Research on Low Strain Detection Technology of Foundation Pile in Residential Building Engineering

A low strain detection technology for residential building engineering foundation piles was studied. Reflected wave signals reflected from the pile body were collected through sensors. In order to improve signal quality, the EEMD method was used to denoise the signals. Time domain and frequency domain characteristics of reflected wave signals from foundation piles of different defect categories were extracted and used as input. BP neural network was used to classify and identify models to detect the extent of foundation pile defects, and locate the defect location based on relevant parameters to complete low strain detection of foundation piles. The test results show that the error between the detection results of foundation pile defects and the actual defect location is smaller, and the detection results are more accurate, which has good application value.