基于PID神经网络的无人机纵向姿态控制.pdfVIP

基于PID神经网络的无人机纵向姿态控制 陈锋，秦世引 北京航空航天大学自动化科学与电气工程学院100083 1002@163．corn e—mail：chenfen9015 摘要：本文针对某小型无人机智能自主飞行的要求，提出了一种无人机纵向姿态的PID 神经网络控制方法，利用其优良的在线训练、学习和调整功能，可有效实现该无人机的纵向 姿态控制和纵向航迹跟踪。仿真结果表明，本文所设计的PID神经网络控制器较传统的PID 控制器具有更好的控制性能，其响应快，超调小，精度高，可满足自主飞行的要求。 关键词l自主飞行PID神经网络控制纵向姿态控制 Controlandsimulationofthe attitudeforUAV longitudinal basedPIDneuralnetwork Chen FengQinShi—yin ScienceandElectrical SchoolofAutomation University,100083 Engineering，Beihang 1002@163．com e-mail：chenfen9015 thedemandofautonomousforasinail aerial Abstract：Tomeet flight UAV(unmanned is forthe attitudeoftheUAV． PIDneuralnetworkcontroller longitudinal vehicle)，a proposed Witllthe functionofthe controllerforitson·line and superior proposed training。learning attitudecontrolandthe controlforUAVare ability,the trajectorytracking regulating longitudinal the PID network simulationresultsshowthat neural implementedeffectively．’nle proposed controllerCanattain control as overshootand higher performance，suchquickresponse，small high better thantraditionalPIDcontrollerSOasto robustnessand ability precision，better adaptive theautonomicdemands． satisfy flight words：au