Optimization of Fractional-order Stochastic Resonance Parameters Based On Improved Genetic Algorithm

Fractional-order stochastic resonance (FOSR) system can use for noise in detecting weak signals and make them produce positive effect, so as to enhance the amplitude power of weak signals. In this system, the parameters of the bistable system, the fractional order and the noise intensity of the input all have a certain influence on the output of the system. For the purpose of achieving the best effect of the output, an improved genetic algorithm(GA) is proposed in this paper. This algorithm introduces the idea of simulated annealing(SA), and makes adaptive adjustment to multiple parameters. Numerical simulations show that the algorithm has a stronger global optimization capability than traditional genetic algorithms, and it improves the convergence speed and reduces the amount of calculations, which is conducive to the use of fractional stochastic resonance systems in practical applications.