论文部分内容阅读
针对倒立摆摆起这类复杂非线性系统的优化控制问题,提出了一种基于遗传算法的近似全局最优化的控制方法。在导轨长度和控制力受限的情况下,从倒立摆的物理特性出发,以摆杆进入上半平面前的最后一次角速度为零的时刻为分界时刻,将摆起过程分成两个阶段,分别使用自适应交叉、变异算子的遗传算法搜索每个阶段的最优控制决策,实现了1.9 s内一级倒立摆的成功摆起。这种方法可以简化控制规则,优化摆起时间,并使摆杆更加平稳地进入稳摆控制。仿真和实时控制实验,说明该方法是可行的。
In order to solve the problem of optimal control of inverted complex pendulums in this kind of complex nonlinear systems, an approximate global optimization control method based on genetic algorithm is proposed. In the case of the limited length and controllability of the guide rail, starting from the physical characteristics of the inverted pendulum, the moment when the last angular velocity is zero before the swing rod enters the upper half plane is the moment of demarcation, and the swing up process is divided into two stages, Adaptive crossover and mutation operator genetic algorithm to search the optimal control decision of each stage, to achieve the success of the inverted pendulum within 1.9 s. This method can simplify the control rules, optimize the swing time and make the pendulum more smoothly enter the pendulum control. Simulation and real-time control experiments show that the method is feasible.