Research On PI Parameter Tuning Of The AC Servo System In Ultra-high Speed Cigarette Making Machine

With the rapid development of servo drive control technology, permanent magnet synchronous AC motor drive to replace the traditional mechanical transmission gear, has been widely used in modern high-speed and high-precision industrial systems. Control parameters of the servo system are closely related to control performance, when it comes to multi-axis servo system, it is cumbersome with repeated parameter tuning process. In this paper, the Siemens servo system of load wheel of ultra high-speed cigarette making machine is taked as the research object. For the problem of inaccuracy when directly modeling and simulation with the actual system parameters, this paper turns the control parameters of actual Cigartte making machine according the identified discrete model of controlled object from system operating data, and then its established simulation model.First of all, the mathematical model and the structure of the vector control of permanent magnet synchronous motor are analyzed, and the tricyclic transfer function model of servo system is established in this paper. Then, according to the classical control theory, the general optimization formulas of current loop, velocity loop and position loop control parameters are gained from interior to exterior and lay the foundation for the identification analysis.Secondly, based on the mathematical model of permanent magnet synchronous AC servo system and closed loop excitation sequence of actual system, the controlled object discrete model of velocity loop and the identification problem under closed loop conditions are analyzed. The innovation estimation adaptive Kalman filtering method is proposed to approximately identify the controlled object of speed loop under closed loop conditions, in which algorithm can effectively suppress the influence of disturbances on the accuracy of identification. Though the recognition result is not the only solution of actual system, it can better reflect the dynamic input-output characteristics of the servo system of actual Cigartte making machine. According to the identified second order discrete transfer function simulation model of controlled object, a pole placement PI control parameters tuning algorithm of third order in closed loop is proposed to turn parameters and scope fastly. Then the self-tuning strategy using rule-based genetic algorithm is to tuning PI control parameters of speed loop globally.Thirdly, the discrete mathematical models of speed loop controlled object of no-load, fixed load slump wheel and variable load separating wheel are identified and analyzed under closed loop conditions. The validity of adaptive Kalman filter system identification algorithm and the feasibility of approximate identification model of actual servo system under closed loop conditions are verified though comparing. According to the established approximate equivalent simulation models of actual servo system, the control parameters of speed loop and position loop of no-load and fixed load slump wheel are carried through corresponding self-tuning analysis, and the control parameters of variable load separating wheel are analyzed through electromechanical co-simulation. A more optimized simulation control parameters of the servo system are obtained.Eventually, the control parameters of speed loop and position loop through simulation tuning analysis are verified on the ultra-high speed Cigartte making machine experimental platform, and the influence of different control parameters on the servo control performance are also analyzed by simulation and experiment. After tuning, the servo system appears to smaller speed overshoot and position following error, and the overall control performance of the servo system are improved effectively.