基于BP算法的石化设备可靠性计算及评价模型

Reliability Evaluation of Petrochemical Equipment Model based on BP Algorithm

提出了一种基于BP算法的石化设备可靠性建模与评价新方法,利用BP算法对复杂非线性系统的拟合能力,通过网络训练自适应地调整单一预测模型的权重,并应用MATLAB神经网络工具箱实现了基于BP算法的石化设备可靠性评价软件。结合兰州石化设备管理与预警系统实时监控数据,将该方法应用于计算某石化设备通道的故障概率和评价设备的整体可靠性,通过评价各种训练方法的学习效率,评价结果对比表明该方法具有平均计算时间短和收敛快的优点,在设备故障和可靠性评价中具有广泛的应用前景。

This article proposes a new model to evaluate the reliability of petrochemical based on BP algorithm.By taking advantage of BP algorithm to approximate complex nonlinear systems,adjusting weight of single forecasting model,furthermore,and by using MATLAB Neural network toolbox,the Reliability Evaluation Software of Petrochemical Equipment Model based on BP Algorithm is realized.The method is applied to Compute malfunction probability and reliability of a petrochemical equipment by means of real-time monitoring data of Lanzhou Petrochemical facility management and early warning systems to evaluate the learning efficiency of all training methods.The results shows that the average time of this method is short and the advantages of fast convergence.Therefore,this model of reliability evaluation can be widely applicable.