Font Size: a A A

Research On Issues Of Mill Position Servo State Feedback Steady State Error And State Acquisition

Posted on:2015-05-24Degree:MasterType:Thesis
Country:ChinaCandidate:S LiuFull Text:PDF
GTID:2181330422470497Subject:Mechanical and electrical engineering
Abstract/Summary:PDF Full Text Request
Steel is the major industrial materials in areas of today’s production, strip productsaccount for about half of total production, thickness control accuracy is one of the mainindicators of the quality of strip.Improve the mill hydraulic servo control system isimportant for improving thickness accuracy. Because the system description and controllerdesign makes use of incomplete information basing on output feedback of mill hydraulicservo system limited control optimization, so it need to improve the control performancebased on state feedback. However, at present state feedback applied to the mill hydraulicservo system still have problems of steady state error and status difficult toobtain.Although using system output correct and compensate state feedback system caneliminate steady-state error, it can effect dynamic performance; Although using observercan observe observability of the system state, it is difficult to meet rapidity requirementsof feedback state when the system is not certain.To solve the problem of difficult to eliminate steady-state error, at first analyze theroot cause of steady-state error based on the state feedback in the mill hydraulic servosystem, and obtain steady-state error condition existed in state feedback. Correctinghydraulic servo system model and designing online identification algorithm eliminate thesteady state error, and deduced identification algorithm convergence conditions.To solve the problem that feedback state is difficult to obtain, rebuild the model intwo ways: First, the servo valve spool displacement and cylinder displacement can bedetected as state variable reconstruct model to get the second-order state space expression;Second, the servo valve spool displacement pressure and cylinder displacement can bedetected as state variable reconstruct mode to obtain higher accuracy third-order statespace expression.Experiments were conducted on300rolling mill. From the experimental curves,it canbe drawn that online identification can eliminate the steady-state error in the system, andtake into account the stability and speed of the system. From the thickness differencecurves, it can be drawn that the second-order system and third-order system whose state variable can be detected roll out the strip after pole assignment is better than state observerbased on pole placement, and can replace the state observer system, so that the statefeedback system with uncertainties quick access issues is solved.In a word, online identification principles apply to System gain recognition toeliminate steady state error, and ensure the speed and stability of the original system. Onthe basis of tuning system each element control performance, depressed position servosystem is remoded after reasonable simplification,and ensure the selected feedback statevariables can be detected, and achieve real-time acquisition of feed-back state under theuncertain system. Experiments conducted on300rolling mill verify the effectiveness ofthe control strategy.
Keywords/Search Tags:AGC hydraulic, state feedback, identification, steady state error, condition detection
PDF Full Text Request
Related items