deformation prediction model of surrounding rock based on ga-lssvm-markov基于ga-lssvm-markov围岩变形预测模型.pdfVIP

Vol.4, No.2, 85-90 (2012) Natural Science /10.4236/ns.2012.42013 Deformation prediction model of surrounding rock based on GA-LSSVM-markov * Duo-Dian Wang , Guo-Qing Qiu, Wen-Bin Xie, Yue Wang Engineering Institute of Corps of Engineers, PLA University of Science Technology, Nanjing, China; * Corresponding Author: diandian2829@163.com Received 2 January 2012; revised 8 February 2012; accepted 12 February 2012 ABSTRACT technology, etc. [4]. Research method on time series con- tains, polynomial regression, gray theory, the neural net- Command protection engineering is the impor- work and genetic neural network method etc [5-10]. The tant component of national protection engineer- methods have made certain effects. But these methods ing system. To raise the level of its construction, usually have good effect only when the dependent and a deformation prediction model is given based independent variable is linear relationship or some sim- on Genetic Algorithm (GA), Least Square Support ple function relationship. When datum are little and have Vector Machines (LSSVM) and markov theory. no statistical significance, the above methods can’t be used. Genetic algorithm is used to improve the para- Support vector machine, with the foundation of struc- meter of LSSVM. Markov predict method is used tural risk minimization principle, has advantages that the to improve the precision of the prediction model. others, with the foundation of empirical ris