An Expert System for One Phase Induction Motor Diagnostics Kebri Kein Moudy Pajung
Universitas Pendidikan Indonesia
Abstract
This research is to study the possibility of developing an expert system application program for system diagnostics. The case of one phase cage rotor induction motor is studied as an example. Various conditions of a one phase induction motor are measured using 5 (five) sensors including a tachometer for measuring the motor speed, a thermometer for measuring the stator temperature, a voltmeter for measuring the line voltage, an ampere-meter for measuring the line current and a microphone for measuring the sound level. The measurement of these five operating parameters will indicate the conditions of 3 (three) motor elements, i.e. the bearing, the main winding and the auxiliary winding. Each element has 3 (three) possible abnormal conditions, so the total diagnostics covers 512 possibilities. The measurement of each sensor is categorized into 5 (five) possible level, i.e. very low, low, medium, high and very high, using a fuzzy logic algorithm. Thus, there are 3125 possibilities of symptoms to make an expert system with 3125 inputs and 512 outputs. A new knowledge can be added to the database whenever a new symptom is produced by a laboratory experiment or a field experience.
