Investigating the Performance of Some Tracking Filter Schema for the Advanced Shipboard Command and Control Technology (ASCACT) Project
Abstract
Tracking maneuvering targets is a complex problem which has attracted a great deal of effort over the past several years. It has been now well established that, in terms of tracking accuracy, the Interacting Multiple Model (IMM) algorithm where state estimates are mixed perform significantly better for maneuvering targets than the other types of filters (Adaptive Single Model, Input Estimation, Variable Dimension, etc.). However, the complexity and computation cost of the IMM algorithm can prohibit its use in some applications for which simpler algorithms can provide us with the necessary accuracy at a lower computation cost. This document presents the evaluation of the tracking accuracy of a multiple model track filter using three different constantvelocity models running in parallel {3CVPAR) and a maneuver detector. The output estimate is defined by selecting the model having its likelihood function lower than a Target Maneuver Threshold (TMTH). This approach is recommended for the MultiSensor Data Fusion (MSDF) problem to be tested in the Advanced Shipboard Command and Control Technology (ASCACT) testbed. The tracking performance of the 3CVPAR track filter is compared with: 1) an adaptive single motion model Kalman filter (ASMMKF) ; 2) an IMM algorithm using the same three CV models than the 3CVPAR filter; 3) an IMM filter using a CV model and a constant acceleration (CA) model producing a CVCA filter; 4) an IMM filter using a CV and two CA models (CA 1, CA2) differing only by the level of process noise producing a CV2CA filter. Calculations of the average Root Mean Square Error (RMSE) on 100 Monte Carlo runs permit to evaluate the tracking accuracy of the 3CVPAR track filter compared with simpler (ASMMKF) or more complex (IMMs) algorithms on a challenging multiple-sensor scenario.
Document Details
- Document Type
- Technical Report
- Publication Date
- Oct 01, 2000
- Accession Number
- ADA593299
Entities
People
- A. Jouan
- E. Bosse
- J. Roy