中文摘要：

利用高灵敏度霍尔器件，设计研制了多通道阵列式漏磁检测传感器及信号处理电路．对不同几何参数的铁磁性试件缺陷进行了检测实验研究，该漏磁检测系统可实现地磁场激励和人工弱磁激励下的缺陷信号图像显示．探讨了基于多通道漏磁信号的缺陷表示方法，并利用人工神经网络技术对基于多通道传感器漏磁信号的缺陷反演问题进行了初步研究，表明利用霍尔元件阵列检测装置和人工智能信息处理方法，可以实现多通道漏磁信号与缺陷参数的非线性拟合，进而实现漏磁检测中的缺陷定量化分析．

英文摘要：

By using Hall devices with high sensitivity, multi channel array magnetic flux leakage sensor and its signal processing circuit were designed and developed. The experiments of testing defects in ferromagnetic samples with different parameters and sizes were carried out. The developed magnetic flux leakage detection system realized the image display of defect signals under geomagnetic excitation and artificial weak magnetic excitation. The display method of subject defects based on signals from testing system was discussed. In addition, the nonlinear inversion between flaw parameters and tested signals could be mapped by using the technique of artificial neural network, that could be used to quantitatively analyze the subject flaw.