系统工程理论与实践
繫統工程理論與實踐
계통공정이론여실천
Systems Engineering—Theory & Practice
2013年
6期
1557~1562
,共null页
李月光 吴小萍 吕安涛 聂敏
李月光 吳小萍 呂安濤 聶敏
리월광 오소평 려안도 섭민
农村公路 养护管理评价 模糊神经网络
農村公路 養護管理評價 模糊神經網絡
농촌공로 양호관리평개 모호신경망락
rural road; maintenance management assessment; fuzzy-neural network
综合考虑影响农村公路养护管理的各要素,采用层次分析方法,建立了包含目标层、准则层、指标层的农村公路养护管理评价三级指标体系.针对综合评价中知识学习积累问题,研究了基于人工智能的模糊神经网络方法在农村公路养护管理评价中的应用.结合模糊理论和神经网络方法,采用模块化设计思想初步建立了农村公路养护管理评价的结构模型,在评估系统中嵌入专家知识,采用模糊理论对评价指标进行模糊化处理,再利用多层神经网络进行数值分析,最后将结果反模糊化,实现对农村公路养护管理系统的综合评价.同时通过实例说明了系统学习、系统评价的过程,实例验证了所建模糊神经网络模型的可行性与有效性.
綜閤攷慮影響農村公路養護管理的各要素,採用層次分析方法,建立瞭包含目標層、準則層、指標層的農村公路養護管理評價三級指標體繫.針對綜閤評價中知識學習積纍問題,研究瞭基于人工智能的模糊神經網絡方法在農村公路養護管理評價中的應用.結閤模糊理論和神經網絡方法,採用模塊化設計思想初步建立瞭農村公路養護管理評價的結構模型,在評估繫統中嵌入專傢知識,採用模糊理論對評價指標進行模糊化處理,再利用多層神經網絡進行數值分析,最後將結果反模糊化,實現對農村公路養護管理繫統的綜閤評價.同時通過實例說明瞭繫統學習、繫統評價的過程,實例驗證瞭所建模糊神經網絡模型的可行性與有效性.
종합고필영향농촌공로양호관리적각요소,채용층차분석방법,건립료포함목표층、준칙층、지표층적농촌공로양호관리평개삼급지표체계.침대종합평개중지식학습적루문제,연구료기우인공지능적모호신경망락방법재농촌공로양호관리평개중적응용.결합모호이론화신경망락방법,채용모괴화설계사상초보건립료농촌공로양호관리평개적결구모형,재평고계통중감입전가지식,채용모호이론대평개지표진행모호화처리,재이용다층신경망락진행수치분석,최후장결과반모호화,실현대농촌공로양호관리계통적종합평개.동시통과실례설명료계통학습、계통평개적과정,실례험증료소건모호신경망락모형적가행성여유효성.
An analytical hierarchy process (AHP)-based three-level assessment system (including target, criterion, and index) was established to cover all elements relevant to rural road maintenance management. This paper applies the artificial intelligence (AI)-based fuzzy neural network approach to the rural road maintenance management evaluation to handle problems related to knowledge acquirement and accumu- lation that is essential to comprehensive evaluation. Modular design, coupled with fuzzy theory and the neural network approach, was employed to tentatively develop a rural road maintenance management as- sessment model with built-in expert knowledge. Indexes in this model are initially fuzzified according to fuzzy theory, then analyzed in the multi-layer neural network, and conversely defuzzified to produce data that support and finalize the rural road maintenance management assessment. An example was given to illustrate the working mechanism of this assessment system, and to prove the feasibility and validity of the fuzzy neural network-based assessment model.