Fuel Economy Simulator

[GasSavers_maximilian]

I have extensive experience with genetic algorithms and want to create a program to evolve the most fuel efficient driving strategy for various terrain scenarios. A key part of this program will be a simulator to evaluate each potential strategy. I figure all I need are the rolling resistance and drag coefficients, terrain inclines, and engine efficiency and power output vs throttle position and I should be able to write such a simulator. Can anyone see any potential problems with this approach? I'll need to hunt down a reasonably representative set of graphs for the engine performance, so any suggestions on that front would be appreciated. Thanks.

[GasSavers_maximilian]

Oh, yeah, I guess I need an estimate for the extra frictional losses when applying power (vs when coasting in neutral) as well. I'm neglecting DFCO (at least for now) and assuming neutral coasting.

[GasSavers_maximilian]

Found this chart of si engine power, torque, and specific fuel consumption vs RPM in The Fundamentals of Vehicle Dynamics. Since I only want to simulate qualitative results from different hypermiling strategies, it doesn't matter if it matches my car. I guess next I need to get good estimates for transmission losses and then road load parameters. Then I should be able to start writing code.

[mikehallbackhoe]

the calculation for horsepower and torque, appears wrong on that chart. horsepower and torque, regardless of what engine is used, should cross paths, or intersect, at 5252 rpm.

[GasSavers_maximilian]

Thanks! I'll look for other charts to compare against.

[palemelanesian]

BSFC article: http://www.autospeed.com/cms/A_110216/article.html

[samandw]

Originally Posted by mikehallbackhoe

the calculation for horsepower and torque, appears wrong on that chart. horsepower and torque, regardless of what engine is used, should cross paths, or intersect, at 5252 rpm.

Unless the hp and tq are on a different scale Notice 120 hp is about the same place as 150 tq.

[samandw]

The really challenging part about this will be finding BSFC numbers in the form of a 3D plot. You need a plot of BSFC against a range of HP @ each RPM. Most plots only give WOT data . . . throttling losses can have a very major effect on BSFC.

[GasSavers_maximilian]

Upon closer examination I've found this sufficiently daunting that I'm probably ditching the project. I wanted to answer some pretty specific questions about correct driving strategy with it, but I guess I can just use my SG2 to do it empirically.

[soda_pop503]

Originally Posted by maximilian

Oh, yeah, I guess I need an estimate for the extra frictional losses when applying power (vs when coasting in neutral) as well. I'm neglecting DFCO (at least for now) and assuming neutral coasting.

neutral coasting is best, just short of shutting down the engine while coasting. but in my car I have to turn it back to the run position after a second, or the odometer doesn't turn. its very strange. and its not digital. DFCO is great and all, but it slows you down also. or at least wont let you gain much speed going downhill.