Error Performance Analysis of IMM-Kalman Filter for Maneuvering Target Tracking Application
Meutia Yunita, Joko Suryana, Ahmad Qois Izzuddin · 2020
Maneuvering target is one of the important issues in the tracking system algorithm development. Various techniques are used to improve algorithm performance in maneuvering target tracking. The most popular technique is the Interacting Multiple Model (IMM) algorithm. IMM algorithm uses a combination of several types of filter models in the tracking process. In this paper, the error performance of three types of IMM algorithms namely IMM - Kalman Filter (KF) constant velocity (CV) and constant acceleration (CA), IMM-KF constant velocity (CV) and constant turn (CT), and IMM - KF CV, CA, and CT will be compared. The three types of algorithm will also be compared with a single filter KF- CV to see an increase in the performance of the Kalman Filter algorithm on target maneuvers after being combined. The four types of algorithm will be tested in 3 types of generated target trajectories. Based on the simulation results, it is concluded that the IMM-KF that uses 2 types of filter models provide the best error performance compared to other algorithms. These four types of algorithms will also be tested on real measurement data, which is Automatic Dependent Surveillance - Broadcast (ADS-B) measurement data commonly used to support secondary surveillance radar (SSR) at the airports. From the results of the implementation (using ADS-B data), the same conclusions are obtained from the simulation results.