Monitoring, diagnosis and control of industrial processes

We present a framework for monitoring, diagnosis, and control of industrial processes. This framework utilizes the multiresolution analysis capability of wavelet theory. Wavelet coefficient patterns at different scales, under a variety of process conditions, are noted to form process fingerprints, these fingerprints yield process fault diagnosis. This knowledge facilitates efficient control.