Tracking algorithm

On Saturday, we will be on the edge for a EME transmission at 5GHz operated by Guy. That will our first  tracking operated in a fully automatic mode.

We worked hard recently to develop our own ephemeris prediction models. Priority is predicting accurate positions for the Moon and the Sun. The outputs of our trajectories have been tested against the NASA/JPL Horizon model, the fit is perfect (+/-0.001°).

If we consider tracking a target, constraints on PB8 are:

  • ON/OFF commands for AZ and EL motors in both directions (clockwise/counterclockwise, up/down).
  • Speed on AZ can be changed, but manually only.
  • Speed on EL is constant, around 0.028°/s (it depends on EL and whether the movement is up or down).

Given these constraints, we consider the following general guidelines for a tracking algorithm:

  • AZ and EL trajectories  should not be reversed (exception for rallying a point)
  • As ON/OFF commands cannot be really part of a PID control loop, tracking will be done in an open loop.

Next figure illustrates our strategy for tracking a target:pb8-tracking-1First, we define an antenna centric crosshair of what pb8 is seeing. The width of the crosshair can be set separately in both directions. The center point of the crosshair represents the current (az,el) position of pb8.

A moving target can be observed within this crosshair. The position of the target and its velocity are estimated by means of the ephemeris prediction models.

Given the AZ/EL projections of the velocity vector we can anticipate the need to move pb8 or not on both axis. The main goal is to keep the target inside the crosshair. The strategy is to move pb8 so that the crosshair (meaning the position of pb8) is slightly ahead of the target position. Antenna is stopped to let the target moves inside. Process is repeated.