基于微粒群优化的船舶柴油机T-S模糊模型

提出了一种基于微粒群优化(PSO)的T-S模糊模型的非线性系统辨识方法,并用于船舶柴油机的动态建模.该辨识方法采用GK模糊聚类算法确定模糊模型的前件结构及参数,利用PSO算法来辨识模糊模型的结论参数.利用6160-All船舶柴油机模型,获得柴油机各主要参数在油门尺度和负载发生小偏差扰动时的试验数据,再利用该组数据辨识出柴油机转速、涡轮增压器转速、增压压力、空冷器压力、进气管压力、排气管压力等参数的T-S模糊动态模型.仿真结果表明了该方法的有效性.

T-S Fuzzy model of marine diesel engine based on particle swarm optimization

An identification algorithm based on a Takagi-Sugeno(TS) fuzzy model is proposed for nonlinear systems using particle swarm optimization(PSO).This identification algorithm can be used for ship engine dynamic modeling.In this algorithm,a Gustafsson and Kessel(GK) fuzzy clustering algorithm is applied to decide the configuration and initial parameters of a fuzzy model, then a PSO algorithm is used to identify the final parameters.Experimental data was obtained from a 6160-A11 engine model while throttle levels and loads were subjected to small perturbations.This data was used to build a T-S fuzzy dynamic model relating engine speed,turbocharger speed,turbocharger pressure and air condenser pressure,intake pipe pressure,and exhaust pipe pressure of a ship's diesel engine.Simulation results show that this algorithm is effective.