A performance evaluation of three inference engines as expert systems for failure mode identification in shafts

• We examine performance of expert systems in failure mode identification.
• Increasing symptoms and evidence will increase complexity of diagnosis.
• Bayesian inference had better performance for failure mode identification.
• Procedure could be used for evaluation of other failure analysis expert systems.

This paper aims to present performance evaluation of three different inference engines (rule based reasoning, fuzzy based reasoning and Bayesian based reasoning) for failure mode identification in shafts. This research was done with a focus on the validation cases and results after their use in failure cases from several industries where the three systems were tested under the same conditions.Each system was implemented using the same user interface and knowledge base, with different frameworks and techniques as follows: rule based inference reasoning (prolog, C#), Mamdani-fuzzy based reasoning (C, MATLAB®) and Bayesian based reasoning with a variable elimination algorithm (C, MATLAB®).The best performance was obtained using the Bayesian inference engine. The conditional probabilities give flexibility when evidence is not listed, while the fuzzy and classical IF-THEN systems depend on the rules in the inference engine.The process presented in this paper could be used for validation of any expert system or for comparison with other expert systems (inference engines) when the knowledge base is the same.