用神经网络研究柴油机性能

人类当前所面临的重大科学技术研究任务之一,就是要揭示人脑的工作机理和智能本质,开发智能应用技术。神经网络从脑的神经系统结构出发来研究脑功能,研究大量简单的神经元的集团信息处理能力及其动态行为,具有大规模并行,分布式存储和处理,自组织,自适应和自学习能力。特别适用于处理需要同时考虑许多因素和条件,不精确和模糊的信息处理问题。将神经网络技术应用于柴油机领域,可以得到一个或多个发动机参量对某个或某几个研究对象的影响效果。更重要的是,通过它可以实现对柴油机性能的预测,而且预测结果令人满意。该论文系统的介绍了什么是神经网络系统,神经网络的一般结构模型及怎样将神经网络系统应用于柴油机领域中。最后用一些实验数据对所建立的神经网络进行了验证与说明。

The Discussion of Diesel Performance using the technique of Neural Networks

One of the most important research tasks for science and technique is to discover the working mechanism of human brain and the essence of human intelligence, and develope application of intelligence. Neural Networks are one of intelligence controlles. They study the abilities to handle the information of lots of simple nerve elements and their dynamic behavior. They can organize, adapt and study themselves and can particularly solve the problems of ambiguities that need considerations of many factors and conditions. If the technique of Neural Networks is applied to motor, we can know the influences of one or more motorparameters on others. More importantly, the motor function can be predicted with Neural Networks, and the prediction is satisfactorily precise. This paper presents what are Neural Networks, what is the Common structure model of Neural Networks and explains how Neural Networks are applied in motor field.Finally Neural Networks are further tested and proved by some experimental data.