【作者】 韩晓军；

【导师】 陈守煜；

【作者基本信息】 大连理工大学 ， 水文学及水资源， 2008， 博士

【摘要】 水是生命活动的物质基础,是人类赖以生存、发展的基础性自然资源。作为战略性的经济资源,水资源始终是影响人类社会发展的重要因素。因此,水资源科学在认识自然、改造世界的斗争中,有着重要的意义和广阔的前途。现实的水资源系统是一个具有社会性、自然性、开放性和动态性的复杂巨系统。影响水资源系统的因素很多,涉及到自然、社会、经济、技术等多个相互联系但又相互制约的因素,这些因素的不确定性与模糊性使得经典数学理论难以完善地解决水资源系统中的实际问题。本文在可变模糊集理论的基础上,研究将该理论的相关模型和方法应用于水文水资源系统的环境污染评价、污染负荷分配、水旱灾害评估、涝区分类研究、南水北调工程中黄河上、中游水量重新分配决策等各个方面。同时,结合应用对模型参数及其变化、模型拓展等方面进行了相应的研究,取得了一定的成果。主要研究内容成果概括如下:1.模糊可变评价模型参数研究及其在水环境质量评价中的应用研究模糊可变评价模型的参数,建议一种实用的确定M值位置及区间[c,d]范围值的方法。应用模糊可变评价模型及建议方法于水环境质量综合评价领域,通过哈尔滨江段的实例,对模型的理论研究成果进行验证,并结合模型参数α、p四种组合情况,综合四个子模型的评价结果,利用模型的可变性对级别特征值的稳定性进行分析,取得了较好的评价结果,一定程度消除了模型参数不同对评价结果的影响。2.水污染负荷总量分配方法研究在水污染负荷总量分配领域,由于影响因素众多,多数原有的理论和方法由于片面的依据经济效益或者公平原则难免会对分配方案的具体实施带来障碍。本文将总量控制理论与可变模糊集决策单元系统理论相结合,充分考虑了影响水污染负荷分配所涉及的社会、经济、环境和技术的多方面要求。将可变模糊集决策单元系统理论应用于水污染负荷分配问题中,提高了分配方案的可操作性,为污染负荷的合理分配提供了一种新方法。3.可变模糊聚类迭代模型在干旱程度评估中的应用结合可变模糊集的相关理论,考虑模糊聚类循环迭代模型中参数的可变特性,通过聚类模型权重参数的改变,分别进行聚类运算,相互印证了分类结果的稳定性,提高了评价结果的可信度。并结合水文干旱的实例,说明其可行性与实用性,为水文干旱程度评估提供了一种新的方法。4.基于模糊优选神经网络模型的辽宁省涝区分类研究指出涝灾与洪灾的不同,分析辽宁省的涝区现状,采用模糊优选神经网络模型,选取辽宁省35个县(区)的实测资料进行分类研究,既能用模糊概念来表达人的知识和经验,又可利用神经网络较强的学习能力,获得了比较符合实际的分类结果。为涝渍灾害的研究提供了一种新的方法。5.可变模糊集决策理论与方法在黄河置换水量分配中的应用研究在南水北调东、中线供水情况下,黄河下游置换出的水量在黄河上、中游地区的重新分配问题。依据可变模糊集理论与模糊集决策单元系统理论,对黄河置换水量进行分配决策,既考虑了经济发展、缺水状况等定量指标,又考虑了科技、人口、医疗等诸多难以量化的定性指标;把多地区、多层次、多指标复杂的黄河置换水量决策问题转化为对于一系列单元系统模糊可变优选问题的求解,所得成果可供南水北调中、东线工程黄河置换水量决策支持参考之用。6.模糊可变识别模型的拓展及其在清晰识别中的应用拓展模糊可变识别模型优化准则参数α的变化范围,使得模糊可变识别模型既可以应用在模糊识别,也可以应用在清晰识别问题中。在清晰识别问题中增强了识别结果的说服力,论述了模糊可变识别模型中相对隶属度,受模型参数α变化的影响的一般趋势,并结合实例论述了该拓展方法在遥感系统地物清晰识别中的应用。更多还原

【Abstract】 The water resource is the substance foundation of the life activity and one of the basic natural resources that the mankind depended on for existence and development. As an important strategic economic resource, water resource is always an important factor that influences the development of human society. So, water resource science is of great importance and wide future in the process of understanding nature and transforming the world.In fact, water resource system is a kind of complex and dynamic system, which have large amount of uncertain factor. The factors effecting water resource systems arise from nature, society economy and techniques. Several problems of water resource systems are uncertain and fuzzy. Traditional mathematics can’t be used to deal with a lot of fuzzy events. Based on variable fuzzy set theory, this paper concentrates on relative models and approaches for water resource systems. Some achievements have been made on model parameter variations and spreading models.The major contents and research results are as follows:1. Research on parameter variation of the fuzzy assessment model and its application for water environmental qualityA suggestion is provide to determine the position of M and range [c,d]. In the field of evaluating water environmental quality, this suggestion is used to solve practical problems. Four kinds of situations of model parameterαand p are considered verifying the model stability. The assessment result can help eliminate the effect of model parameters difference.2. Research on the method for allocation of water waste loadsIn the field of allocation of water waste loads, because of all kinds of influencing factors, many old methods and theories become unfortunate. Combining the total amount control theory and unit systematic theory of variable fuzzy sets decision-making, multi indexes of social, economic, environmental and technical system are fully considered to allocate the water waste loads. Operability of allocation plan is improved. A new method is provided for allocation of water waste loads.3. Research on a drought degree evaluation method based on the interactive computing model of variable fuzzy classification Based on the variable fuzzy set theory, parameter variation of the interactive computing model of variable fuzzy classification is considered to verify the model stability. The reliability of assessment result can be provided. A new method is provided for the drought degree evaluation too.4. Research on the classificatory of Water Logging Regions in LiaoNing Province based on neural network model of fuzzy optimizationIt is pointed out that waterlog is different from flood. The neural network model of fuzzy sets is proposed in this paper for the classificatory calculation of 35 water logging regions in Liaoning province. It expresses people’s knowledge and experience with fuzzy concepts and utilizes the strong learning ability of neural network. The result is objective. A new method is provided to study waterlog.5. The application of variable fuzzy sets decision-making theory and method to replaced water of the Yellow RiverA problem is studied that a part of water diverted from Yellow River to the areas served by the east-route and mid-route of the South-North Water Transfer Project may be reallocated to the upstream and midstream areas of the Yellow River after commissioning of the east-route and mid-route and be-fore commissioning of the west-route of the South-North Water Transfer Project. The method fully takes account of not only quantitative indexes but also qualitative indexes. The complicated decision-making problem of the water reallocation is predigested to the unit system series. The result can be used in practice.6. The expansion of variable fuzzy pattern recognition model and its application to clear recognitionThe range ofαcan be expanded on condition that the recognition is clear. This idea can expand the applicability of variable fuzzy recognition model. The changing trend has been found about relative membership degree. This trend is important for variable fuzzy recognition model, the example can specify this idea.更多还原