Research On Reliability Design And Security Of Wheeled Service Robot Control System

The urgent needs of society and the rapid development of technologies such as computers,artificial intelligence,and sensors have greatly promoted the process of R&D and industrialization of service robots.At present,service robots play an increasingly important role in the areas of cleaning,care,security and rehabilitation.However,the low level of intelligence,low reliability,and low security have greatly hindered the large-scale deployment of service robots.Service robot hardware and software reliability design and system fault self-diagnosis is an important means to improve product safety.Therefore,this paper designs and researches the reliability and fault self-diagnosis of a new generation of wheeled cloud-based intelligent escort robot bottom control system developed by the laboratory.First of all,from the perspective of reliability and security,the underlying control system of the intelligent companion robot is designed.Based on the high reliability design of the hardware,the reliability design of multiple monitoring software,communication software and motion control software in the control system was emphasized.And completed the reliability test of the entire underlying control system,providing an experimental platform for fault self-diagnosis research.Then,for the problem that the optimal parameters in the SVM failure prediction model are difficult to determine,an improved particle swarm optimization algorithm based on Sigmoid inertia weight reduction strategy is proposed to optimize the parameters of the SVM model,and an SVM based on the improved PSO optimization is established.Prediction model.The simulation results of single-peak and multi-peak test functions show that the improved PSO algorithm presented in this paper has faster convergence speed and higher accuracy;The artificial data set training experiment validates the effectiveness of the improved PSO-SVM prediction model.Finally,taking the encoder fault in the close-loop motion control system of the companion robot as an example,the fault diagnosis and fault-tolerant control of the robot are studied.Based on the improved PSO-SVM and residual decision method,a data-driven fault self-diagnosis and fault-tolerant control system is established.Taking the simulation platform and the accompanying robot system built in the laboratory as the experimental platform,it provides data sets for fault diagnosis research and establishes a PSO-SVM prediction model.Based on the residual signal obtained by the experiment,the SPRT test method is used to detect the fault.When the fault occurs,the hardware is closed-loop to the software closed-loop to fault-tolerant.Simulation experiments show that this method can quickly and accurately detect sensor failures,achieve closed-loop control switching,and greatly improve the reliability and safety of the control system.