【摘要】 针对某油气混合型悬架试验系统,建立了油气悬架负载的电液伺服驱动力控制系统的数学模型。提出基于Widrow-Hoff学习算法的ADALINE神经网络控制定步长幅相控制正弦波复现方法。在对算法收敛速度影响因素分析的基础上,提出了应用相对量实现幅相控制的改进方法。以油气悬架加载试验台系统特性为仿真对象,对比了相对量控制方法与绝对量控制方法在收敛速度上的不同,给出了系统收敛时间与输入信号频率和幅值的关系,算法收敛终值与幅值及频率的关系。结果证明,将相对量作为幅相控制的输入,可以有效地提高系统的收敛速度,间接提高了正弦波复现的精度,使幅相控制方法具有更好的实用效果。更多还原

【Abstract】 A mathematic model of electro-hydraulic servo driving force control system for a hydro-pneumatic suspension was established.This paper analyzed the control method usually used in wave form reappeared.The fixed step phase and amplitude control method use ADALINE neural network based on Widrow-Hoff learning algorithm was emphatically introduced aiming at sine wave reappearing.This paper put forward phase and amplitude control improved method use relative quantity based on the analyses of the algorithm the convergence speed influencing factors.This learning algorithm and absolute amount control method were used to control a kind of hydro-pneumatic suspension force loading system model.The convergence rate,the relationship between convergence time and input signal’s frequency and amplitude,the relationship between algorithm convergence value and input signal’s frequency and amplitude was presented.Outcome shows that using the relative value as the input of the phase and amplitude control could improve the convergence speed and the reappeared precision of sine wave effectively.The phase and amplitude control can get better practical effect.更多还原