Abstract: The location of high⁃speed railway stations has always been in a contradiction. The government should not only reduce the impact of high⁃speed railway stations on people's lives, but also consider the cost of demolition. How to make the location of high⁃speed railway stations more scientific has become a problem that needs careful consideration. Therefore, the model of economic contribution degree was established by using linear regression analysis and grey prediction method to obtain the economic contribution degree of high⁃speed railway station to the city;based on the relevant data of high⁃speed railway in Liaoning Province, considering the construction cost, construction time, transfer convenience and economic contribution, a high⁃speed railway station location evaluation model based on principal component analysis was proposed and the evaluation model was used to analyze the siting of three high⁃speed railway stations in Liaoning Province. The results show that the siting of Shenyang North Railway Station is the most successful,and the passenger flow and distance from the city center are the important factors affecting the development of high⁃speed railway.

Key words: High?speed railway； Station location, Grey prediction, Principal component analysis, Machine learning

摘要： 高铁站的选址一直处于矛盾之中，政府既要减少高铁站对民众生活的影响，又要考虑拆迁成本，如何让高铁站的选址更加科学成为一个需要深思熟虑的问题。为此，运用线性回归分析法、灰色预测法建立了经济贡献率模型，得到高铁站对该市的经济贡献率；以辽宁省已建成的高铁相关数据为基础，从建设投入、施工耗时、换乘方便度、经济贡献等角度进行考虑，提出了一种基于主成分分析法的高铁站选址评价模型，并运用评价模型对辽宁省3个高铁站选址进行实例分析。结果表明，沈阳北站选址较为成功，客流量、距市中心距离是影响高铁开发的重要因素。

关键词: 高速铁路, 车站选址, 灰色预测, 主成分分析, 机器学习