ImuFactor + improve estimate using an approximate aircraft model

[Suresh Kannan]

Greetings,

I am utilizing the CombinedIMUFactor to estimate the state of an aircraft. What is the best factor to use to improve the estimate when an approximate model of the aircraft is known?

I looked at 

1) the between-factor between adjacent estimates.

2)  also looked at the pendulum example in gtsam_unstable.

3) gtdynamics

The IMU preintegration (Forster paper) is formulated for an arbitrary angular velocity. However, if we have an approximate expression for both the angular acceleration and linear acceleration, what is the best way to incorporate them?

Sincerely,