自改革开放以来，中国的城市化经历了前所未有的快速发展。城市化推动中国经济取得了举世瞩目的巨大成就，同时也造成了诸多严重的环境、生态问题。在城市扩张的过程中，大量的农业用地如耕地、园地和生态用地如林地和草地被城市建设用地所侵占。土地滥用现象普遍存在，导致土地资源浪费，后备土地资源严重不足。作为人类生活的载体，土地生态系统不仅为人类的生产，消费提供了一系列原材料和物质基础，而且提供了诸如气候调节，废物处理，水土保持，娱乐休闲，观光旅游等一系列不可替代的非物质性服务。然而，由于城市化过程中粗放的土地利用模式，导致土地生态系统不断降级，所提供的生态系统服务逐年减少。虽然将生态系统服务价值纳入土地利用决策中的重要性已得到广泛的认知，但将生态系统服务价值纳入到土地利用规划和管理中的方法严重不足。鉴于此，本文主要进行了以下几个方面的研究： 从土地生态系统所提供的生态服务的角度出发，利用生态系统服务价值评估理论，评估了不同土地生态系统所提供的生态服务的价值。然后将其纳入到土地利用规划中，建立数学模型，优化土地利用结构，实现土地生态价值最大化。具体而言，本文首先建立了灰色预测模型，预测了土地利用的变化，并检验了预测模型的可靠性。然后根据所建模型的预测结果，构建灰参数线性规划模型进行土地利用结构优化。最后，将所建立的模型应用在实例研究中，以验证所提出方法在土地利用规划中的可行性和适用性。本文比较分析了优化的土地利用结构与政府土地利用规划的优劣。分析了不同土地利用方案对生态系统服务价值的影响。结果表明，将灰色预测模型与灰参数线性规划结合起来进行土地利用规划是一种有效的方法。优化后的土地利用结构与政府的规划相比，可以增加生态用地如林地、园地和草地的面积，也可以增加土地生态系统所提供的生态系统服务的价值。从而有利于维持地区生态系统的完整性和生态系统组成成分的稳定性。 从土地利用综合效益的角度出发，为解决快速城市化下偏重土地利用的经济效益而造成土地利用结构不平衡的问题，本文综合考虑了土地利用的经济效益、社会效益和生态效益，建立了多目标的土地利用规划模型。同时，为处理土地利用规划中的不确定因素，本文将模糊规划和多目标规划结合起来并纳入到一个优化框架中，建立了模糊多目标线性规划(FMLP)模型，并且简化了模型的求解过程，提高了模型的可行性。最后选取了一个典型的地区-武汉市作为实例，将模型应用在武汉市的土地利用规划的研究中。结果表明，与政府制定的土地利用规划相比，优化的土地利用结构在增加土地利用的经济效益和生态效益中优势明显。本研究所建立的模型不仅可以实现不同土地利用类型的社会效益、经济效益和生态效益的整合，也能够处理土地优化过程中的一些不确定性因素。因此，FMLP模型是土地利用规划中一种有效的方法，可以应用在土地利用结构不平衡的城市或地区，以优化土地资源配置。 基于前面的研究，为协调经济系统和自然生态系统之间的土地资源配置，本文将经济系统和自然生态系统耦合为一个复合系统，进行土地资源优化配置研究。为处理规划过程中大量的不确定性因素，以及分析不同的政策情景，本文根据系统工程理论，建立了模糊两阶段随机规划(FTSP)模型，并且提出用模糊函数排序法处理模糊集，该方法可以减少传统模糊集求解方法中附加变量和约束的数量，从而简化模型求解过程，增强模型的可行性。最后，本文选择了一个经济发展与自然生态系统失衡的地区武汉市，作为案例，应用FTSP模型进行武汉市所有工业园区和湿地生态系统之间土地资源可持续利用的规划研究。本文引入了土地交易作为一种生态补偿机制来平衡经济系统和自然生态系统之间土地资源的分配，考虑了一系列工业污染物的排放，也考虑了一系列生态环境因素如降雨径流系数、湿地保湿系数、植被覆盖率等。此外，本文将生态系统服务价值评估纳入到优化过程中，评估了11种生态系统服务的价值。根据不同的污水排放标准，设置了五种减排方案，分析了不同的政策情景下，工业结构、经济效益、生态系统服务价值、环境成本和废水排放之间的关系。结果表明，FTSP模型是经济-自然复合系统土地可持续利用规划中一种有效的方法，它不仅可以处理规划过程中的不确定性和随机性，也能够分析当预期目标被违反时，不同的经济处罚变化的情景。研究结果也表明，土地交易是一种有效的生态补偿方式，它有助于将不同土地利用模式的外部性内部化，从而有效地抑制过度开发活动。此外，研究结果也为该地区的工业发展和湿地生态系统保护提供了一些现实启示和实际建议。 总之，本文在中国快速城市化的背景下，响应国家生态文明建设的重大战略要求，从节约土地资源，保护生态环境的角度出发，评估土地生态系统所提供的生态服务价值，并将土地的生态价值纳入到土地综合效益的评估中，进行土地利用规划研究。本文从不同的角度，建立数学模型，优化土地利用结构，实现土地资源的最优配置。以平衡经济发展与自然生态环境之间的矛盾，丰富城市可持续发展和人地关系研究的理论内涵，为我国土地资源的可持续利用提供理论依据和实际政策导向。 关键词：生态系统服务价值；土地利用规划；城市化；不确定性。

 Since the reform and opening-up policy was implemented, the urbanization in China has experienced unprecedented rapid development. Urbanization has pushed China's economy to achieve great achievements that have attracted worldwide attention. Meanwhile, it has also caused numerous serious environmental and ecological problems. In the process of urban expansion, a large number of agricultural lands (e.g. cropland, orchard) and ecological lands (e.g. woodland, grassland) are occupied by built-up lands. The abuse of lands is widespread, resulting in a waste of land resources and a serious shortage of land resources reserve. Land ecosystems not only provide a series of raw materials and material foundations for human production and consumption, but also provide a series of irreplaceable non-material services such as climate regulation, waste treatment, recreation and tourism. However, due to the extensive land use pattern in the process of urbanization, the land ecosystems have been continuously degraded, and the provided ecosystem services have been reduced year by year. In spite that the importance of integrating ecosystem services values into the decision-making of land use has been widely recognized, the method of ecosystem service values integrated into land use planning and management is seriously inadequate. In view of this, this paper mainly investigates the following aspects: From the perspective of the ecological services provided by the land ecosystems, the ecosystem service values provided by different land ecosystems are evaluated. Then the ecosystem service values are integrated into land use planning, and a mathematical model is established to optimize the land use structure so that the ecological values of the lands can be maximized. In detail, firstly, a grey prediction model is established in this paper, by which the changes of land use can be predicted, and the accuracies of the prediction model are tested. Then based on the predicted results of the grey prediction model, a gray parameter linear programming model is constructed to optimize the land use structure. Finally, the established model is applied into a case study to verify the feasibility and applicability of the proposed method in land use planning. The advantages and disadvantages between the present optimized land use structure and government's land use planning are compared in this paper. The impacts of different land use plannings on ecosystem service value are analyzed. The results show that combining GM (1, 1) model with grey parameter linear programming model is effective for land use planning. Compared with government's planning, the optimized land use structure can increase the area of ecological lands such as woodland, orchard and grassland, and can increase the ecosystem service values provided by different