基于GA/LM混合优化的直升机全机配平算法

Optimal trim for helicopter based on GA and LM hybrid algorithm

基于牛顿迭代法的直升机配平算法简单、高效,但配平结果依赖于初始值。为减小对初始值的依赖,基于直升机飞行动力学模型,通过构造优化目标函数,将全机配平问题转化为优化问题,并针对非线性方程组全局最优解不唯一,将GA算法计算的配平较优解作为LM算法的初值,并构建解空间约束条件,提出一种基于GA/LM混合的优化配平算法。通过计算UH-60A直升机平飞状态和协调转弯状态的配平解,并与单一优化算法和飞行测试数据比较,验证了优化配平算法的准确性和高效性。计算结果表明,基于GA/LM混合的优化配平算法不仅继承了GA算法良好的全局收敛性,也继承了LM算法高效的局部寻优特点,是一种高效、全局最优的直升机配平算法。

Trim algorithm was usually based on the Newton iterative algorithm which was simple and effi- cient, but the results of the trim depended on the initial value. Since the objective function was construc- ted on the basis of UH-60A flight dynamic model, trim calculations were transformed into optimal compu- tation. Taking advantages of the GA and LM algorithm, a hybrid algorithm and the solution space had been developed with constrained, non-unique trim solutions. In this paper, UH-60A helicopter in level flight and coordinated turn flight was trimmed. Comparison between calculated and flight test data were made. The analysis indicates that the GA/LM hybrid algorithm, which not only inherits global conver- gence of the GA algorithm but also inherits excellent characteristics of fast local search of the LM algo- rithm, is a correct and effective algorithm for helicopter trim problem.