Use of TimeSpanDragForce

Orekit development
21

Hello again everyone

I’m trying to set up DragForce to emulate the behavior of TimeSpanDragForce. I created my own DragSensitive model where the drivers are now TimeSpanMap instances instead just numbers. This works fine when propagating, however, when I use the force model in my OD, the CDs estimated are wrong, already from the first iteration. Any idea of where the problem could come from?

This is my DragSensitive model (in Python):

class MyDragSensitive(PythonDragSensitive):
    def __init__(self, cd, area, dates):
        SCALE = FastMath.scalb(1.0, -3)
        
        dragCoef = float(cd)
        dragArea = float(area)

        drivers = ArrayList().of_(ParameterDriver)
        drivers.add(ParameterDriver(DragSensitive.GLOBAL_DRAG_FACTOR,
                                    1.0, SCALE,
                                    0.0, Double.POSITIVE_INFINITY))
        
        namesMap = TimeSpanMap('drag before')
        valueMap = TimeSpanMap(Double(dragCoef))
        for ii, date in enumerate(dates):
            name = 'Cd%02d'%ii
            namesMap.addValidAfter(name, date, False)
            valueMap.addValidAfter(Double(dragCoef), date, False)
        drivers.add(ParameterDriver('Cd',
                                    namesMap, valueMap,
                                    dragCoef, SCALE,
                                    Double.NEGATIVE_INFINITY, Double.POSITIVE_INFINITY))
        # drivers.add(ParameterDriver('drag coefficient',
        #                             dragCoef, SCALE,
        #                             Double.NEGATIVE_INFINITY, Double.POSITIVE_INFINITY))
        
        self.area = area
     

        self.drivers = drivers
        super().__init__()

    def dependsOnAttitudeRate(self):
        return False
    
    def dragAcceleration(self, state, rho, relVel, params):
        mass = state.getMass()
        if isinstance(relVel, Vector3D):
            cdA = params[0]*params[1]*self.area
            val = self.ore.Vector3D(relVel.getNorm()*rho*cdA/(mass*2), relVel)
        else:
            cdA = params[0].multiply(params[1].multiply(self.area))
            val = FieldVector3D(relVel.getNorm().multiply(rho.multiply(cdA)).divide(mass.multiply(2)), relVel)
        # print(val)
        return val

    def getDragParametersDrivers(self):
        return self.drivers

Edit: Never mind, found the issue… I forgot that since I now have to estimate all of the CDs, the first one when instantiating the TimeSpanMap will also be estimated. The way I was constructing it before, that CD had no GNSS data during the OD