The Research On The Prediction Of Fly Ash Concrete Strength And Mixing Ratio Design Based On Genetic Neural Network

Genetic algorithm(GA)is a simple,wide-used, and robust probability searching algorithm, Compared to other optimizing methods, GA adopts some particular methods and techonolgies, and is easy to be mixed with other techniques such as neural network ,fuzzy reasoning, so it has been used widely. Prediction of 28 day's compressive fly ash concrete strength, which is very significant to increasing the efficiency of concrete production, is a typical multi-variable, nonlinear system. The existing methods of the prediction is difficult to be applied widely in practice because its accuracy is lower. In this article the author combines artificial neural networks with GA as a new modeling and predicting method of fly ash concrete strength , and on the basis of the above, a new optimizing design method is presented. The experiment results show the method above mentioned is effective and feasible. The model for the prediction of concrete strength and the model for optimizing design, which is established by making use of before-mentioned methods is handy for operating, high precision in prediction and strong adaptability and so on.