基于涡流效应的线圈阵列传感器设计与神经网络检测算法

The Design of Coil Array Sensor Based on Eddy Current Effect and Neural Network Detection Algorithm

针对传统混凝土内部检测方式,只能测量混凝土保护层厚度,而无法同时对钢筋直径进行有效测量的问题,提出了一种基于脉冲涡流阵列的新型检测装置。使用ANSOFT仿真软件对检测模型进行仿真分析,并通过试验,验证了该检测模型对于混凝土保护层厚度以及钢筋直径具有较好的响应效果。通过BP神经网络建立了线圈信号和保护层厚度、钢筋直径之间的数学模型。试验结果基本满足精度要求,弥补了传统涡流检测仪器无法同时完成对保护层厚度与钢筋直径测量的不足。

In respect that the diameter of steel bar inside of concrete could not be measured effectively in traditional method,a new type of detection device based on pulsed eddy current array was proposed to detect both the thickness of concrete protective layer and the diameter of steel bar simultaneously.The detection model was simulated by ANSOFT simulation software,and physical experiments verified that the model had a good response to the thickness of the concrete protective layer and the diameter of the steel bar.Then,the BP neural network was used to establish the mathematical relationship of the coil signal to the thickness the protective layer and the diameter of the steel bar.The result of neural network experiment satisfied with the requirement basically.The new type of detection device could measure both the thickness of concrete protective layer and the diameter of steel bar at the same time,which the traditional eddy current detection equipment could not complete.