分会场

高效清洁燃烧

摘要

In the mid-20th century, the concept of neural networks was introduced. Entering the 21st century, with the rapid advancement of computer technology, machine learning began to be utilized across various fields. In the realm of vehicle engines, numerous studies have applied it for predicting or optimizing engine performance. However, most of these studies focused primarily on the use of neural networks, overlooking how to enhance the efficiency of neural networks without compromising the accuracy of the results. In this paper, a single hidden layer GA-BP (Genetic Algorithm-Backpropagation) neural network was established for diesel spray data in a high-pressure constant volume vessel. The number of nodes in the hidden layer was determined using existing empirical formulas. Ultimately, the accuracy of the prediction results was analysed and compared from multiple perspectives to determine the best-performing empirical formula. This study offers recommendations for improving the use of neural networks, assisting engineers in working more efficiently.

关键词

Machine learning;Fuel spray;Penetration Prediction;hidden layer nodes

电子U盘全文仅限大会已缴费参会代表下载。

您还没有登录，请您先 点击这里登录