CAJ | 학술논문

针对无人直升机的辨识建模问题,提出了把频率响应辨识法和遗传算法相结合的一种新的辨识方法.该方法充分结合了频率响应辨识法抗噪声能力强、对输入信号通用性强、以及遗传算法的全局寻优特性的优点,极大地提高了辨识的准确性.与最小二乘法、Levy方法的对比仿真结果表明,这种方法辨识精度更高,具有重要的工程使用价值.
침대무인직승궤적변식건모문제,제출료파빈솔향응변식법화유전산법상결합적일충신적변식방법.해방법충분결합료빈솔향응변식법항조성능력강、대수입신호통용성강、이급유전산법적전국심우특성적우점,겁대지제고료변식적준학성.여최소이승법、Levy방법적대비방진결과표명,저충방법변식정도경고,구유중요적공정사용개치.
A new identification method is proposed by combining classical frequency response method and genetic algorithm based on Unmanned Helicopter. The combination of the two methods has strong ability of resistance to the noise, good input signal generality, good excellent optimal ability and faster convergence ability, owning higher identification precision. Comparing with least square method and Levy method, this scheme has the advantages of high parameter identification precision, so it has important practical values.