PSO-NP Maching Dimension Type Generalized Predictive Control Based On PI

In the process of machining quality on-line monitoring, data processing size prediction is particularly important, and predict machining dimenison data is the key to realize the error feedback compensation control, so the research of high precision and the ability to play well in the actual industrial process control effect of machining dimension data to predict algorithm of excellence is particularly important.Mechanical processing to reform the workpiece geometry parameters can effectively improve the efficiency of social production to produce greater economic benefit, so the machining dimensions of predictive control is particularly important.machining process for machining size detection and control can effectively ensure the accuracy of parts processing size,the first condition is to establish the appropriate model to simulate the procssing, the change of the size of the accurate description and prediction of the multstep prediction and the generalized predictive control, the system according to the size of the input and output data, and selected the future size of the input values, constantly to predict the future size of the output value which can overcome the uncertainty of the system and enhance the system robustness.But the selected parameters in GPC contact engineering practical requirements of the indicators is not quite close together, and for the interference of random sudden cannot achieve the realtime control effect.PI type of generalized predictive control can improve the realtime tracking, improve the control quality of system, combining with the advantages of two kinds of control technology can produce control effect is more in line with the actual requirements of PI type of generalized predictive control.However, various factors in the process of machining mechanism is very complex, and by many of the actual conditions of constraint processing, existing in control system of the constraints, is bound to control quantity of solving difficult, complicated problems, increase the amount of calculation and affect the performance of the algorithm, plus PI type of generalized predictive control the scaling factor and integral factor setting more difficult, so the parameter setting has the very vital significance.For workpiece size prone to error in the process of machining defects, this article through to the machining dimension online modeling sequence of testing data, this paper proposes a of nonlinear programming based on particle swarm optimization (pso) algorithm and PI type of generalized predictive control with constraints.The algorithm through nonlinear programming processing machinery processing control system in the process of input and output constraints, type PI constrained generalized predictive control is obtained control law, and secondary search by particle swarm optimization (pso) algorithm, the constrained generalized predictive PI type of scaling factor and integral factor to optimize the setting, through matlab simulation results show that the algorithm can accurate projections for processing the change of the size and make the processed parts size basically controlled within permissible error. On the nc machine tool for processing of a batch of specimens to predict the processing size of the output value, with GAPIGPC optimization algorithm by matlab simulation to this algorithm, and optimize the parameters in the algorithm not setting algorithm is concluded that the new algorithm can effectively reduce the system’s overshoot, shorten the adjusting time, so as to achieve better control effect, and has better tracking performance and adaptability, output fluctuations in smaller, control more ease and reduce the amount of control oscillation caused by model mismatch.