中文摘要：

人工神经网络技术具有自动获取知识、联想记忆、自适应性、良好的容错性和推广的能力，处理内燃机特性曲线具有效果好、程序推广性强等特点，在内燃机台架性能试验自动化方面有着非常重要的作用。采用改进的神经网络BP算法，对内燃机万有特性曲线进行了训练和模拟，取得了良好的效果。

英文摘要：

Artificial neural networks have the abilities of obtaining knowledge automatically, associate anamnesis, self-accommodating, accommodating mistakes well and favorable popularizing. The technology of neural networks has the characteristics of precision and popularized easily in painting the characteristic curve of internal combustion engines, and has important effect in the aspect of bed-test automation of internal combustion engines. The modified arithmetic of BP neural networks is used for the training and painting of the characteristic curve of internal combustion engines this paper. The results showed that the method of painting characteristic curve of internal combustion engines by the technology of neural networks is feasible.