进化算法多目标优化的飞控参数自整定方法  

Self-tuning of FCS Parameters with Multi-objective Algorithm

作　　者：何磊[1] 马建仓[1] 李广文[1] 

机构地区：[1]西北工业大学电子信息学院,陕西西安710129

出　　处：《计算机测量与控制》2010年第3期598-600,共3页Computer Measurement &Control

基　　金：国家自然科学基金项目(60672184)

摘　　要：随着飞行品质要求的提高,且由于飞控系统模型的复杂,控制器参数的传统整定方法耗时长且精度低;引入多目标进化,针对某飞机横侧向控制系统模型中的协调转弯模态,采用改进的非劣排序进化算法对飞行控制器参数进行了整定;此优化策略具有经典控制系统简洁实用的特点,且能同时优化多个目标,能够快速得到符合要求的控制器参数;仿真结果表明所设计的控制器具有良好的性能指标和鲁棒性,具有一定的实用价值。With the improving of the flying quality and the flight control system （FCS） becoming more complicated, the typical method to get the parameters of the FCS is time--consuming and hardly up to scratch. In order to solve this problem, we introduced the multi--objective genetic algorithm to the design of coordinating--turn in the aircrafts＇ flight control system, based on the method of non--dominated sorting. This optimizing strategy is intelligible and practical because it keeps the typical control system unchanged. It can also deal with several flying qualities and find the parameters fast. The simulation shows that this controller has a good performance and robustness. The optimization strategy is useful and can be used in other fields.

关 键 词：飞行控制系统 多目标优化 控制参数自整定 进化算法 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术] TP273.2[自动化与计算机技术—计算机科学与技术]