Study Of The Deep Foundation Pit Construction Control Based On Neural Network And Genetic Algorithm

Recently, allowing to the development of modern city construction, the foundation pit developed in depth and large area will be a trend. The following foundation pit accidents calling attention to us to thinking highly of the problem of constructing control in foundation pit. Today, the most control ways is monitoring. Because of the monitoring process lasting long time and have many uncertainty factors, how to using the monitoring data before to predicate the later monitoring data better is the important content in this paper.Firstly, this paper introduces monitoring aims, contents, methods and data processing in deep foundation pit construction detailed. Now, there are many prediction methods of monitoring data processing. For the complexity of deep foundation pit construction, it builds the prediction model using neural network that have excellent non-linear mapping ability in this paper. At the same time, it using genetic algorithm optimize the whole prediction model to making up the shortage of neural network.After introduces the basic theories of neural network and genetic algorithm, it tells how to using BP neural network to build deep foundation pit deforming prediction model. Because BP neural network has a difficult to choose the best weight, it settles this problem by using genetic algorithm in paper. It using Matlab language makes a program for genetic algorithm optimizing BP neural network.At last, it associating engineering example build BP single-element predicting analysis model and multi-element predicting analysis model. It using the two models predicts a few of important target in deep foundation pit constructing control. The using result showing that no matter single-element BP prediction model or multi-element BP prediction model all applies in engineering. It builds genetic algorithm optimizing BP neural network model -GA-BP model for the shortage of BP neural network in prediction. To verify the GA-BP model, it compares the predicting result of BP model and GA-BP model for those important targets in deep foundation pit contrasting. The result shows GA-BP model have strong engineering predicting ability. It can remedy the shortage of BP prediction model completely. It has broad spread value in engineering.