Validation of Worst-Case and Statistical Models for an Automotive EMC Expert System
This document has been relocated to http://scholarsmine.mst.edu/ele_comeng_facwork/1324
There were 22 downloads as of 27 Jun 2016.
Abstract
Previous papers have presented algorithms for an EMC expert system used to predict potential electromagnetic compatibility problems in a vehicle early in the design process. Here, the accuracy of inductive and capacitive coupling algorithms are verified through representative measurements of crosstalk within an automobile. Worst-case estimates used by the algorithms are compared to measured values and are compared to values estimated using statistical methods. The worst-case algorithms performed well up to 10-20 MHz, but overestimated measured results by several dB in some cases and up to 10-15 dB in others. An approximate statistical variation of the current expert system algorithms also worked well and can help avoid overestimation of problems; however, worst-case estimates better ensure that problems will not be missed, especially in the absence of complete system information.