land ecosystems. Therefore, it will help maintain the integrity of the regional ecosystem and the stability of the components of ecosystems. From the perspective of comprehensive benefits of land use, in order to solve the problem of unbalanced land use structure caused by more emphasis on economic benefits of land use under rapid urbanization, a multi-objective land use planning model is established in this paper with comprehensive consideration of the social, economic and ecological benefits of land use. In addition, in order to deal with the uncertainties in land use planning, the fuzzy planning model and multi-objective planning model are combined into an optimization framework, and a fuzzy multi-objective linear programming (FMLP) model is established. Besides, the solving process is simplified and the feasibility of the FMLP model is improved. Finally, a typical area (Wuhan) is selected as a case study to apply the FMLP model to the land use planning in Wuhan. The results show that, compared with the land use planning formulated by the government, land use structure optimized by the FMLP model has obvious advantages on increasing the economic and ecological benefits of land use. The FMLP model established in this study can not only achieve the integration of social, economic and ecological benefits for different land use types, but also deal with uncertainties in optimization process. Therefore, the FMLP model is an effective method in land use planning and can be applied to other cities or regions with unbalanced land use structure to optimize land resource allocation. Based on the previous researches, in order to coordinate the land resources allocation between economic system and natural ecosystem, the economic system and natural ecosystem are coupled into a composite system in this paper to optimize land resources allocation. In order to deal with a large number of uncertainties in planning process and analyze different policy scenarios, a fuzzy two-stage stochastic programming (FTSP) model is established in this paper based on system engineering theory, and a fuzzy function ranking method is proposed to deal with fuzzy sets. This method can reduce the number of additional variables and constraints in the traditional fuzzy set solving methods, thus simplifying solving process and enhancing the feasibility of the FTSP model. Finally, a region (Wuhan) with serious unbalance between economic development and natural ecosystems is selected as a case study in this paper. The FTSP model is applied to planning the sustainabilities of land resources of all industrial parks and wetland ecosystems in Wuhan. Land trading is introduced as an ecological compensation mechanism to balance the land resource allocation between industrial parks and natural ecosystems, with consideration of a series of industrial pollutant emissions, as well as a series of ecological environmental factors such as rainfall runoff coefficient, vegetation coverage rate and so on. In addition, the ecosystem services value assessment was incorporated into the optimization process and 11 types of ecosystem services were evaluated. According to sewage discharge standards for different industries, five kinds of emission reduction schemes were proposed to analyze the relationships between industrial structure, economic benefits, ecosystem service values, environmental costs and wastewater discharges under different policy scenarios. The results show that the FTSP model is an effective method for sustainable land use planning in economic-natural composite system. It can not only deal with the uncertainty and randomness in the planning process, but also analyze scenarios under different economic penalties when the expected targets are violated. The results also show that land trading is an effective way on ecological compensation, which helps to internalize the externalities of different land use patterns, thus suppressing excessive development activities effectively. In addition, the findings provide practical implications and suggestions on industrial development and wetland ecosystem protection for local policy makers in Wuhan. In summary, responding to the major strategic requirements of national ecological civilization construction, the values of ecological services provided by land ecosystems are evaluated, and the ecosystem service values are incorporated into land use planning. Several mathematical models are established from different perspectives to optimize the land use structure. Three major goals are committed to be achieved: Balancing the contradiction between economic development and natural ecological environment; enriching the theoretical connotation of urban sustainable development; providing theoretical basis and practical policy guidance for the sustainable use of land resources in China. Key words: Ecosystem service value; Land use planning; Urbanization; Uncertainty.