A Fuzzy Logic-Based Automobile Fault Detection System Using Mamdani Algorithm

Abstract

Due to advancement and complexity of modern automobiles, fault detection has gone beyond manual or trial by error methods. The fault detection technologies in automotive industry is used to identify any potential or already existing fault in automobiles. Faults in automobiles are usually mechanical or electrical faults that may include airbag control unit, radiator, gearbox, transmission control unit, tyre pressure, brakes, air conditioner, cylinder casket, alternator, hubs malfunctions etc. Each fault has a specific or related sign and symptoms. There are several methods of fault detections in automobiles like the binary logic technique, the fuzzy logic method technique and artificial intelligence technique with different algorithms. In this research work, we employed a fuzzy logic based technique that uses a Mamdani Algorithm which presented a better fault detection mechanism. Mamdani’s algorithm was proposed by Ebrahim Mamdani as a fuzzy inference method which has a rule-bases that are more intuitive and easier to analyse and implement. Mamdani’s algorithm produces fuzzy sets that originate from fuzzy inference system’s output membership function for decision making. This research work is a web-based technology that was implemented using JavaScript, JQuery and SQL server, ASP.Net, Bootstrap 3.5 and CSS. The output of the system showed a greater improvement from other existing methods of fault detections in automobiles.