test_planetNav

test_planetNav.planetNavTestFunction(show_plots)[source]

Test method

test_planetNav.test_gauss_markov_properties()[source]

Test the statistical properties of the Gauss-Markov noise model in planetNav. Tests: 1. Error bounds are not violated too often (<0.3% of samples) 2. Error bounds are actually being used (at least one >80% excursion)

test_planetNav.test_planetNav(show_plots)[source]

Validation Test Description

This unit test is designed to functionally test the simulation model outputs as well as get complete code path coverage. The test design is broken up into three main parts:

  1. Error Bound Enforcement: The simulation is run for 2.4 hours and the error bounds for all of the signals are tested. This test length is long enough to see both the walk in the signal and the noise, all the while not being so long as to slow down the test. The test ensures that the bounds are crossed no more than 30% of the time.

  2. Error Bound Usage: The error signals are checked for all of the model parameters over the course of the simulation to ensure that the error gets to at least 80% of its maximum error bound at least once, ensuring that noise is indeed properly introduced.

  3. Corner Case Check: The simulation is intentionally given bad inputs to ensure that it alerts the user and does not crash.

Test Parameters

These tests are considered to pass if during the whole simulation time of 144 minutes, all the variables need to stay within an allowable statistical error. This means that they must stay within their bounds 30% of the time.

At the same time, we want each of the errors to get to 80% of their respective error bounds at least once during the run.

The test used for the planetNav module tests the statistics of the Gauss Markov process, making sure that we are getting the variability we want. In order to do so, no specific scenario is necessary. Therefore the position of the planet is set generically. The planet position is [10000, 0, 0]^T