基于BP神经网络对BT项目风险的评价

BT融资模式是加快城镇化发展过程中,大型基础设施项目建设的主要融资方式,准确地对BT工程项目的风险等级进行评价,有助于降低投资人的风险损失。为了更好地评价BT工程项目的风险等级,克服传统风险分析方法的不足,建立BP神经网络模型。本文以A市地铁BT工程项目为例,首先建立该项目的风险评价指标体系,然后确定合适的BP神经网络模型参数。在大量经验数据的基础上,对BP神经网络模型进行了训练与检测,通过试验发现构建18-17-1三层BP神经网络模型对风险等级的评价结果精度最高。最后通过此模型对A市地铁BT工程项目的风险等级进行了评价,取得较好的效果。

Evaluation of BT Project Risk Based on BP Neural Network

BT financing mode is the main financing large infrastructure projects in the rapid development of urbanization. Accurate evaluation of BT project’ s risk helps to reduce the investor’ s risk loss. In order to better evaluating the BT project’ s risk level and overcoming the shortcomings of the traditional risk analysis methods, a BP neural network model has been used in this paper. BT project’s risk level of a city’s subway as an example in the paper. Firstly,establish the project risk evaluation index system and determine the appropriate parameters of BP neural network . Meanwhile, based on a large amount of empirical data,the BP neural network model is trained and detected. The experiment showed that the evaluation results on the risk which base on the construction of 18-17-1 three layers BP neural model has a high accuracy. Finally,BT project’ s risk level of a city’ s subway was evaluated by the model and good results have been received.