弹箭与制导学报
彈箭與製導學報
탄전여제도학보
JOURNAL OF PROJECTILES, ROCKETS, MISSILES AND GUIDANCE
2015年
1期
1-4,8
,共5页
制导控制一体化%反演%神经网络%弹道仿真
製導控製一體化%反縯%神經網絡%彈道倣真
제도공제일체화%반연%신경망락%탄도방진
integrated guidance and control%backstepping control%neural network%trajectory simulation
设计了一种适用于制导控制一体化的小波神经网络反演方法( WNNBM )。首先，将目标运动学特性与导弹动力学的系统模型相结合，建立并简化了俯仰通道制导控制一体化模型，综合设计制导律与控制器；其次，利用自适应反演方法对一体化控制器进行总体设计，并针对一体化模型简化程度高，未建模动态难以确定的特点，引入小波神经网络( WNN)逼近系统模型；最后，进行弹道仿真。结果表明，该方法可以有效打击目标，提高系统性能。
設計瞭一種適用于製導控製一體化的小波神經網絡反縯方法( WNNBM )。首先，將目標運動學特性與導彈動力學的繫統模型相結閤，建立併簡化瞭俯仰通道製導控製一體化模型，綜閤設計製導律與控製器；其次，利用自適應反縯方法對一體化控製器進行總體設計，併針對一體化模型簡化程度高，未建模動態難以確定的特點，引入小波神經網絡( WNN)逼近繫統模型；最後，進行彈道倣真。結果錶明，該方法可以有效打擊目標，提高繫統性能。
설계료일충괄용우제도공제일체화적소파신경망락반연방법( WNNBM )。수선，장목표운동학특성여도탄동역학적계통모형상결합，건립병간화료부앙통도제도공제일체화모형，종합설계제도률여공제기；기차，이용자괄응반연방법대일체화공제기진행총체설계，병침대일체화모형간화정도고，미건모동태난이학정적특점，인입소파신경망락( WNN)핍근계통모형；최후，진행탄도방진。결과표명，해방법가이유효타격목표，제고계통성능。
A wavelet neural networks backsteppting method ( WNNBM) was designed for integrated guidance and control. First, the inte-grative control model of missile was built and simplified, the target kinematics characteristics and missile dynamics model of the system were combined, and guidance law and controller were put together. Controller was designed for integrative model by using backstepping method ( BSM) , in view of high degree of model simplification and uncertainty of unmodeled dynamics, wavelet neural networks ( WNN) in backstepping method was used. Finally, trajectory simulations were carried out and simulation results show the method can hit the target effectively and improve maneuverability and efficiency of missile.
