Augmented proportional navigation — gravity compensation
80/20 core: Pure PN steers on the line-of-sight (LOS) rotation only. But gravity constantly bends the missile's flight, adding a spurious LOS rate that PN "chases" — wasting acceleration and causing miss. Augmented PN (APN) with gravity compensation simply adds a term that cancels the known gravitational acceleration so the guidance law only reacts to the real target maneuver.
WHY this note exists
- WHAT: Add a command term equal and opposite to the component of gravity acting perpendicular to the LOS (the part that curves the trajectory).
- WHY: So the effective closed-loop system behaves like a gravity-free PN engagement → zero-effort-miss driven only by target maneuver.
- HOW: Take standard PN, add the target-maneuver term (this is "augmentation"), then add a gravity-cancelling term.
Building the law from first principles
Step 1 — Recall pure Proportional Navigation
Why this form? If , the LOS direction is fixed in inertial space → this is the constant-bearing, decreasing-range collision condition. PN drives , guaranteeing intercept.
Step 2 — Where does gravity break it?
Write the missile's acceleration as commanded + environmental:
The LOS-rate dynamics (linearized, planar) are governed by the acceleration perpendicular to the LOS. Let denote the direction normal to the LOS. Then: where is target acceleration LOS and is gravity's projection LOS.
Step 3 — Augment for target maneuver
If the target accelerates with known/estimated , add the term that keeps zero-effort-miss at zero:
Why the ? Solve the linear ZEM (zero-effort-miss) equation. For a constant target acceleration over time-to-go , the miss integrates as . Feeding this forward requires the factor to exactly cancel it. (This is standard APN.)
Step 4 — Add gravity compensation
Now treat gravity as another known acceleration we must not let corrupt . We command an extra lateral term that cancels the gravity component perpendicular to LOS:
Why the minus sign and why (not full )?
- The full has a component along the LOS (changes closing speed, harmless to ) and a component perpendicular (, the guilty party). Only bends the LOS, so only needs cancelling.
- We subtract it because we want , i.e. the commanded normal accel must supply an extra so that after gravity adds back , the net perpendicular acceleration equals exactly what pure PN wants.
HOW to compute in practice: If the LOS elevation angle is (from horizontal) and gravity is , then When the LOS is horizontal (), (worst case — all of gravity bends the LOS). When shooting straight up/down (), (gravity is along LOS, harmless).

Worked examples
Recall Feynman: explain to a 12-year-old
Imagine you throw a dart at a moving toy car. The dart falls a little as it flies. If you keep aiming straight at where the car is right now, the falling dart always lands low. A smart dart-thrower aims a bit higher to cancel the fall, so the only thing left to worry about is the car moving. Gravity compensation is the missile aiming "a bit higher" all by itself — it knows exactly how much gravity will pull it down and adds the opposite push, so its brain can focus 100% on the real target.
Flashcards
What LOS quantity does pure PN try to drive to zero?
Why does uncompensated gravity hurt PN?
Write the full APN-with-gravity-compensation command.
Why only and not full ?
For LOS elevation in the vertical plane, what is ?
Why the factor on the target-maneuver term?
Why do we SUBTRACT ?
In a straight-up shot, how much gravity compensation is needed?
Connections
- Proportional Navigation (PN) — the base law this augments.
- Zero-Effort-Miss (ZEM) guidance — derivation source of the factor.
- Line-of-Sight rate estimation — how is measured (seeker gimbal / Kalman filter).
- Missile Autopilot & Acceleration limits — why saving g-capacity by not chasing gravity matters.
- Coordinate frames & projections — how is computed onboard.
- Ballistic trajectory & gravity turn — same droop physics, uncontrolled case.
Concept Map
Hinglish (regional understanding)
Intuition Hinglish mein samjho
Dekho, missile bhi ek thrown ball ki tarah gravity me neeche girta hai — usko "droop" bolte hai. Normal Proportional Navigation (PN) sirf ek cheez dekhta hai: line-of-sight (LOS) ghoom raha hai ya nahi. Agar LOS ka rotation rate zero hai to missile seedha target pe collision course pe hai. Problem ye hai ki gravity continuously missile ko neeche kheechti hai, jiske wajah se LOS thoda-thoda ghoomta rehta hai. PN samajhta hai "arre target bhaag raha hai!" aur bekaar me lateral acceleration burn karta hai — asal me to sirf gravity ka natak chal raha tha.
Gravity compensation ka funda simple hai: computer ko bata do ki gravity kitni hai, aur us ka opposite push add kar do. Formula banta hai . Yahan pehla term normal PN hai (LOS spin follow karo), doosra term APN augmentation hai (agar target maneuver kar raha hai to ko ke saath feed karo), aur teesra term hai — yahi gravity compensation hai. Minus sign isliye ki gravity jitna neeche kheechegi, hum utna upar push kar denge, so net effect cancel.
Important baat: sirf subtract karte hai, poora nahi. Kyunki gravity ka jo component LOS ke along hai wo sirf closing speed change karta hai, harmless hai. Jo component LOS ke perpendicular hai wahi LOS ko bend karta hai — bas usko maarna hai. Horizontal shot me () poora bend karta hai, worst case. Seedha upar ya neeche shoot karo to , koi compensation nahi chahiye.
Isse fayda kya? Missile ka precious acceleration budget bachta hai (autopilot ke g-limits hote hai), aur droop se hone wala steady miss khatam ho jaata hai. Short me: gravity compensation = "ghost chasing band karo, sirf real target pe focus karo."