The HWFET cycle, which includes the least transient behaviors, exhibits smaller differences in all data categories while still following the trend of matching the EDIL data the least closely in negative electric energy.
The voltage of the EDIL tests was constantly changing in response to the current drawn by the electric drive.
Data about the cycle as a whole cannot fully describe the differences in the results from the offline simulation and the EDIL testing.
For the majority of each test cycle, the behavior of the offline simulation and the EDIL tests are nearly indistinguishable from one another even under transient conditions.
However, there are certain circumstances that cause the simulation to differ meaningfully from the EDIL tests.
During coasting events, slight differences in vehicle speed between simulation and EDIL cause driver model pedal commands to differ by a percentage point and this can excite a non-linear behavior of the strategies: the simulation stays in coast regeneration mode whereas the EDIL experiment phases it out resulting in a motor torque reduction.
This is where EDIL testing can capture all electric drive characteristics because the electric drive is physically present.
The modelled (dark blue line) and EDIL (dotted red line) torque commands match very closely but the torque feedbacks do not: the EDIL experimental torque feedback (dotted pink line) temporarily drops before recovering, whereas the model torque feedback (line blue line) follows the command with a slight delay.
During early stages of accelerations, torque commands and feedbacks for both simulation and EDIL testing match very well but currents do not.
During regeneration events examplified in figure 17, there is often a mismatch between torque command and feedback (red dotted lines) during EDIL testing because the electric drive delivers more braking torque that it is commanded, whereas the simulation assumes a perfect open loop control of torque so it delivers exactly the torque that was commanded (blue lines).