基于在线PINNN的电子设备剩余使用寿命预测
1Beijing Aerospace Automatic Control Institute, Beijing, China. 3120225045@bit.edu.cn.
Scientific reports
|December 16, 2025
概括
这项研究引入了在线物理信息神经网络 (PINN),用于航空航天电子的准确剩余使用寿命 (RUL) 预测. 这种新的方法增强了对不断变化的条件的稳定性,显著优于传统模型.
科学领域:
- 航空航天工程 航空航天工程
- 机器学习 机器学习
- 可靠性工程可靠性工程
背景情况:
- 准确的剩余使用寿命 (RUL) 预测对于航天器电子系统至关重要.
- 传统的RUL方法在一般化和数据要求方面存在局限性.
- 物理信息神经网络 (PINNs) 提高了准确性,但在不断变化的条件下遭受了灾难性的遗忘.
研究的目的:
- 开发一个新的在线PINN框架,与持续学习相结合.
- 提高RUL预测模型对新降解模式的动态适应性.
- 为了提高航空航天动力电子产品的RUL估计的稳定性和准确性.
主要方法:
- 提出了一个新的在线物理信息神经网络 (PINN) 框架.
- 集成的持续学习,使动态适应不断变化的运营条件.
- 使用NASA的智能门双极晶体管 (IGBT) 数据集验证了框架.
主要成果:
- 与传统的LSTM模型相比,RUL预测错误减少了27.2%.
- 显示在线更新时间不到每周期800毫秒.
- 显示了RUL估计的稳定性和准确性的显著改进.
结论:
- 提出的在线PINN与持续学习为RUL估计提供了更好的解决方案.
- 该方法有效地解决了由于运行条件的变化而导致的性能恶化.
- 这种方法在极端环境中为航空航天动力电子提供了更高的可靠性.
相关概念视频
Electrical Energy
1.6K
Using electric appliances for a longer period of time consumes more electrical energy and results in a higher electric bill. The energy produced by the transfer of electrons from one point to another is known as electrical energy. If power is delivered at a constant rate, the electrical energy can be defined as the product of power used by the device for a period of time. The energy unit on electric bills is the kilowatt-hour, where one kilowatt-hour is equivalent to 3.6 × 106 joules.
1.6K
Insulation Coordination
541
Insulation coordination is the process of matching electric equipment's insulation strength with protective device characteristics to protect the equipment against expected overvoltages. This selection is based on engineering judgment and cost. Equipment can generally withstand short-duration high transient overvoltages, but repeated tests with identical waveforms can yield inconsistent results. As a result, standard impulse voltage waveforms are used for testing, defined by specific times...
541


