Energy feature extraction for stealing oil signal of pipeline with wavelet package transform

The key to carry out real-time monitoring of stealing oil events is the effective identification of stealing oil stress wave signal. The test system of stress wave signal about stealing oil was briefly introduced, and on-the-spot data collection steps were given. With wavelet package transform, features of energy distribution for stealing oil signal on different band were obtained. Analysis of experimental results show that identification of stealing oil signal can be done through the difference of energy distribution characteristics at 0Hz ~ 640Hz band. The research provides a new approach for identification of stealing oil signal.