ระบบผู้เชี่ยวชาญในการแก้ไขปัญหาข้อขัดข้องสำหรับโรงไฟฟ้าพลังน้ำเขื่อนนเรศวร

Abstract

This research studies fault diagnosis on an unmanned 8 MW bulb turbine generator of the run of river hydropower plant locates far from the control and maintenance center by expert system scheme. The aim is to assist the inexpert maintenance crews to solve the failure by collecting knowledge from power plant equipment data, maintenance instruction manual, maintenance standard, historical failure events, and body of knowledge from expertise. The studied system translates this information to the knowledge base of an expert system by propositional logic together with failure analysis of mechanical and electrical equipment that initiates lockout relay signal to shut down the power plant by fault tree analysis technique for more accuracy and correctness of fault diagnosis. The implementation of an inference engine comprises of failure search rules and problem-solving rules with developing a graphical user interface that similar to the SCADA graphic of the operator station. This helps inexpert maintenance crews to familiar with the expert system software and easy to use. These will speed up the inexpert maintenance crews to correct the problem as fast as the expertise does. For example, such a case that the generator cannot synchronize to the grid that took the inexpert maintenance crews about 43 hours to correct the problem which causes a loss income of about 1,036,080 THB. With the help of the studied expert system, the inexpert maintenance crews could take less than 4 hours to correct this problem and could save a loss of about 900,000 THB.