Abstract: Based on the coupling phenomena between the batch drilling process quality fluctuation and monitoring signals features changes, a kind of signals feature extraction method based on bi-spectrum feature is proposed to solve the quality consistency control and testing problem of the high-precision batch drilling step. Theoretically, the acoustic emission signal and acceleration vibration signal of a normal drilling process is viewed as a stochastic process to meet with Gaussian distribution, and there could be a corresponding relationship between the deviation degree from Gaussian distribution of signal and the drilling step quality fluctuation. The acoustic emission signals and acceleration vibration signals of batch drilling are taken as research objects, the quantitative analysis of the deviation degree from Gaussian distribution under different conditions could be expressed by the average bi-spectrum amplitude of each signal. The ReliefF algorithm is used to assign the weights for every feature, these step quality classification is performed using feature weighted fuzzy cluster algorithm and to be contrast with the manual detection result. The results show that there are organic connections between the bi-spectrum feature of monitoring signals and drilling step quality, and the consistency quality testing of batch drilling step is realized by fusion of clustering bi-spectrum features.

Key words: batch drilling, bi-spectrum, Gaussian distribution, monitoring signals, step quality