摘要
对时变随机线性系统在坐标原点附近任意小范围内的摄动问题进行分析:当干扰噪声为连续函数时,统计规律服从高斯正态分布,并给出控制规律或控制函数;当该随机线性系统在控制过程中缺少先验约束条件和干扰噪声的采集时,通过补偿以消除噪声干扰给系统带来的不利影响,并通过决策因子递推迭代,得到最优控制策略和决策值。仿真结果表明,该系统在二次型性能指标条件下,能够获得线性二次型最佳对策。
The paper is concerned with analysis of moving a random linear nonstationary system to a specified vicinity, which can be as small as desired, of the origin of coordinates. If the disturbing noise is in continuum function, then the rules of statistics accord with Gauss normal distribution, and the controlling rules or functions are provided. If the random linear system lacks in conditions of prior constraints and the collection of noise during the course of control, then, the adverse effects of noise to the system can be eliminated by moans of compensati- on, and the optimal control stratagem and decision value can be obtained by recursive iteration of decision factors. The simulation result reveals that optimal decision of linear quadratic game can be achieved with the system under quadratic performance.
出处
《茂名学院学报》
2007年第4期28-30,共3页
Journal of Maoming College
关键词
线性系统
随机控制
摄动分析
linear system
random control
move analysism