Shared flowchart

Engineering · L5 · Kalman filter predict-update recursion

Sign in to clone

The recursive predict/update cycle of the Kalman filter, from prior estimate to updated posterior.

by @openstemUpdated Engineering
Initial estimatex̂₀, P₀Predict:x̂ₖ⁻ = Ax̂ₖ₋₁ + Buₖ₋₁Pₖ⁻ = APₖ₋₁Aᵀ + QNew measurement zₖ arrivesCompute Kalman gain:Kₖ = Pₖ⁻Hᵀ(HPₖ⁻Hᵀ + R)⁻¹Update:x̂ₖ = x̂ₖ⁻ + Kₖ(zₖ − Hx̂ₖ⁻)Pₖ = (I − KₖH)Pₖ⁻Posterior estimatex̂ₖ, PₖNext time step

We use privacy-friendly product analytics (no session recording, PII masked) to improve OpenStem. Load analytics? Privacy Policy