PSAS/ uncertainty/ to-do
Create 2 dimensional example
Create Plots of Filtering Last Flight Launch Data [Judy]

Notes from Bart: --- Take data from last flight. Run though Kalman Filters (smoothers). See if it matches. Use pressure altimeter, imu and gps data. Use Dave's simulator to generate data that we could run through the filter.

Notes from Tim: --- To get running quickly, Assume that the IMU data is in XYZ/Theta,Psi,Phi format. (Ignore complexities such as Q.) Model the sensor outputs as

Where (value) is the sensed parameter (e.g. accelleration, angular rate), (output) is the sensor output in some units, (G) is the sensor gain, (B0) in the measured initial bias, and (B) is the unknown bias modeled as zero mean independent Guassian noise with known standard deviation.

GPS data should be presented as Position(XYZ), Velocity(XYZ), and altimeter will be meters above ground level.

Possibly work in the Local tangent plane, unless your're sure some other coordinate system will work better for you.

Write Up Web Documentation [Judy]
Update RocketSim to generate bias data [Unassigned]

Suggestions from D. Allen:

If this bias is "real" then modify the RocketModel, if this is a measurement bias then modify the file trunk/c/tools/cppsim/RocketDes/src/AccelSensor.cpp, method DoAccelSensorEvent() for the x, y, z and q accelerometer bias.

Adding sensor bias for they gyro will be more difficult. The gyro sensor is not fully implemented. The code is trunk/c/tools/cppsim/RocketDes/src/GyroSensor.cpp, and the method is DoGyroSensorEvent(), but the gyro sensor data created by this method is static and not tied in with the current rocket model.

The code is located from: http://cvs.psas.pdx.edu/trunk/c/tools/cppsim/. The readme.txt helps explain how to build the code.

The version of the simulator currently in use is under the directory RocketDes.

Adding a bias is probably easy to do. You may want to make the change in trunk/c/tools/cppsim/RocketDes/src/RocketModel.cpp, in the method DoRocketModelEvent(). But it really depends on what kind of bias you are attempting to add.