论文部分内容阅读
ＥＳＷＣＭ的建立是在分析处理近１００万个实验数据和５００条知识数据的基础上完成的。ＥＳＷＣＭ把模型技术和专家系统技术有机结合起来，是一个基于模型的专家系统，ＥＳＷＣＭ在组织结构上由自然、社会、经济数据库系统（ＤＢＳ）、知识库系统（ＫＢＳ）、模型库系统（ＭＢＳ）、推理机（ＩＥ）、人机界面（ＩＣＥ）及系统的维护程序等部分组成。知识表达采用了规则、框架、逻辑谓词和模型等方式，并采用了基于内存和基于文件的数据结构存储；模型库包９种类型７２个，用以描述结构性较强的问题，克服了领域专家对预测方面的困难；推理机全部采用Ｔｕｒｂｏ－Ｐｒｏｌｏｇ、Ｔｕｒｂｏ－Ｃ语言设计，使用了元级控制、数据驱动、目标驱动和混合控制策略，并采用了知识规则的优先排序和专家启发式的冲突消解策略。程序设计时，引进了软件工程的概念。软件各个模块可在一个集成环境下运行，同时又可单独使用。通过分析ＥＳＷＣＭ的决策方案的可行性和实施效果，对ＥＳＷＣＭ决策方案的科学性和实际应用的可能性进行了评价。
The establishment of ESWCM is based on the analysis and processing of nearly 1 million experimental data and 500 knowledge data. ESWCM combines model technology with expert system technology. ESWCM is a model-based expert system that is composed of natural, social, economic database systems (DBS), knowledge base systems (KBS), model library systems (MBS) , Reasoning machine (IE), human-machine interface (ICE) and system maintenance procedures and other components. Knowledge representation uses rules, frameworks, logical predicates and models, etc., and uses memory-based and file-based data structures for storage. The Model Library contains 9 of 72 types for describing structurally strong problems and overcoming areas Experts predict the difficulties; reasoning machine using Turbo-Prolog, Turbo-C language design, the use of meta-level control, data-driven, target-driven and hybrid control strategy, and the use of knowledge rules prioritization and expert-inspired Conflict resolution strategy. Programming, the introduction of the concept of software engineering. Each module of the software can run in an integrated environment, at the same time can be used alone. By analyzing the feasibility and effectiveness of ESWCM’s decision-making plan, the feasibility of ESWCM’s decision-making scheme was evaluated in terms of scientificity and practical application.