Development of an anticorrosion alloy selection expert system

To help engineers and designers efficiently and precisely to select the optimal anticorrosion alloys (ACAs), this work is carried out to develop an expert system with the Hierarchical Case-Based Reasoning (HCBR) approach which simulates the logics of the materials selection process from the lower precious embodiment stage(materials class) to the finally precious detail stage (materials member). In this study, the key to the establishment of an expert system is to acquire knowledge from the domain experts or documents to help identify the materials. Knowledge engineers use the knowledge acquisition approach (e.g. decision matrices and k-means methods) to analyze quantitative (e.g. mechanical or physical properties) and qualitative (e.g. corrosion resistance, environment) properties to each class/member of ACA. This acquisition approach allows format the knowledge into the forms that can be effectively adopted by computer programming. The developed system is then demonstrated by using the real cases from China Steel Corporation (CSC). At the embodiment stage the user is guided through a preliminary questionnaire, the feasible materials sets are first screened. As the user completes more detailed questions the ranked detailed materials from the screened materials sets are then proposed to the users.