运载火箭飞行的自适应多维泰勒网滤波与连续时间控制  

Adaptive Multi-dimensional Taylor Network Filtering and Continuous-time Control of Launch Vehicle Flight

作　　者：陈宇浩 严洪森[1] 张开宝 

机构地区：[1]东南大学复杂工程系统测量与控制教育部重点实验室,江苏南京210096 [2]上海宇航系统工程研究所,上海201109

出　　处：《工业控制计算机》2020年第8期65-67,共3页Industrial Control Computer

基　　金：上海航天科技创新基金项目(SAST2019-020)。

摘　　要：为了提高火箭主动段飞行的精度指标,采用MATLAB仿真技术搭建运载火箭模型,这对于火箭性能的研究具有重要的意义。提出了运载火箭飞行的自适应多维泰勒网滤波与连续时间控制方案。首先建立火箭的三通道仿真模型,基于反步法设计自适应多维泰勒网控制器。分析比较PD+滞后校正优化控制器、模型参考自适应控制器和自适应多维泰勒网控制器对火箭飞行姿态的控制效果。在这三种控制器初始参数均用粒子群算法进行优化的条件下,仿真结果表明,自适应多维泰勒网控制器在三者中具有最好的控制效果。In this paper,adaptive multi-dimensional Taylor network filtering and continuous-time control of launch vehicle flight is proposed.Firstly,the simulation model of the launch vehicle with three channels is established,adaptive multi-dimensional Taylor network(MTN)controller is designed based on backstepping.The control effects of PD+lagging-compensator optimal controller,model reference adaptive controller and MTN adaptive controller for flight attitude of launch vehicle attitude are analyzed and compared.Under the condition that the particle swarm optimization(PSO)is applied to optimize initial parameters of controllers of all the three kinds.The simulation results show that the adaptive MTN controller has the best control effect of them.

关 键 词：运载火箭 三通道仿真 粒子群算法 多维泰勒网 

分 类 号：V475.1[航空宇航科学与技术—飞行器设计]