Genetic algorithm-based study on model parameters identification for multi-functional sensor

In order to avoid the inadequate extremum convergence of the evaluated parameters, a method about multi-functional sensor model parameters estimation is proposed which is based on genetic algorithm (GA). Compared with the traditionally used least square strategy, GA is of overall optimal, emulation results indicate that the dualistic variables of a multi-functional sensor could probably be reconstructed with the nonlinear input/output structure parameters being evaluated in a proper computation speed and veracity.