Analysis of optimal sampling period based on multi-rate predictive control systems
Ma Ailiang,Zhou Lifang
DOI: https://doi.org/10.3969/j.issn.1001-4160.2010.01.018
2010-01-01
Abstract:With the development of industry,in complicated chemical process,multi-rate control system is used in many more fields,while sampling period is an important parameter with aspect influence in digital control system.Now,the method for the selection of sampling period is still limited in experiential selection,and there is not a effectual method to define the optimal sampling period which the system most needs.In this paper,based on a multi-input -multi-output system and the algorithm of dynamic matrix predictive control,through the establishment of a two input-two output multi-rate input model,obtain a universal model of multi-input and multi-output under the strategy of multi-rate input;Using the classical optimization algorithm - genetic algorithm to realize the optimization of sampling period,when the basic sampling period is defined as T,each optimal sampling period which can make the system performance best is found,the optimal control performance index is called integral square error, the corresponding two inputs are respectively five times and one time of the basic sampling period,compare the performance under the single-rate control strategy,the simulation for a speed-coupling system indicates that,the method proposed in this paper is effective to obtain the optimal sampling period,when the sampling-period is selected properly,the stability and system performance with multi-rate input system are better than single-rate.