| Inverted pendulum system has the characteristics of high-order, multi-variable, non-linearity, strong-coupling and un-stability, to keep inverted pendulum system stable, thus, it is necessary to adopt effective control algorithm. As a typical experiment device, it can make the abstract control theory more visualized, such as the robustness problems, anti-interference problems, nonlinear problem. Meanwhile, the inverted pendulum also has been widely applied to such as aerospace and robot areas, so the study on the inverted pendulum system has profound theoretical and practical significance.The paper regards the rotary double inverted pendulum as the experiment subject provided by the Googol Technology of Shenzhen, carrying out linear quadratic optimal control based on artificial bee colony algorithm and the research on the improved artificial colony algorithm, the main contents are arranged as the following:First, According to the characteristics of the rotary double inverted pendulum system, by using Nuton-Euler method,the mathematical model of rotary double inverted pendulum is established. Then, the stability, controllability and observability of system are also analyzed. Given of the simulation study, the results present the open-loop characteristics of inverted pendulum system.Secondly, we has studied and designed the LQR controller of inverted pendulum system based on its state space equation, meanwhile, the artificial bee colony algorithm(ABC) is introduced to optimize the weighing matrices Q and R. To obtain better control effect, the strategies are adopted from the three phase of ABC algorithm, then the simulation study on the inverted pendulum system is performed.Finally, To verify the effective in own control algorithms, we implement the real-time control of inverted pendulum based on the control software produced by Google Company. From the actual operating results, the LQR controller optimized by improved ABC algorithm obtain better performance. |
PDF Full Text Request