Orbit Determination - performance evaluation w/ Range, RRate and AzEl

1

Hello All,

I am writing an Orbit Determination code based on the great laser orbit determination tutorial: Jupyter Notebook Viewer

In the tutorial the Range performance is extracted from the following piece of code:

for estMeas in estimatedMeasurements:
    estMeas = EstimatedMeasurement.cast_(estMeas)
    observedValue = estMeas.getObservedValue()
    estimatedValue = estMeas.getEstimatedValue()
    pyDateTime = absolutedate_to_datetime(estMeas.date)
    observedRangeSeries[pyDateTime] = observedValue[0]
    estimatedRangeSeries[pyDateTime] = estimatedValue[0]

However, my code also includes RangeRate and AzEl besides Range, and I have tried extracting the respective values via, for example:

for estMeas in estimatedMeasurements:
    estMeas = EstimatedMeasurement.cast_(estMeas)
    observedValue = estMeas.getObservedValue()
    estimatedValue = estMeas.getEstimatedValue()
    pyDateTime = absolutedate_to_datetime(estMeas.date)
    observedRangeRateSeries[pyDateTime] = observedValue[1]
    estimatedRangeRateSeries[pyDateTime] = estimatedValue[1]

i.e., trying to vary the positional factors of observedValue and estimatedValue, but without success.

My question is, how can I extract the respective lists of RRate and EzEl from the estimated orbit in order to evaluate residuals, in a similar way that is done for Range in the tutorial?

Many thanks in advance.

/Leonardo.

2

Hi @lghizoni

Is there a way in Python to know if an object is an instance of a class type?
If think you can try using the Python method isinstance().

I ask this question because in the Java tutorials of Orekit we look at this condition (i.e. instance of) to extract the statistics for the different measurement types. Indeed, the estimator only returns a list of estimated measurements without differences on the type. Please find the following example used in the Orekit tutorials:

// Compute some statistics
for (final Map.Entry<ObservedMeasurement<?>, EstimatedMeasurement<?>> entry : estimator.getLastEstimations().entrySet()) {
     if (entry.getKey() instanceof Range) {
         final EstimatedMeasurement<Range> evaluation = (EstimatedMeasurement<Range>) entry.getValue();
         // Fill range statistics
     } else if (entry.getKey() instanceof RangeRate) {
         final EstimatedMeasurement<RangeRate> evaluation = (EstimatedMeasurement<RangeRate>) entry.getValue();
         // Fill range rate statistics
     } else if (entry.getKey() instanceof AngularAzEl) {
         final EstimatedMeasurement<AngularAzEl> evaluation = (EstimatedMeasurement<AngularAzEl>) entry.getValue();
         // Fill az-el statistics
     }
}

The isinstance() method could help you to translate this code in Python.

Best regards,
Bryan

1 Like
3

Hi @lghizoni,

Just a quick remark on:

Beware that RangeRate is a one dimensional measurement, so you should still use index 0 for this one (instead of 1 in your example).
AzEl is two dimensional, with index 0 you’ll get the azimuth and with index 1 the elevation (both in radians).

1 Like
4

Dear @bcazabonne,

Many thanks for the input! I will try to implement it in my code, and let you know the outcomes :slight_smile:

Best regards,
/Leonardo.