基于多智能体(Multi-agent)之间的动态信息反馈提出了交互式多智能体城市空间规划模型。在模型中,多智能体之间可即时交流信息,并经南反馈与协调等行为获取一致决策意见,最终得到各方可接受的城市规划方案。以长沙市岳麓区内建设项目选址为例,开展了模型实验。研究结果表明:多智能体交互式城市空间规划模型能够较好的模拟不同情景下参与城市空间规划的各类Agent之间的交互行为,并可提供多角色参与城市规划时空间冲突问题的解决方案。
In order to improve public participation in urban spatial planning and properly deal with spatial conflicts among all kinds of actors participating in urban spatial planning, an interactive urban spatial planning model based on multi-agent systems and dynamic information feedback mechanism is proposed. The model consists of a virtual urban spatial system (planning areas) and a social system (multi-agent systems). Agents participating in urban spatial planning can exchange knowledge and informa- tion with each other frequently, and finally achieve a consistent decision to form an acceptable urban planning scheme through feedbacks and coordinations. The model is further applied to simulate site selection of construction project in Yuelu District of Changsha City and results show that this model not only better simulates the interaction among various agents under different scenarios, but also provides solutions to spatial conflict in the process of multi-actor urban planning.