Modern Calibration: The Need for Automation
Worldwide engine manufacturers are under pressure to provide engines that meet tougher government-mandated emissions regulations, and increased customer expectations for durability, noise and performance, and low fuel consumption. To meet these demands, engine complexity is on the rise. Current engines are using variable valve timing, EGR (internal or external), multiple injections, turbocharger controls (variable geometry or wastegate) as well as advanced control strategies and sophisticated exhaust aftertreament systems to achieve these demands.
When earlier engines had only a few controlled parameters, the skilled calibrator could determine near optimal settings using a variety of manual methods and general-purpose software. Plots of the effects and the cross effects of parameters could be generated in Excel or other mainstream analysis or plotting tools. From these plots and the experience of the calibrators, good calibrations could be generated and refined.
However, now that engines have many more controlled parameters, it is becoming increasingly difficult to efficiently generate tests and analyze the results to produce an optimal calibration. Where, using traditional full factorial mapping, three variations of three parameters with potential interactions at 11 speed/load conditions would take ninety-nine (33*11=99) experiments, five variations of five variables at 11 conditions takes over thirty-thousand (55*11=34,375) experiments. Engineers now have at least this number of variables to deal with. Going beyond three-dimensional space (more than three variables) usually requires analytical tools. Clearly, new engine and engine controls design requires new methods of mapping, optimization and calibration.
[ back to top ]