This paper combines the global optimization ability of the symbiotic parallel genetic algorithm and the local optimization ability of the improved LMBP algorithm to research,proposes an neural network structure optimization symbiotic parallel genetic algorithm and to testify the correctness and validity of this algorithm by the simulation experiments. This algorithm realizes unequal length coding, large probability cross, small probability variation, cross and variation between sub-populations, information exchanging between sub-populations etc, and successful implements the optimization of neural network structure. The experimental results shows that this algorithm having reliable performance, searching a large space, be able to find the feasible solution within the specified generalization and approximation error range.
Study on Real-Time Porosity Defect Detection Through Neural
Network Structure Optimization using Genetic Algorithm in GMAW
