| In recent years,the selection of geotechnical engineering parameters in China’s engineering construction projects is unreasonable,facing the hidden danger of design and construction biased to danger,which is easy to cause a certain degree of economic loss and potential risk.For a long time,due to the limitations of field sampling and laboratory test conditions,it is difficult to accurately measure the physical and mechanical parameters of soil in China’s geotechnical industry.In most cases,it is obtained by empirical estimation.Seismic Piezocone Penetration Test(SCPTU)is a kind of in-situ testing technology,which has the advantages of no sampling,high precision,continuous testing,high repeatability,simple and fast,and is favored by domestic and foreign counterparts.Therefore,based on the SCPTU testing technology,the geotechnical engineering mechanical parameters are evaluated,and the SCPTU parameters are used to identify and classify the soil layers.On this basis,the bearing effect of single pile is analyzed,which is of great scientific and practical significance to improve the foundation design in the geotechnical engineering practice in China.This paper attempts to use the advantages of collaborative representation theory,through the MATLAB platform to develop the corresponding program,and puts forward the soil layer identification method based on probability collaborative representation and weighted collaborative competition representation,which provides support for further evaluation of bearing effect of single pile.The main research contents include:(1)In order to better characterize the similarity between soil mechanical parameters,with the help of the possibility that probability quantization parameters belong to each category,the paper discusses the soil layer identification research based on the combination of sparse automatic coding technology of CPTU parameters and probability cooperative representation.Through the verification with the field drilling sampling results,it shows that the method has good recognition effect in soft soil,and can quantify the proportion of each component in the mixed layer,which is a useful supplement to the classical chart method.(2)In order to avoid the problem of high parameter correlation in probabilistic cooperative representation,a soil layer recognition method based on weighted cooperative competitive representation is proposed.Through the identification of several typical sites,it is shown that the weighted cooperative competitive representation can better represent the process of soil deposition.Applying this method to the macroscopic evaluation of the regional site,the general situation of the distribution of geological layers along the borehole depth can be obtained intuitively,which has certain engineering application value.(3)Several methods to determine bearing capacity of piles directly based on CPT/CPTU data are discussed,and evaluates the suitability of each prediction method for static pressure pile in combination with a project in Jiangsu Province.Discusses the range of the influence area of the pile end resistance,the influence of the side resistance degradation effect on the side wall friction,and the influence of the timeliness on the bearing capacity of a single pile.Based on the CPTU data,a new method for predicting the bearing capacity of a single pile based on the weighted cooperative competition representation is proposed.Compared with the static load test under the same geological conditions,the validity of the new method is verified. |
