test(tecs): explain why the airspeed step down is not first order

The step down is not first order because the SEB rate to pitch
conversion ignores the total energy rate: the pitch feedforward climbs
with the kinetic energy the throttle is already removing, and the pitch
integrator takes it out again. The loops interacting is not the cause;
with a consistent conversion the response is first order.

Assisted-by: Claude:claude-opus-5-5
Signed-off-by: RomanBapst <bapstroman@gmail.com>
This commit is contained in:
RomanBapst authored and Roman Bapst committed 2026-11-20 06:33:50 +03:00
1 parent 2f5f3f74fb
commit c7c8f29d56
1 file changed
+4 -3
+4 -3
View File
@@ -527,9 +527,10 @@ TEST_F(TECSClosedLoopTest, AirspeedStepDownDefaultTuning)
setAirspeedSetpoint(_params.equivalent_airspeed_trim - 1.f);
// Capture the rough shape of the airspeed step response. This one
// is not a nice first order response, because the throttle and pitch
// loop both interact, including the pitch integrator.
// Capture the rough shape of the airspeed step response. It is not first
// order: the SEB rate to pitch conversion ignores the total energy rate, so
// the pitch feedforward climbs with the kinetic energy the throttle is
// already removing, and the pitch integrator has to take it out again.
run(5.0f); EXPECT_NEAR(_state.V, _V_sp, 0.3f) << "at t=" << _t << " s";
run(5.0f); EXPECT_NEAR(_state.V, _V_sp, 0.2f) << "at t=" << _t << " s";
run(5.0f); EXPECT_NEAR(_state.V, _V_sp, 0.1f) << "at t=" << _t << " s";