论文部分内容阅读
经常性的现场测试是掌握电子系统实时指标和电磁兼容状况的基本手段。针对电磁兼容测试数据较多但应用不充足的现状,提出了电磁兼容模型综合的概念,结合电磁兼容原理,实现了运用数理统计方法和人工神经网络技术对电磁测量结果进行数据挖掘,对统计分析方法进行了实用性设计,研究了神经网络训练方法的实用性及学习率确定方法。电磁兼容模型综合可建立或修正模型,归纳综合设备电磁参数及系统和平台电磁兼容性能变化规律,实现电磁兼容信息综合与状态预测,为增强系统电磁兼容针对性测试、主动管理和预知性维护与保障提供技术手段。
Regular field test is to grasp the real-time indicators of electronic systems and electromagnetic compatibility status of the basic means. Aiming at the current situation that there is a lot of electromagnetic compatibility test data but not enough application, the concept of electromagnetic compatibility model synthesis is proposed. Combined with electromagnetic compatibility principle, the data mining of electromagnetic measurement results by using mathematical statistics and artificial neural network technology is realized. Method of practical design, study the practicality of neural network training methods and learning rate determination method. Electromagnetic compatibility model can be established to build or modify the model to summarize the electromagnetic parameters of equipment and system integration platform and electromagnetic compatibility changes in the law to achieve the electromagnetic compatibility of information and status prediction, in order to enhance the system of electromagnetic compatibility-oriented testing, active management and predictive maintenance and Guarantee to provide technical means.