Pattern recognition of acoustic sea-bed profiling records (part 2:a fuzzy logic reasoning expert system)

An expert system based on the fuzzy set theory has been developed for the computer-basedpattern recognition system of Acoustic Sea-bed Profiling Records.The standard pattern characteristicstate sets appearing on the ASPRs corresponding to 10 geological categories of marille sediment layershave been defined by experienced specialists,and the weight of each pattern characteristic-state in determining the geological classification calculated through a detailed analysis.Through inquiring aboutthe pattern characteristic-state appearing on an ASPR being interpreted and then calculating the anproximating factors of this ASPR to the standard ones corresponding to the 10 categories of sedimentlayers respectively,the geological classification of the sedimellt layer can be determined by selecting themaximum approximating factor.