对不同的试井解释模型，把压力导数曲线作为训练样本，应用ＢＰ网络进行训练，训练后的网络能根据现场的实际试井数据识别试井解释模型．本文用模拟的数据、不完整的数据、有噪声的数据和一个现场试井数据对这个ＢＰ网络进行了测试．结果表明人工神经网络能够正确地识别试井解释模型，也能识别不完整的、有噪声的数据．人工神经网络技术有效地改进了目前在试井解释模型中广泛采用的模式识别方法，是一个非常值得推广和使用的技术 For different well interpretation models, the pressure derivative curve is used as a training sample and BP network is used for training. The trained network can identify the well testing interpretation model according to the actual well test data in the field. This paper tests the BP network with simulated data, incomplete data, noisy data and a field well test data. The results show that ANN can correctly identify well testing interpretation model and also identify incomplete and noisy data. Artificial neural network technology effectively improves the current pattern recognition method widely used in well testing interpretation model, which is a technique worth popularizing and using