论文部分内容阅读
在模型未知的情况下,估计过程的重要变量尤为重要.鉴于此,采用不敏卡尔曼滤波(UKF)与神经网络相结合的方法,解决一类未知模型非线性系统的状态估计问题.采用动态神经网络对非线性系统进行建模,利用UKF对状态和权值进行同时更新,从而达到神经网络逼近真实模型,估计值跟随真实值的目的.通过两个仿真实例表明了所提出的方法具有良好的估计效果,并且状态在输出中的比重越大,其估计精度越高.
In the case of unknown model, the important variables of the estimation process are particularly important.In view of this, the state estimation problem of a class of unknown model nonlinear system is solved by the combination of unscented Kalman filter (UKF) and neural network, The neural network models the nonlinear system and uses UKF to update the states and weights at the same time, so that the neural network approximates the real model and the estimation value follows the true value. The simulation results show that the proposed method has good performance Of the estimated results, and the state of the larger the proportion of the output, the higher the estimation accuracy.