论文部分内容阅读
关于用内燃机油模拟实验数据预测其台架实验结果的经验知识,在内燃机油配方的开发过程中有着重要的作用。但是,由于内燃机油在模拟实验和台架试验中分别处在非常不同的条件下发生物理化学变化,影响因素复杂,因而很难用数学模型准确地描述这些结果间的关系。近年来,由于人工智能的发展,为知识的获取和利用提供了更为有力的工具。本文介绍用人工神经网络的方法,根据内燃机油模拟实验数据预测其台架评定结果的知识。
The experience of predicting the experimental results of bench test using internal combustion engine oil simulation experimental data plays an important role in the development of internal combustion engine oil recipes. However, since the internal combustion engine oil undergoes physicochemical changes under very different conditions in simulated experiments and bench tests, respectively, and the influencing factors are complex, it is very difficult to accurately describe the relationship between these results by mathematical models. In recent years, due to the development of artificial intelligence, it has provided a more powerful tool for the acquisition and utilization of knowledge. This paper introduces the method of using artificial neural network to predict the knowledge of bench evaluation results based on the simulation data of internal combustion engine oil.