Based on genetic algorithm and BP network to fault diagnosis of fermentation process

The proposed method is adopted by genetic algorithm and BP neural network combine together, so GA-BP network is structured and used to the study of fault model and fault diagnose. It combines the ability of genetic algorithm with global optimizing and neural network with learning and memory, and resolves to problems that are slowly study speed and local minimum solution to produce easily for the BP network alone. Finally, it takes the fault diagnosis of the Fermentation Process as the example. GA-BP network is trained Compared to BP network alone, it proves GA-BP network has the quick convergence speed and stable the ability of overall optimizing.