Knowledge Representation Issues Pertaining to Mine Ventilation KBS Development

"Engineering design involves, in addition to numerical calculations, considerable amount of expertise and judgement. The expertise and judgement is required to conceptualize and idealize the problem, select a suitable mathematical model for the problem and to interpret the model results in terms of stipulated design constraints. In recent years, knowledge based systems (KBS) have provided the capability to integrate the qualitative experiential and expert knowledge with ~he quantitative knowledge for improved analysis and design of systems. A possible KBS architecture for mine ventilation system design was discussed by Ramani and Prasad · (1987). The expert system (ES) component of the KBS supports the analysis and interpretation phases of the ventilation design process with an algorithmic program used for the solution of alternate designs. The purpose of this paper is to discuss the data and knowledge needs for developing such a system and a formal means of their representation on the computer. A number of ES shells were evaluated to explore possible development and delivery environments for the system. A brief summary of this evaluation is also provided.INTRODUCTIONLuxbacher and Ramani (1979) have discussed the design process of mine ventilation systems and have provided the interrelationships that exist between ventilation, strata control, drainage and material handling systems and mining methods. They have also outlined the importance of correct analysis of the problem, choice of the right solution method and a rigorous interpretation of the solutions. The expertise required for analysis and interpretation phases was identified as the key element in developing effective and efficient ventilation system designs. Such expertise is embedded in the right choice of questions to be asked of the user as well as in directing the user to a wide array of alternative methods that may satisfy the design requirements. With advances in artificial intelligence applications in recent years, particularly expert systems, the integration of expertise into traditional mine planning and design algorithms has been facilitated. These planning and design aids, employing the integration of numeric and symbolic logic and called coupled systems, are powerful tools for engineers to tap the benefits of experts (Kitzmiller and Kowalik, 1985; Fenves, 1985). The purpose of this paper is to discuss the means of representing knowledge for the design of mine ventilation systems through the use of a coupled system. The KBS is visualized as consisting of the following components:"