Double inverted pendulum system control strategy based on fuzzy genetic algorithm
Chengjun Ding, Duan Ping, Minglu Zhang, Yanfang Zhang · 2009
The double inverted pendulum is a MISO complication system. If only adopting the traditional fuzzy control, then the total of fuzzy rules will bring the exponential increase with the increase of input variables, which is disadvantaged to the real-time control. So in this paper, we firstly adopt the add weighting coefficient to reduce the input variables of fuzzy controller, then using genetic algorithm to optimize the weighting coefficient and membership function parameters of fuzzy controller by the off-line way, finally, using the optimizing result to control the double inverted pendulum system. In this paper, we introduce the whole optimizing operations with GA, which are coding strategy, determination of the fitness function, selection, crossover and mutation. In the end, by the validation of simulation platform Matlab gives out the curve of simulation and shows the optimizing control strategy is more effective in the aspect of speed response and stability than traditional fuzzy controller.