Soil Slope Stability And Reliability Analysis Based On Immune Genetic Algorithm

Soil slope stability analysis is one of the problems encountered frequently in various engineering constructions. How to improve the accuracy, efficiency, and reliability of soil slope stability analysis has been constantly what researchers are concerned. It is the critical technique of soil slope stability analysis to locate the most dangerous slip surface, and to locate the most dangerous slip surface is a nonlinear optimization problem that takes integral stability safety factor of soil slope as a target function. Intelligent optimization algorithms have obvious advantages over traditional optimization techniques in solving complicated nonlinear optimization problems due to their unique mechanisms from nature and strongly robust, highly parallel, distributive characteristics. With genetic algorithm as a basic frame, a kind of improved immune genetic algorithm has been put forward benefiting from the thoughts of vaccine selection, vaccination, immune memory, gene affinity mutation, gene recombination. Compared with simple genetic algorithm and existing immune algorithm, some modifications and improvements have been made mainly in such aspects as crossover mode, algorithm structure, vaccine selection mode, vaccination mode, immune selection. Proper testing scheme has been designed to test the performances of the improved immune genetic algorithm in this paper by utilizing five typical functions which are usually used to test intelligent optimization algorithms. The result shows that the algorithm has better improvements in convergence, search accuracy, and efficiency than simple genetic algorithm and simulated annealing algorithm. Based on the immune genetic algorithm, the models of soil slope safety stability analysis have been studied. And such main problems as genetic variables designing, genetic coding, fitness function designing, automatic realization of the calculation from designing variables to fitness function in the problem of most dangerous slip surface search of soil slope have been solved. The most dangerous slip surface search models of circle and non-circle shapes, which can be used to solve problems of various soil slope safety stability analysis, have been developed based the immune genetic algorithm. The comparison of engineering testing examples and the calculating result of an engineering case show that the models are effective, and have good performances.The models can objectively respond to the influences of slope angle, groundwater, seismic effect. Considering that stochastic variability of parameters has great influence on soil slope stability analysis, a new method and model for soil slope reliability analysis has also been put forward and developed with the basic theory and methods of structural reliability analysis as starting points and by a combination of Monte-carlo stochastic simulated technique and soil slope safety stability analysis method based on the immune genetic algorithm. The calculating results of an engineering example show that the model has a good performance of convergence and can give proper responses to the effects caused by stochastic factors in soil slope stability analysis. And it also verifies the validity of the model that the calculating result is in agreement with the actual case. Besides, the problem of soil slope reliability designing is also further discussed based on the model and the example in this paper. It is pointed that the soil slope stability and reliability can be effectively improved by controlling the stochastic variability of strength indexes of backfill soil or by using anchor cables (rods) timbering.