To improve the limitation of Adomian finite term series solution reducing the convergence for nonlinear dynamical systems, a recursive algorithm for nonlinear systems analysis based on Adomian Decomposition Method( ADM) with suitable truncation order is proposed. The recursive algorithm makes use of Differential Transformation( DT) theory to convert the analytic solution from series into matrix,and then the solution matrix is used in each discrete interval to compute numerical solution iteratively. The maximum stable step-size criterion using recursion percent error( RPE) is developed for good convergence in each iteration. As classic nonlinear dynamical equations,the second-order equation with one RPE and the coupling Duffing equations with two RPEs are illustrated. Comparison results demonstrate that the presented algorithm is valid and applicable to nonlinear dynamical systems analysis.
在全面讨论了以Matlab为软件平台实时仿真方法的基础上,提出了双机RTWT(Real-Time Windows Target)模式的实时仿真系统构架,设计了双目视觉两轮机器人实时目标跟踪系统.相比于传统RTWT模式,该构架大幅度地提高了实时仿真系统的数据处理能力,拓宽了硬件的适用范围.在实时系统测试前,对所使用的位置角度同步、异步控制器及同步补偿算法对目标捕捉性能的影响进行了全仿真研究,仿真结果表明:采用同步控制策略捕捉时间较短,而采用异步控制策略运行轨迹较短,两者各具优势;当扰动不是很大时,同步补偿算法对同步控制的影响较大,能有效缩短捕捉时间和运行轨迹,而对异步控制的影响较小.
