【摘要】 以准确、高效评估城市总体规划方案为目标,提出基于深度学习技术的城市总体规划方案评估方法。从气象环境可行性、环境影响评价构建城市总体规划评估指标体系,并通过专家打分、城市规划实况分析以及城市规划模拟等方式采集评估指标数据;采用深度学习技术构建稀疏编码器,将评估指标数据作为稀疏编码器的输入,通过学习训练输出城市总体规划方案评估结果,并通过迭代训练过程提升稀疏编码器输出精度。实例应用分析结果显示,所提方法能够准确找到评估城市总体规划方案,城市总体规划方案评估时间短、评估精度最高。更多还原

【Abstract】 A method of urban master planning scheme evaluation based on deep learning technology is proposed to obtain an accurate and efficient evaluation. In this method,the evaluation index system of urban master planning is constructed by taking account of the meteorological environment feasibility and environmental impact assessment first,and then the evaluation index data are collected by expert scoring and situation analysis and simulation of urban planning. Next,a sparse encoder is constructed with deep learning technology. The collected evaluation index data is taken as the input of the sparse encoder to output the evaluation results of the urban master planning scheme by learning and training. The output precision of the sparse encoder is improved by iterating the training process. Finally,the analysis results of example application show that the proposed method can accurately assess the urban master planning evaluation scheme. In addition,it has the shortest evaluation duration and highest evaluation precision among the evaluation methods.更多还原