Ask us
All work

Takeoff and climb, predicted from weight, air and runway.

What
Nonlinear regression
For
An aviation engineering client
Built with
MATLAB
Status
Delivered

Set the conditions, watch the climb

Ground distance
Height reached
Fuel used
Time

Scenario from the dataset. 142 scenarios, the data the model was fitted on.

Graph: fuel-burn rate over time for five takeoff weights
01 / 03
01 / 03

The problem

Fuel burn during takeoff depends on weight, air density, wind and runway elevation together. A lookup table won't do.

What it was

A fuel-burn model that holds within 1.5% across 142 cases.

How it was done

  1. Fitted a two-term tanh model of fuel-burn rate over time.
  2. Made each coefficient a function of initial mass, air density and wind.
  3. Validated against every scenario in the dataset.

The result

RMSE under 1.5% and maximum error under 10% across 142 scenarios.

Want something like it?