论文部分内容阅读
提出了一种适用于非线性系统结构和参数估计的新方法．首先用变遗忘因子的最小二乘法辨识参数，导出了准则函数与参数估计一起进行递推计算的方法，进而实现对模型结构的确定．结构确定中采用了ｃｈｏｌｅｓｋｙ分解方法，大大降低了计算量，并给出了算法收敛性的证明．作为仿真应用对Ｆ８歼击机纵向运动系统进行了建模，结果表明，该方法可以有效地确定多项式、样条函数模型结构，参数辨识的结果满足工程要求．
A new method suitable for nonlinear system structure and parameter estimation is proposed. At first, the least square method is used to identify the parameters of the forgetting factor, and the recursive calculation method of the criterion function and the parameter estimation is deduced. Then the determination of the model structure is achieved. The cholesky decomposition method is adopted in the structural determination, which greatly reduces the computational complexity and proves the convergence of the algorithm. As a simulation application, the longitudinal movement system of F8 fighter is modeled. The results show that this method can effectively determine the structure of polynomial and spline function model, and the result of parameter identification meets the engineering requirements.