个体动态空间决策及其驱动下的城市空间增长模拟研究

Simulating urban sprawl by modeling individual’s adaptive spatial behavior from micro-perspective is an important tendency of urban sprawl simulation. It is where GIScience focuses to contribute to urban study and support sustainable urban planning. Geographers have paid much attention to modeling individual decision-making dynamic process as it is one of the key processes in urban sprawl micro-simulation. Facing the lack of knowledge of individual spatial decision-making dynamics, modeling methodology, and complete model system, we focus on modeling the human-environment interaction and modeling the interaction between individuals to interpret and simulate the dynamic human spatial decision-making process. We learn individual decision-making dynamic mechanism by using Piaget’s recognition development theory, social learning theory and artificial computing method et al. We construct series of models to evaluate land use utility value and its dynamics, to simulate multiple temporal scale learning behavior, to analyze the spreading social scope of individual learning and decision-making, and to simulate individual spatial decision-making dynamic process by coupling these multi-process models with support of GIS data. Furthermore, we depict the response of individual decision-making to urban land use planning. Finally, we propose and develop urban sprawl micro-simulation model by coupling our decision-making dynamic models with land use conversion model and other models. We apply calibrated model and decision-making model to Nanjing city and Changzhou city to test their validity. The study is expected to contribute to the theory and methodology of individual spatial decision-making modeling and to the perfection of urban sprawl micro-simulation framework, and consequently to assist rational urban planning and implementation.

从微观个体适应性行为过程的视角研究和建模城市空间增长是地理信息科学推动城市空间研究、支撑空间规划的重要方向，其核心是个体动态空间决策过程的建模。针对个体空间决策模拟存在的机制认知不足、模型描述不完整、模型中决策模式静态单一等问题，研究拟从人—地交互、人—人交互过程入手，借鉴认知发展、社会学习、智能计算等理论方法，深入解析个体空间决策动态过程，构建个体动态空间决策模拟模型，重点研究个体认知环境变化的土地利用效用评价模型、多时间尺度的个体学习模型、个体知识和决策传播网络模型、集成多过程模型的个体空间决策建模等关键技术方法。通过集成个体空间决策模型，土地利用转换模型，空间规划响应分析方法等实现城市空间增长微观动力模拟模型，并开展模型应用与评价。预期成果能够支持城市空间增长模拟方法的发展，服务于我国新型城镇化背景下城市与区域规划和发展战略的编制。

建模个体的动态决策过程，推演由此驱动的宏观空间格局是理解城市增长微观动力过程的重要手段，能够为相关空间规划、基础设施布局等提供支撑。然后，当前个体空间决策建模和城市增长模型都面临着机制认知不足、模型描述不完整、考虑因素有欠缺等问题。研究针对上述不足，关注个体学习—决策动态建模、决策环境影响因素拓展等方面，一方面实现了个体拓展学习模型及其支撑的动态决策模型和城市增长模拟模型；另一方面，面向城市化带来的生态环境问题，建立生态格局、生态环境质量评价模型，实现了考虑生态安全要素的城市增长模拟模型。上述模型在南京、常州、深圳等研究区应用实验中取得了良好的模拟效果，研究成果已初步拓展至城市群空间增长模拟。本项目实施过程中发表论文9篇，其中SCI/SSCI检索论文6篇；登记软件著作权1项，获省部级科技奖励2项。