search

ARCLab-MIT / kspdg

ARCLab-MIT participation in the KSPDG challenge
MIT License
4 stars 0 forks source link

Are the pe1_base.py constants useful for the LLM? #6

Closed vrodriguezf closed 10 months ago

vrodriguezf commented 11 months ago

I'm referring to this:

 PARAMS.EVADER.CONTROL_RANGE = 2200  # [m]

    # Specific impulse assumes all RCS thrusters are identical RV-105
    # parts operating in vacuum
    PARAMS.PURSUER.RCS.VACUUM_SPECIFIC_IMPULSE = 240 # [s]
    PARAMS.PURSUER.RCS.VACUUM_MAX_THRUST_PER_NOZZLE = 1000 # [N]

    # Assumed number of thrusters creating thrust in each direction
    PARAMS.PURSUER.RCS.N_THRUSTERS_FORWARD = 8
    PARAMS.PURSUER.RCS.N_THRUSTERS_REVERSE = 8
    PARAMS.PURSUER.RCS.N_THRUSTERS_RIGHT = 4
    PARAMS.PURSUER.RCS.N_THRUSTERS_LEFT = 4
    PARAMS.PURSUER.RCS.N_THRUSTERS_UP = 4
    PARAMS.PURSUER.RCS.N_THRUSTERS_DOWN = 4

    # computed max thrust in each direction [N]
    PARAMS.PURSUER.RCS.VACUUM_MAX_THRUST_FORWARD = \
        PARAMS.PURSUER.RCS.N_THRUSTERS_FORWARD * PARAMS.PURSUER.RCS.VACUUM_MAX_THRUST_PER_NOZZLE
    PARAMS.PURSUER.RCS.VACUUM_MAX_THRUST_REVERSE = \
        PARAMS.PURSUER.RCS.N_THRUSTERS_REVERSE * PARAMS.PURSUER.RCS.VACUUM_MAX_THRUST_PER_NOZZLE
    PARAMS.PURSUER.RCS.VACUUM_MAX_THRUST_RIGHT = \
        PARAMS.PURSUER.RCS.N_THRUSTERS_RIGHT * PARAMS.PURSUER.RCS.VACUUM_MAX_THRUST_PER_NOZZLE
    PARAMS.PURSUER.RCS.VACUUM_MAX_THRUST_LEFT = \
        PARAMS.PURSUER.RCS.N_THRUSTERS_LEFT * PARAMS.PURSUER.RCS.VACUUM_MAX_THRUST_PER_NOZZLE
    PARAMS.PURSUER.RCS.VACUUM_MAX_THRUST_UP = \
        PARAMS.PURSUER.RCS.N_THRUSTERS_UP * PARAMS.PURSUER.RCS.VACUUM_MAX_THRUST_PER_NOZZLE
    PARAMS.PURSUER.RCS.VACUUM_MAX_THRUST_DOWN = \
        PARAMS.PURSUER.RCS.N_THRUSTERS_DOWN * PARAMS.PURSUER.RCS.VACUUM_MAX_THRUST_PER_NOZZLE

    # computed maximum fuel consumption rate in each direction [kg/s]
    PARAMS.PURSUER.RCS.VACUUM_MAX_FUEL_CONSUMPTION_FORWARD = \
        PARAMS.PURSUER.RCS.VACUUM_MAX_THRUST_FORWARD / (C.G0 * PARAMS.PURSUER.RCS.VACUUM_SPECIFIC_IMPULSE)
    PARAMS.PURSUER.RCS.VACUUM_MAX_FUEL_CONSUMPTION_REVERSE = \
        PARAMS.PURSUER.RCS.VACUUM_MAX_THRUST_REVERSE / (C.G0 * PARAMS.PURSUER.RCS.VACUUM_SPECIFIC_IMPULSE)
    PARAMS.PURSUER.RCS.VACUUM_MAX_FUEL_CONSUMPTION_RIGHT = \
        PARAMS.PURSUER.RCS.VACUUM_MAX_THRUST_RIGHT / (C.G0 * PARAMS.PURSUER.RCS.VACUUM_SPECIFIC_IMPULSE)
    PARAMS.PURSUER.RCS.VACUUM_MAX_FUEL_CONSUMPTION_LEFT = \
        PARAMS.PURSUER.RCS.VACUUM_MAX_THRUST_LEFT / (C.G0 * PARAMS.PURSUER.RCS.VACUUM_SPECIFIC_IMPULSE)
    PARAMS.PURSUER.RCS.VACUUM_MAX_FUEL_CONSUMPTION_UP = \
        PARAMS.PURSUER.RCS.VACUUM_MAX_THRUST_UP / (C.G0 * PARAMS.PURSUER.RCS.VACUUM_SPECIFIC_IMPULSE)
    PARAMS.PURSUER.RCS.VACUUM_MAX_FUEL_CONSUMPTION_DOWN = \
        PARAMS.PURSUER.RCS.VACUUM_MAX_THRUST_DOWN / (C.G0 * PARAMS.PURSUER.RCS.VACUUM_SPECIFIC_IMPULSE)

Would a manual expert solution require this information? Or is it secondary?

escharf320 commented 10 months ago

I spoke to one person about this. They are more knowledgeable than I am in this realm. It sounds like for a human user they might not be as important. However, If we already have the data would there be a reason to exclude it in the models we train?

vrodriguezf commented 10 months ago

Yeah I agree they are meaningful too. Peng also told me about it.