Introduction to Kalman Filtering: Estimating True Values in Digital Signal Processing
Download as PPTX, PDF0 likes142 views
This slide, part of a presentation on "Introduction to Kalman Filtering: Estimating True Values in Digital Signal Processing" for a Digital Signal Processing course (EEE321), introduces Kalman filtering. It presents the method as an iterative mathematical process that estimates true values—such as position or velocity—from noisy or uncertain measurements. The slide underscores its ability to enhance accuracy despite errors, paving the way for a deeper exploration of its theoretical foundations and practical applications, including GPS and radar tracking.
Introduction to Kalman Filtering: Estimating True Values in Digital Signal Processing
- 1. Kalman Filtering Page 1/8 Presented by: Partho Prosad - 2022371 Sojib Ahmed Rifat – 2030936 Marjan Al Haque - 2020369 Alif Khan - 2022246 Shadid Ahmed - 1820990 Department of Electrical & Electronic Engineering INDEPENDENT UNIVERSITY, BANGLADESH EEE321 Group Presentation Introduction to Kalman Filtering: Estimating True Values in Digital Signal Processing Instructor: Prof. Dr. Md Kafiul Islam
- 2. What is Kalman Filtering? It is an iterative mathematical process that uses a set of equations and consecutive data input to quickly estimate the true value, position, velocity etc. of the object that is being measured, when the measured value contain unpredicted random error or uncertainty or variation. Kalman Filtering Page 2/8
- 3. Theoretical Background Kalman Filtering Page 3/8 Let's say we want to measure the temperature of a thermometer, but the thermometer is not reliable, therefore the data readings are doubtful to some extent. Let's assume that the horizontal axis represents time or the input of successive samples, and the vertical axis represents temperature. Actual temperature Estimate temperature using Kalman filter Initial estimate +- error or uncertainty
- 4. Presenters’ short name Short Title Page 4/24 Flowchart of a simple example (single measured value ) #1Calculate the Kalman gain #2 Calculate the current estimate #3 Calculate new error in estimate Error in estimate Error in data measurement Previous estimate Measured value (data input) Update estimate Kalman Filtering Page 4/8
- 5. The Kalman gain: A closer look 0<kg<1 est=es(t-1)+kg (mea – es(t-1)) Estimate are unstable Measurement are Accurate. Kg 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 Measurement are inaccurate. Estimate are stable (small error) Current estimate=est Previous estimate=es(t-1) Measurement=mea Kalman Filtering Page 5/8
- 6. Estimating GPS location Temperature of exhaust system Tracking airplane through radar Tracking satellite position Applications Kalman Filtering Page 6/8
- 7. State Estimator & State observer Before estimating unknown value, we estimate a known value State estimator is a model to estimate measured value State observer is a feedback system to eliminate error between measured value and estimated value In terms of real measurement now we can estimate the unknown value. Kalman Filtering Page 7/8
- 8. Any questions, comments or suggestions? THANK YOU Kalman Filtering Page 8/8