Empirical Analysis And Forecast Of The Concrete Diaphragm Wall And Unit

In recent years in our country, some projects have collapsed frequently after completed, which leads to enormous economic losses to the state and the people. The reason is that they did not establish theoretical models among the materials, the geological formation of the flooding rate, unit volume of water leakage and the stability volume of water leakage, but to carry on the project by experience accumulated, leading to the materials used unreasonable and infrastructure unstable, even though they understood the local geological structure to some extent. Hence, according to the local geological conditions, studying the relationship between the grouting with the impact of factors, giving a reasonable model to forecast the unit grouting has very important significance, as well as the cost savings, to prevent the occurrence of disasters and the safety of the people.Firstly, in section 2, we studied the relationship between the unit grouting with the geological structure, correlation of unit grouting of diaphragm wall of a dam, unit volume of water leakage, grouting pressure, pressing water pressure, the initial cement slurry filter rate, and eliminated the outliers by one-by-one method.In section 3 and section 4, based on the analysis the relationship of the concrete diaphragm wall and unit grouting quantity of the five key factors, we constructed stepwise regression model of unit grouting of diaphragm wall, the flooding rate of geological formation, unit volume of water leakage, grouting pressure, pressing water pressure, the initial grouting rate of cement slurry. We also tested the established models by mathematical statistics methods and used those models to forecasting.Finally, this paper compared those models by goodness-of-fit, validity of forecast, and other aspects as well. The results showed: the absolute value of partial correlation coefficient between unit grouting and unit volume of water leakage is lager than the absolute value between unit grouting and unit volume of water leakage, and the absolute value between unit grouting and grouting pressure. The absolute value between unit grouting and grouting pressure is the smallest one. In regression models, we can see that the absolute value of the coefficient of grouting pressure is large than that of the leakage, the rate of flooding, so the forecast results in practice should be difficult to be accepted, and the PLSR may be better. Basically, the regression coefficient is equals to the partial correlation coefficient. The goodness-of-fit of two models have a smaller gap. Basically, the signs of the regression coefficients and the partial correlation coefficient are consistent in those two methods, but the PLSR method has higher accuracy of forecast than that of stepwise regression method.