与研制新型装备相比,现役装备改进具有成本低、周期短的优势,但仍存在一定的风险。采用BP神经网络评价方法,利用MATLAB的神经网络工具箱,把非线性的风险评价原理模型化,给出了评价的学习算法,并通过测试数据验证了模型的可靠性。采用BP神经网络模型,不但解决了非线性数据方面的问题,而且通过训练模型的训练和建立,找出输入与输出的对应关系,减弱了人的主观因素,相对于其他方法更客观可信。 Compared with the development of new equipment, existing equipment improvements have the advantages of low cost and short cycle, but there are still some risks. Using BP neural network evaluation method, the neural network toolbox of MATLAB was used to model the nonlinear risk assessment principle. The learning algorithm of evaluation was given. The reliability of the model was verified by the test data. The BP neural network model not only solves the problem of non-linear data, but also finds out the corresponding relationship between input and output through the training and establishment of training model, weakens people’s subjective factors and is more objective and credible than other methods.