Improvement of ant colony algorithm for predicting uneven settlements of shield tunnel in soft soil and its parameters selection

Zhou Shun-hua
2011-01-01
Abstract:According to monitoring settlement data,the law of the uneven settlement of shield tunnel in soft soil and its causes are analyzed;and then the leading information with regulation is obtained.Together with the auxiliary information with great randomness,it can be used to improve the previous prediction model based on ant colony algorithm.The principle for selecting partial parameters in the prediction model is preliminarily conferred by an instance.The results show that:(1) The leading information with regulation is considered to have something to do with the amount and cumulative settlement of positions with significant settlement in the vertical sedimentation tank,and the distance between the positions with significant settlement and the positions with non-significant settlement.(2) The improved prediction model is clear in calculation process,simple in formula,easily to use and has high prediction accuracy;and its maximum error can be controlled about 0.5%.(3) The influence of match between evaporation coefficient of pheromone and the total number of ants on prediction accuracy is relatively significant;whereas the correction function reflecting auxiliary information has a little role when there is no random factor;based on these,a reasonable process of parameter selection in the model is given.
What problem does this paper attempt to address?