Expert system for diagnosing damage of prestressed concrete pile

The damage of prestressed concrete pile (PCP) during the installation process has resulted in injuries, time delay, and cost overruns. Diagnosis of the damage is one of the most important problems in foundation engineering. This paper describes a diagnosis expert system PCPILE (prestress concrete pile) for diagnosing the damage of PCP during the construction process. Because there is no way to avoid subjective judgment, imprecise data, and experience-based knowledge in this problem, fuzzy logic is employed to solve it efficiently. In PCPILE, the inference is mainly implemented by the deductive inference mechanism with depth-first backward chaining. But, when the deductive inference mechanism cannot satisfy a diagnosis goal directly, a specific case base is opened for the analogical influence mechanism. The expert system is novel in at least two respects: (1) Case base is employed for knowledge representation; (2) analogical inference technique is applied when there is a lack of sufficient experience in getting reliable rules.