Diagnosis System for Detecting Plant and Sensor Deterioration Using Immune Network Integrated with Plant Knowledge

For large plants such as iron steel, chemical and public institution, we have been studying sensor diagnosis system using the immune network, which is autonomous distributed system known as high scalability and maintainability. Meanwhile, plant time deterioration causes changes to value detected by sensors compared to one detected before. Generally, it is difficult to tell which causes sensor changes, between sensor faults and plant deterioration. In this paper, we describe a diagnosis system that detects sensor and plant deterioration separately using the immune network combined with plant knowledge. After constructing two kinds of network based on plant knowledge, a network for diagnosing plant time deterioration and another for sensor faults, the proposed system operates them sequentially. We also show the efficiency of proposed system with simulations applied to the continuous galvanizing line plant using actual plant data.