Current forces.maneuvers package only proposes the ConstantThrustManeuver class as a numerical force model that can be integrated during a propagation and estimated with orbit determination.

It has been discussed in the following thread on the forum that it would be a good addition to have a more versatile implementation of numerical maneuvers' models that would allow for:

• Estimating the direction of the maneuver as well as its norm, with a 3D model that could be customized;
• Implementing a variable thrust maneuver model (see this other forum thread);
• Using a ΔV-based definition instead of a thrust-based definition to model, estimate or optimize quasi-impulse maneuvers;
• Using other triggers than dates to define start and stop of maneuvers;
• etc.

In that perspective a first implementation of a new maneuver package was proposed and is available on branch maneuver-package.

It is based on the organization proposal given in the first forum thread mentioned here.