基于机器学习的CT设备故障自动化识别系统

CT Equipment Automatic Fault Automation Identification System Based on Machine Learning

CT设备由于故障点数量存在不确定性,导致故障识别率低,提出基于机器学习的CT设备故障自动化识别系统。根据故障特征,设计系统硬件的框架和功能模块;软件单元设计根据CT设备正常状态信号分布,设定故障判定依据,采用神经网络算法,检测和划分故障信号的类型,结合大数据融合算法,优化故障识别系统,输出识别结果,实现CT设备故障自动化识别。实验结果表明:设计的故障自动化识别系统将电压波动控制在0值上下,峰值控制区间在[-0.5,0.5],并且最低故障识别准确率达到了90.3%,准确识别和划分了故障类型,因此,设计系统有效提升了故障识别的准确率和抗干扰能力,运行稳定性好。

Due to the uncertainty of the number of fault points in CT equipment,the overall fault recognition rate of the system is low.According to the fault characteristics,it designs the framework and functional modules of the system hardware.According to the normal state signal distribution of CT equipment,the software unit design sets the basis for fault determination,uses neural network algorithm to detect and divide the types of fault signals,and combines with big data fusion algorithm to optimize the fault identification system,output identification results,and realize automatic fault identification of CT equipment.The experimental results show that the design of fault automatic identification system to control the voltage fluctuation zero value,peak control interval in[0.5,0.5],anti-jamming is strong,and the minimum fault identification accuracy reached 90.3%,and accurately identify and classify fault types,therefore,design system effectively improve the fault identification accuracy and anti-interference ability,and good overall stability.