The Intelligent Control Of Internet-based Teleoperation Pipeline Robot

At present, the pipeline has been widely used and has played a huge role in the fields ofpetroleum, chemical industry, city water supply, ect. In order to improve the life of thepipeline and avoid accidents, periodic detection is very essential which draws more and moreattention of engineering personnel. This provides the opportunity for the research andapplication of pipeline robot. The pipeline robot mostly works in the environment wherepeople hardly reach and very complex, so it requires strict control of the pipeline robot. Withthe rapid development of teleoperation technology and network, the combination ofteleoperation technology and internet provides a new method for the control method, andexpands the application fields of teleoperation technology as well as human operation ability.Therefore, applying the teleoperation technology to implement remote control of the pipelinerobot and do timely detection and maintenance to avoid accidents is feasible. The papermainly includes following aspects:This paper mainly researches on the network time delay in internet-basedcommunication of teleoperation system. Firstly, it introduces the teleoperation technology andthe research status, and summarizes the control methods of solving the delay problem incontrol areas. On the basis, the paper introduces the calculation methods of the delay of thenetwork transmission and the parts of delay. The paper also tests the delay by using the testprocedure of network delay, and analyses the delay in details on the basis of test results.Secondly, as the delay in the network transmission process will lead to instability of thesystem, the delay is the key problem of the control system and the operator can predict thedelay by using the neural network. The paper highlights the GMDH network used in theforecast area. GMDH network has such an advantage of being able to establish the networkstructure dynamically that it leads to very good predictive effects. Then the paper introducesthe Smith predictive control which can be used in large delay system and the improvementprogram. The ordinary regulator used in Smith Predictive control is generally the PIDcontroller. According to the three adjustable parameters of PID controller, the paper adjuststhe parameters by using BP neural network on line, and verifies the effectiveness of thismethod through concrete examples.Finally, the paper introduces the model of the master-slave teleoperation control system and brings up the control system which is based on the neural network and the modifiedSmith predictor. Through theoretical analysis and simulation, the paper shows that thismethod can improve the stability and dynamic quality of the control system.