【摘要】 针对电力通信网风险评估的影响因素繁多,复杂性和重要程度不同,建立了基于因子分析法与神经网络的风险评估模型;运用因子分析法对指标体系进行了分析和约简;考虑到历史的预测误差与未来预测误差的映射关系,利用RBF神经网络对未来的预测误差进行预测。实例验证结果表明,采用因子分析和神经网络结合的方法对电力通信网进行风险评估,减少了评估指标的数量,提高了评估的精度。更多还原

【Abstract】 Aiming at the situation that risk assessment on power communication network is affected by many factors with different complexity and importance,a risk assessment model is established based on factor analysis and neural network.By means of factor analysis,the analysis and reduction of index system is carried out.Considering the relation of historical prediction error and future prediction error,an error forecasting model is developed based on radial basis function(RBF) neural network.Validated results show that the risk assessment method on power communication network combining factor analysis and neural network can reduce the number of evaluation index,and the accuracy of evaluation results are improved.更多还原