attainment of the optimal solution is facilitated by employing a global search technique and an adaptive adjustment strategy,thereby circumventing the possibility of arriving at a suboptimal solution and dynamically modifying the control parameters.Simultaneously,the excavation trajectory is monitored,and the findings indicate that the optimised controller enhances the precision of the operational system,resulting in a substantial improvement in the trajectory tracking effect.