Pursue (§12): seek where a constant-velocity target will be.
The lead time comes from the closed-form interception solution described on
this module's interceptTime helper — not from Reynolds' distance / maxSpeed heuristic, which lags a crossing target — and the predicted point is
targetPosition + targetVelocity · t. Pursue then seeks that point,
so everything true of seek (full-speed desired velocity, truncation, braking
when standing on the point) is true here.
The prediction assumes the target holds its velocity for t seconds; a
manoeuvring target simply gets a fresh prediction every step, which is the
standard treatment and converges because t shrinks as the range closes.
Pursue (§12): seek where a constant-velocity target will be.
The lead time comes from the closed-form interception solution described on this module's
interceptTimehelper — not from Reynolds'distance / maxSpeedheuristic, which lags a crossing target — and the predicted point istargetPosition + targetVelocity · t. Pursue then seeks that point, so everything true of seek (full-speed desired velocity, truncation, braking when standing on the point) is true here.The prediction assumes the target holds its velocity for
tseconds; a manoeuvring target simply gets a fresh prediction every step, which is the standard treatment and converges becausetshrinks as the range closes.