| Based on the project supported by national natural science fund project (51175349)and Shenyang science and technology plan fund projects (F12-277-1-70). With the higherand higher machining efficiency and machining precision demand of modern industry, NCmachining technology is toward the direction of high speed and high precision. Based onthe tracking characteristic, robustness and contour error characteristics problems ofpermanent magnet linear synchronous motor (PMLSM) drive XY table system, a zerophase robust control system and predict robust control system are designed. In single shaftzero phase error tracking controller (ZPETC) and model predictive controller(MPC) areused as feedforward controller to improve the tracking performance of the motor. Thedisturbance observer (DOB) is used to improve the robustness of single shaft motor. In thetwin screw, cross-coupling controller (CCC) is used to reduce contour error of system.First of all, according to the widely read of related literature at home and abroad, thecurrent situation of home and abroad about the direct drive XY platform control systemdevelopment present situation and some main control methods are introduced. Theworking principle of PMLSM and mathematical model is introduced. The contour error ofXY table is introduced. The causes of the contour error are analyzed. At the same time acontrol method combined with feedforward, robust and decoupling is come out accordingto the causes of contour error.Secondly, the zero phase robust control system is designed. An inverse systemmethod by combining zero pole and phase compensation is designed by ZPETC. In thecase of zero pole cancellation a larger bandwidth range of closed-loop dynamic system canbe processed inversely. Thereby the closed-loop system phase error and static gain iseliminated. The system tracking performance is improved. The basic idea of DOB is thatthe difference of the actual objects and nominal model output caused by external torque disturbance and model parameters equivalent to the control input. The equivalentinterference is observed, equivalent compensation is introduced in the control, so thatinterference is suppressed fully. The tracking error is estimated by CCC and the contourerror is calculated, the contour error is compensated, so contour error caused by the twinscrew machine does not match is solved.After that, the high model parameters demand of ZPETC is considered. Some linearmotors are influenced by temperature, environment and zero pole drift. The MPC is used toreplace ZPETC as feedforward controller. Past and the future input/output state andinternal model is used by MPC. The future output state is predicted. After feedbackcompensate by Model output error which have compared with reference trajectory, thesystem is rolling optimized by quadratic performance index. Then the system control in thepresent moment is calculated and the whole action cycle is completed. While the MPCcannot guarantee the compatible of tracking and robustness performance. So DOB is usedto improve system robustness and MPC as feedforward controller. At the same timecontour error can be estimated by MPC, so CCC is optimized and contour error is reduced.At last, the zero phase robust control system and predict robust control system aresimulated in MATLAB and compared with the PID controller. The results show thatsystem tracking performance is improved and system error is reduced by zero phase robustcontrol system and predict robust control system. Compared with predict robust controlsystem zero phase robust control system is worse slightly but not obvious. Zero phaserobust control system is designed more simple. Predict robust control system have lowerdemand of model parameters. |
PDF Full Text Request