摘要
随着近些年来锂离子动力电池的广泛应用,作为生产锂离子电池关键设备的检测系统也成为新的研究热点。但由于系统中大量非线性元件的使用,使得设定基准、采样信号和实际的测量值之间存在较大误差。为了减小干扰、提高数据传输准确性和控制精度,必须对传送的数据进行处理。人工神经网络以其任意非线性函数的任意逼近能力和自学习能力,在控制领域内得到了广泛的应用。用人工神经网络对数据进行处理,修正系统误差。结果表明,经神经网络处理后的采样数据检测精度大幅度提高,为提高电池生产质量提供了可靠保证。
In recent years, lithium-ion electric vehicle battery has been more and more widely used, as a key equipment of Li-ion battery production, battery testing system has become a new research hotspot. Because the use of many nonlinear elements in this system, the setting reference and sampling data were distorted during the data transmission. Aiming at reducing the interference and enhancing controlling accuracy, the data should to be processed. BP neural network with excellent nonlinear approximation and self-learning ability is widely applied to various fields. In order to process the transmitting data and correcting the systematic errors, A 3 --layer BP neural network was adopted in this paper.The testing results show that the network is easy to realize, the forecast data of which is high in accuracy, and thus is better in reducing the distortion of the sampling data.
出处
《电子工业专用设备》
2009年第9期14-18,44,共6页
Equipment for Electronic Products Manufacturing
基金
国家发改委重大高科技产业化项目
项目编号:发改办高技[2007]2456号
