基于LMBP神经网络的国际工程项目风险评价研究

Research on international engineering project risk evaluation based on LMBP neural network

国际工程承包项目面临的风险因素众多,给项目风险评价带来了很大的难度,积极寻求合理有效的方法对国际工程风险进行评价尤为重要。通过建立国际工程项目风险评价指标体系,对传统的BP神经网络模型进行改进,提出LMBP神经网络模型,避免了其难收敛和收敛时间长的弊端,建立LMBP神经网络评价模型,用模糊综合评价方法验证其准确性和可靠性。最后以马来西亚的一个项目为例,实证其风险评价的实用性。

The various risk factors exist in the international engineering contracting project,and bring great difficulties to risk evaluation of the project. It is particularly important to seek the reasonable and effective methods for international engineering project risk evaluation. The index system of the international engineering project risk evaluation is established to improve the traditional BP neural network model. The LMBP neural network model is proposed to avoid the drawbacks of difficult convergence and long convergence time. The evaluation model of LMBP neural network was established,and its accuracy and reliability were verified with fuzzy comprehensive evaluation method. Taking the project in Malaysia as an example,its practicability of risk evaluation was demonstrated.