锂离子电池的剩余寿命预测方法研究

发布时间：2018-03-27 18:45

本文选题：退化数据　切入点：锂离子电池　出处：《西安理工大学》2017年硕士论文

【摘要】：随着卫星技术的发展,卫星载体中的许多关键部件变得越来越复杂。在设备运行的环境中,由于复杂性和多种不确定因素的作用,其功能与性能都将不可避免的发生退化,进而造成设备的最终失效。假如,能够在部件开始出现性能退化的时候对其进行剩余使用寿命(Remaining Useful Life, RUL)估计,并在此基础上确定对设备的最佳维修时机,可以有效地避免可能会发生的重大故障,从而达到降低设备运行的风险的目的,这对于提高设备运行的安全性和可靠性具有重要意义。本课题是来自XXXX测控中心的委托项目,是该项目的子课题,以卫星的关键部件——锂离子电池为对象研究其剩余使用寿命预测方法。首先,本文在对锂离子电池(Lithium Ion Battery,LIB)工作原理和失效机理清楚认识的基础上,针对卫星锂离子电池在轨实际运行失效数据难以获得的问题,采用了美国航空航天实验室(National Aeronautics and Space Administration, NASA)对外公布的锂离子试验退化数据替代失效数据,并分析了试验数据的统计特性,从而采用相应的算法进行剩余使用寿命估计。其次,针对单一算法存在一些不可避免的缺陷,如,扩展卡尔曼滤波、粒子滤波强烈地依赖于锂离子电池的模型,而自回归方法只与样本数据有关。不管是依赖模型还是只依赖数据的统计特性,都会对估计的结果产生影响,因此提出了一种混合模型算法,该模型既依赖于数据的统计特性,又依赖于研究对象的模型,取其各自算法的优势,使混合模型算法的预测性能得到改善。最后,借助一些算法的性能评估指标,量化地分析了本文中使用的RUL预测算法的性能。通过量化指标,更进一步地说明了,提出的混合模型的预测的算法性能相对于单一算法的预测性能有所提高。锂离子电池剩余使用寿命预测问题对许多领域的应用都至关重要,准确实现RUL估计对视情维修和提高系统可靠性具有重要意义。本文中的预测方法都已用MATLAB仿真实现,验证了算法的有效性。
[Abstract]:With the development of satellite technology, many key components in satellite carrier become more and more complex. If it is possible to estimate the remaining service life of the components at the beginning of performance degradation, and on this basis determine the optimal repair time for the equipment, It can effectively avoid the major failures that may occur, thus reducing the risk of equipment operation, which is of great significance for improving the safety and reliability of the equipment operation. This project is a commissioned project from the XXXX Measurement and Control Center. Lithium Ion battery is a sub-project of the project. The residual service life prediction method of Lithium Ion battery is studied with the key component of satellite as the object. Firstly, based on a clear understanding of the working principle and failure mechanism of Lithium Ion Battery Lib, a key component of the satellite, Lithium Ion BatteryLib is studied in this paper. Aiming at the problem that it is difficult to obtain the failure data of the satellite lithium ion battery in orbit, the degradation data of the lithium ion test published by the National Aeronautics and Space Administration (NASAA) of the American Aeronautics and Astronautics Laboratory are used to replace the failure data. The statistical characteristics of the experimental data are analyzed, and the corresponding algorithm is used to estimate the remaining service life. Secondly, there are some unavoidable defects in the single algorithm, such as extended Kalman filter. Particle filter strongly depends on the model of lithium-ion battery, and autoregressive method is only related to sample data. Whether it depends on the model or only on the statistical characteristics of the data, it will have an effect on the estimated results. Therefore, a hybrid model algorithm is proposed, which depends on both the statistical properties of the data and the model of the research object. The prediction performance of the hybrid model algorithm is improved by taking advantage of their respective algorithms. The performance of the RUL prediction algorithm used in this paper is quantitatively analyzed with the help of some performance evaluation indexes of the algorithm. The predictive performance of the proposed hybrid model is better than that of the single algorithm. The residual life prediction of lithium ion batteries is of great importance in many applications. Accurate realization of RUL estimation is of great significance to maintenance of visual condition and improvement of system reliability. All the prediction methods in this paper have been realized by MATLAB simulation, which verifies the validity of the algorithm.
【学位授予单位】：西安理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TM912

【参考文献】