Applied Research Of Qualitative Trend Analysis In Process Fault Diagnosis

Process monitoring and quick fault detection and diagnosis are very important for effective management of process operations management, and Qualitative Trend Analysis is one of the important techniques of fault detection and diagnosis. Trend analysis is very important for fault detection and diagnosis, trend can reflect the running states, the development speed and trend of the important parameters, make sure the starting time of the fault, predict the early occurrence of the fault. Qualitative process trend representation is a useful approach to model temporal evolution of sensor data by computer, Trend base on the three primitives: {Steady, Increasing, Decreasing}, using these seven shapes combined by three primitives to describe the process trend. Two important algorithm of Qualitative Trend Analysis (QTA) for process fault diagnosis is systematically introduced—trend extraction algorithm and trend identification algorithm. To make the algorithm have a larger adaptability, the method of grading threshold is introduced to achieve the threshold dynamically required in trend extraction and identification. A framework of trend match methodology is used to diagnosis fault after capturing the trends. The advantage of Qualitative Trend Analysis (QTA) diagnosis faults compactly and quickly for fault diagnosis of process industry.