针对直升机飞行动力学建模参数优化问题,提出了一种基于粒子群的线性变参数模型预测控制(Line-ar parameter variable and model predictive control,LPV-MPC)优化设计,通过优化建模参数,提高直升机飞行动力学建模精度.基于经过滤波加权的前飞速度和横向速度对操纵量、姿态角、气动导数和操纵导数进行三维插值,对气动力进行实时计算,完成了UH-60 直升机的LPV-MPC系统建模.采用粒子群优化算法,针对前飞速度插值间隔、横向速度插值间隔和滤波常数等三个建模参数进行优化.利用UH-60 直升机非线性模型生成一系列用于优化的线性状态空间模型,对直升机纵向通道、横向通道同时给定激励进行仿真试验.仿真结果表明:在一定范围内任意选取的优化参数,经过粒子群寻优后,建立的LPV-MPV模型精度更高.
Research on LPV-MPC Modeling Method based on Particle Swarm Optimization
Aiming at the optimization of helicopter flight dynamics modeling parameters,a particle swarm based linear variable parameter model predictive control optimization design was proposed to improve the accuracy of helicopter flight dynamics modeling by optimizing modeling parameters.Control variable,attitude angle,aerodynamic derivatives and control derivatives were interpolated according to current lateral speed and flight speed weighted by a low-pass filter,then the aerody-namic force and moments were calculated in real time.Finally,the LPV-MPC system modeling of UH-60 helicopter was completed.Particle swarm optimization was used to optimize three modeling parameters such as forward velocity interpolation interval,lateral velocity interpolation interval and filter constant.A series of linear state-space models for optimization were generated by using UH-60 helicopter nonlinear model.Experiments were carried out on the helicopter longitudinal channel and transverse channel at the same time with given excitation.The simulation results showed that the LPV-MPV model was established with higher accuracy after particle swarm optimization with arbi-trary optimization parameters in a certain range.
万方数据
维普
