Research On Control Of Inverted Pendulum Based On Genetic Algorithms

Inverted pendulum is a typical fast, multivariable, nonlinear, strong-coupling andnaturally unstable system. During its control process, many crucial questions in thecontrol theory can be reflected, such as calm question, nonlinear problem, robustquestion as well as tracking question and so on. The research on inverted pendulumsystem has the far-reaching significance in theory and project application. Thecorrelative scientific research achievement has already applied to astronautics sciencetechnology and subject of robot and so many domains.By using the inverted pendulum as the experiment platform, the purpose of thispaper is utilizing Genetic Algorithms to optimize and design the pole assignmenttechnology, linear quadratic regulator controller and solve the bottleneck problem oftransformation from quantitative to qualitative in human-imitating intelligent controlautomatically and releasing people from the time-consuming trial and error exercises.The computer simulations and real time control on the system are performed withinverted pendulum and show that Genetic Algorithms can efficiently optimize theperformance of controllers and reduce the design cycle time from hours to minutescompared to original optimization method.