Predict and smooth noisy sensor data using a Recursive Kalman Filter. Professional service for engineering and robotics.
Represents how much we trust our mathematical model vs sensor.
Represents the error/variance in the raw measurements.
Real-time recursive processing
The filter projects the current state and uncertainty forward in time.
It calculates the weight given to the new measurement (K).
It blends the prediction with the actual sensor data.
Notice how the Kalman estimate (Blue line) eventually stabilizes near the true value (Dashed line) despite the heavy sensor noise (Rose line). This tool demonstrates the "Optimal Estimator" theory used in Apollo moon missions and modern self-driving cars.
ヒント
クイックアクセスのためにこのツールをブックマークしてください。すべてはお使いのブラウザ上でローカルに動作します。