Yeh page har squiggle ko build karti hai jo topic use karta hai, order mein, bilkul shuru se. Har ek ko milega: woh kya kehta hai, woh kaisa dikhta hai, aur topic ko uski zaroorat kyun hai. Hum target equation se milenge
q˙=21Ξ(q)ω
sirf bilkul end mein, jab iske chaar symbols — q˙, 21, Ξ(q), ω — mein se har ek earn ho chuka ho. Abhi ke liye, ise ek locked door ki tarah samjho; hum ek-ek karke keys kaat rahe hain.
Ek satellite ko space mein floating imagine karo. Woh kahin move nahin ho raha, lekin woh turn kar sakta hai. "Abhi woh kitna turned hai" ka poora description uski attitude hai.
Figure 1 dikhata hai kyun humein do sets of axes ki zaroorat hai. Body ke apne teen axes (uska body frame, orange) fixed lab axes (the inertial frame, gray) se rotate hue hain. Picture jo insight deti hai: attitude koi cheez nahin jo body ke paas hai, yeh woh rotation hai jo gray ko orange pe carry karta hai. Is topic mein sab kuch us ek rotation ko track karta hai jab woh time ke saath change hota hai.
n^ (ek "n" hat ke saath) = woh axis jiske around turn karte ho. Hat ^ ka matlab hamesha hai "is vector ki length 1 hai" — yeh sirf direction carry karta hai, size nahin.
Yeh axis–angle idea quaternions ka bridge hai aur seedha Rodrigues rotation formula se link karta hai, jo ek axis aur angle ko ek actual vector rotation mein convert karta hai.
Topic ko yeh kyun chahiye: ek quaternion ki length hamesha exactly 1 honi chahiye. Parent ka poora final section ("why the norm stays 1") ∥q∥ ko check karta hai ki yeh kabhi drift na kare. Agar tum ∥q∥ nahin padh sakte, toh woh poora safety argument invisible hai.
Ye dono symbols Hamilton product mein throughout aate hain. Ye 3D vectors ko "multiply" karne ke sirf do tarike hain, aur ye alag sawaalon ka jawab dete hain.
Figure 2 split ko visible karta hai: blue aur orange arrows inputs a,b hain; green marker (cross product) seedha us plane se bahar point karta hai jo ye span karte hain; shaded parallelogram ka area equal hai∥a×b∥ ke; aur boxed number dot product a⋅b hai. Ek geometric lesson: dot = ek number plane mein rehta hai, cross = ek vector plane se bahar jaata hai.
Figure 3 dono halves ko side by side rakhta hai: left mein, chaar numbered slots (q0 blue, q1q2q3 orange); right mein, woh slots geometrically kya mean karte hain — ek blue angle θ aur ek orange axis arrow. Picture jo takeaway sikhati hai: chaar numbers ka ek column = ek angle ek axis se glued.
Dhyaan do ki quaternion ke andar angle hamesha half hota hai: cos2θ, sin2θ.
Topic ko yeh ABHI kyun chahiye: jab tum baad mein cos2θ ko time ke saath differentiate karte ho, chain rule 21 bahar kheench leta hai. Woh single 21 target equation ka poora mystery hai. Yeh half-angle mein chhupi thi.
Picture karo ki q ek point hai jo ek curve par slide kar raha hai; q˙ us point ka velocity arrow hai — woh kis direction mein ja raha hai aur kitni tezi se.
Spacecraft par ek gyroscope exactly yahi measure karta hai. Kyunki gyros body se bolted hain, ye ω ko body frame mein report karte hain — ek distinction jo baad mein multiplication order decide karta hai. Angular velocity and the body frame mein aur gehraai se samjho.
Ek hi physical arrow ko alag numbers milte hain depending on which frame se tum use padh rahe ho. Yeh poora subject Euler angles and gimbal lock aur Rotation matrices and SO(3) ka hai.
Topic ko yeh kyun chahiye: derivation tiny rotation δq ko "identity plus a small correction" likhti hai. Woh correction, cleanly factored, hi target equation ka right-hand side hai.
Ab har symbol defined hai, toh locked door khulti hai:
q˙=21Ξ(q)ω.
Left to right padho: q˙ (attitude kaise move karti hai, §6) equals 21 (half-angle se paida, §5) times Ξ(q) (Hamilton product q⊗(0,⋅) ek grid ke roop mein, §9–10) acting on ω (gyros se body-frame spin, §7–8).