由超声脉冲法获得４种检测信息———声速、波幅、频率和波形，从中提取３个特征量：平均声速、声速的离差系数和波幅的离差系数，利用模糊分析方法和人工神经网络相结合对ＧＲＣ、加气混凝土、空心砖、实心砖、砌块５种墙体材料进行种类识别，形成一个智能化的、具有类似人脑功能的系统．经现场实测证明，判别的准确率达９５％以上 Four kinds of detection information—sound speed, amplitude, frequency, and waveform—were obtained by the ultrasonic pulse method. Three feature quantities were extracted: average speed of sound, coefficient of variation of speed of sound, and coefficient of dispersion of amplitude, using fuzzy analysis methods and artificial neural networks. Combining GRC, aerated concrete, hollow bricks, solid bricks, and block types of wall materials for species identification, an intelligent system with human brain function is formed. The field test proved that the accuracy of the judgement is over 95%.