Application of genetic expression programming and artificial neural network for prediction of CBR

作     者：Tenpe, Ashwini R. Patel, Anjan 

作者机构：Visvesvaraya Natl Inst Technol Dept Civil Engn Nagpur Maharashtra India 

出 版 物：《ROAD MATERIALS AND PAVEMENT DESIGN》 (道路材料与铺筑道路设计)

年 卷 期：2020年第21卷第5期

页      面：1183-1200页

核心收录：

学科分类：08[工学] 0805[工学-材料科学与工程（可授工学、理学学位）] 0814[工学-土木工程] 0823[工学-交通运输工程] 

主　　题：Gene Expression Programming artificial neural network California bearing ratio prediction 

摘      要：TheCBR(California bearing ratio) value is an important parameter of the subgrade soil required for the design of pavements. The present study deals with the application of genetic expression programming (GEP) and artificial neural network (ANN) for the prediction ofCBR. Various soil properties, such as gravel percentage (G), sand percentage (S), fine content (FC), liquid limit (W-L), plastic limit (W-P), plasticity index (I-P), optimum water content (Wc(opt)) and maximum dry density (gamma(d)(max)), were considered as the variables input parameters in the analyses. Mathematical expressions were developed for the prediction ofCBRand their dependency over different combinations of variables was obtained byGEP. The same combinations of variables were used forANNprediction. It was observed that both theGEPandANNmethods fit well forCBRprediction and the model consisting of variablesG,S,I-P,W(C)(opt)and gamma(dmax)was found to be the best model. It was found that 80 numbers of chromosomes, 3 head length and 3 numbers of fixed genes is the optimal condition for the prediction ofCBRbyGEP. TheGandSwere found to be the most significant parameters with 28.41% and 39.62% contribution in case ofGEPand 26.83% and 23.37% in case ofANN, respectively. The variableW(P)was not used byGEPduring optimal model construction, which may imply poor dependability ofCBRover it.