论文部分内容阅读
为了使所建立的气动力模型能准确地描述飞行器的动态特性,提出一种基于改进粒子群优化(IPSO)算法的小波神经网络(WNN)飞行数据气动力建模方法。该方法引入邻近粒子信息和变异操作,对标准PSO(SPSO)算法的不足进行改进,以提高WNN参数的全局搜索能力,克服早熟收敛,再按照所设计的飞行数据的气动力建模流程,构建了IP-SO算法训练的WNN模型。试验结果表明:提出的气动力建模方法预测精度高,收敛速度快,能较好控制早熟收敛问题,用于飞行数据的气动力建模是有效的,也是可行的。
In order to make the established aerodynamic model accurately describe the dynamic characteristics of the aircraft, a method based on the improved Particle Swarm Optimization (IPSO) algorithm for the aerodynamic modeling of WNN flight data is proposed. This method introduces neighbor particle information and mutation operation to improve the deficiencies of the standard PSO (SPSO) algorithm to improve the global search ability of WNN parameters and overcome premature convergence. Then, according to the aerodynamic modeling process of the designed flight data, The WNN model trained by IP-SO algorithm. The experimental results show that the proposed aerodynamic modeling method has the advantages of high prediction accuracy, fast convergence rate and better control of premature convergence. The aerodynamic modeling for flight data is effective and feasible.