The Application Of Genetic Algorithm And BP Network In The Fermentation Medium

At present the artificial neural network and genetic algorithm are basic content of computing intelligence science. Aiming the insufficiency of BP neural network, a new neural network based on genetic algorithm is presented. The algorithm makes fully use of BP neural network's local searching and GA's global searching and improves the convergent rate and convergent precision.This paper briefly recalls the histories and the fundamentals of artificial neural network and genetic algorithm. And proposes the optimized design method , unifying the genetic algorithm and the BP neural network, then as an example of the Application of BP neural network and genetic algorithm, we give the simulation for the fermentation medium of L-isoleucine by using genetic algorithms and BP-Neural network. The effect of this method is much better than that in suboptimal case. The optimized formula actually produces L-isoleucine 16.84g/L, enhanced 28.6% compared to the results with the initial 10 stochastic formula of L-isoleucine's most high production value 13.09g/L, and the error is 3.2% between the actual production value with 17.40g/L that is estimated.The results indicate that neural network is an effective and quick method for simulation and predication for the process of L-isoleucine fermentation. With the complication, multiple dimension and nonlinearity of manufacturing scale, there is needed much more optimization technology. It takes the advantage of GA and BP to suggest an optimization design method based on BP and GA to obtain unitary efficiency, which uses the GA's parallel searching and solution space searching, to find the good parameters of the net, and then use the BP's simplicity and plasticity to optimize the sample space.