基于人工神经网络和证据理论的电梯健康状态评估

Health Status Evaluation of Elevator Based on Artificial Neural Network and Evidence Theory

针对目前电梯安全评估结果主观性大的问题,建立了一种基于人工神经网络和证据理论的电梯健康状态评估模型。选取了电梯各个子系统具有代表意义的特征量。基于人工神经网络构建了径向基函数神经网络,应用D-S证据理论对各子神经网络的输出结果进行融合,得到电梯的健康状态结果。实例分析验证了该模型的可行性和有效性。

According to the problem that safety evaluation result is subjective,an evaluation model for the elevator health status is established based on artificial neural network and evidence theory.This paper selects the representative characteristic parameters of each elevator subsystem.The detectior values are processed by radical basis function neural network,and Dempster-Shafer evidence theory was applied to refuse the output results of all sub neural networks,then the evaluation result is obtained by using the method.The example analysis~ showed that the model is correct and effective