Prediction of Displacement Time Series Based on Support Vector Machines-Markov Chain

Fei Xu; Wen Xiong Xu; Ke Wang

doi:10.4028/www.scientific.net/AMM.580-583.436

Paper Titles

Failure Mechanism of Strip Footing on Geotextile-Reinforced Soil
p.415

Horizontal Displacement of GFRP Bars and PC Strand Combination of Technology Supporting Piles
p.420

Numerical Analysis of Stabilization of Slopes Overlying Bedrock Using Piles and Effect of Socketed Length of Pile on Stability
p.424

Numerical Simulation Analysis of Bearing Characteristics on Super-Large Diameter Rock Socketed Pile
p.432

Prediction of Displacement Time Series Based on Support Vector Machines-Markov Chain
p.436

Project Design of Dewatering for Deep Foundation Pit in Dalian
p.440

Reinforcement and Characteristic Analysis of Bolt Technology Applied in Bedding Rock Slope
p.445

Analysis on Influencing Factors of Cement Stabilized Macadam Base of Compressive Strength in Shenyang Area
p.450

Experimental Study on Dynamic Characteristics of Tailings
p.455

HomeApplied Mechanics and MaterialsApplied Mechanics and Materials Vols. 580-583Prediction of Displacement Time Series Based on...

Prediction of Displacement Time Series Based on Support Vector Machines-Markov Chain

Abstract:

A new displacement time series predicting model was proposed by combining the Support Vector Machines and the Markov Chain, which was named as Support Vector Machines and Markov Chain (SVM-MC) model. Through studying the measured displacement, SVM optimized by particle swarm optimization (PSO) was used to forecast the trend of macro development in roll. Markov chain was applied to compute State Transition Probability Matrix. By classifying system state and calculating absolute error and relative error between measured value and SVM fitting value, the predicting results are improved. The model was used on predicting displacement time series of a high slope of a permanent lock. The engineering case studies indicated that the model was scientific and reliable, and there was engineering practical value for displacement time series forecasting.