Compute Jacobian Matrix wrt State Variables


I have a question regarding the computation of the State Jacobian.
Basically I want to compute the Jacobian in order to get the Matrix A of the linearized dynamic expressed as:
x_dot = A * x

To do so I have tried to use the getParametersJacobian method from the MatrixHarvester class. Unfortunatively it seems that it is useful to compute the Jacobian only for additional state parameters that I need to add directly in the force models.

There is a quick way to compute the linearized dynamic Jacobian that I need in orekit?

Thank you in advance,

Hi @DDega,

Welcome to the forum !

Try using: MatricesHarvester.getStateTransitionMatrix(SpacecraftState state) instead :wink:


Hi @MaximeJ, thank you for the quick reply.

I have another question about that, using the get STM method I will obtain the integral of the Jacobian instead of the Jacobian itself right?
Maybe I am missing something, given the non linear dynamics I would like to obtain the matrix A from equation (2).

I have noticed that in utils there is implemented and interface calle StateJacobian but I don’t know how to use it, or if it is a good option for what I am trying to do.

Thank you for your time.

Yes, you’re right I answered your previous post way too fast, thinking you picked the wrong function.
Sorry about that.

I’m not sure how you can get this matrix, I don’t think Orekit has a built-in function for it.

There’s a piece of code in PartialDerivativesEquation.combinedDerivatives that computes partial derivatives of Cartesian state time derivatives w.r.t to itself given the accelerations from force models in input (see here and block matrix in the comment containing dVel/dPos = 0, dVel/dVel = Id(3), dAcc/dPos and dAcc/dVel).
Is that what you’re looking for?


It should do the trick.

Now I have another question regardin the piece of code you have linked, I dont really get what is inside the matrices Adot Bdot etc. I mean it should be the state derivatives but I don’t know how they are mapped inside each matrix.