Focusing on the berth mud-filling phenomenon of ports, especiallymud ports, this dissertation starts from various data relevant to the factorsof berth mud-filling, and discusses about the application of Data MiningTechnology on the process of berth mud-filling knowledge discovery.Considering the diversity and hetero-structure of knowledge sources, amethod for the standardization of berth mud-filling knowledge, includingthe knowledge expression of Data Mining, is brought forward with theapplication of Extension Theory. Furthermore, the pattern of knowledgedecomposition concerning single-factor and multi-factors of berthmud-filling is defined in this dissertation, and the model of berthmud-filling knowledge level is built up.During the course of the formation of Knowledge Base System(KBS), this paper particularly illuminates the contents and functions ofknowledge base and inference mechanism in the KBS of berth mud-filling.Based on Fuzzy Cognitive Map (FCM) and considering the multi-level ofberth mud-filling knowledge, Extended Cognitive Map (ECM) is putforward to realize the inference process of Knowledge Base. Moreover,some problems about the maintenance, update, and interface of berthmud-filling KBS are discussed and the model of customer requirementsolving process is presented.About the application of berth mud-filling knowledge on themanagement of mud-cleaning engineering, this dissertation, based on theabove-mentioned berth mud-filling extended KBS, adds the simulation oflittle-probability or expectative affairs, such as the rebuilding the berths,to the strategies of the present mud-cleaning management, and thenintegrates the research on berth mud-filling with mud-cleaning andmud-reducing engineering. With some special case analyses, a feasiblemethod for transferring the present passive strategies of berth mud-fillingmanagement into active ones is provided, which shares great significancefor the berth mud-filling management of numerous mud-land and inlandports of our nation.
|