Generating fuzzy rules for target tracking using a steady-state genetic algorithm

Radar target tracking involves predicting the future trajectory of a target based on its past positions. This problem has been dealt with using trackers developed under various assumptions about statistical models of process and measurement noise and about target dynamics. Due to these assumptions, existing trackers are not very effective when executed in a stressful environment in which a target may maneuver, accelerate, or decelerate and its positions be inaccurately detected or missing completely from successive scans. To deal with target tracking in such an environment, recent efforts have developed fuzzy logic-based trackers. These have been shown to perform better as compared to traditional trackers. Unfortunately, however, their design may not be easier. For these trackers to perform effectively, a set of carefully chosen fuzzy rules are required. These rules are currently obtained from human experts through a time-consuming knowledge acquisition process of iterative interviewing, verifying, validating, and revalidating. To facilitate the knowledge acquisition process and ensure that the best possible set of rules be found, we propose to use an automatic rule generator that was developed based on the use of a genetic algorithm (GA). This genetic algorithm adopts a steady-state reproductive scheme and is referred to as the steadystate genetic algorithm (SSGA) in this paper. To generate fuzzy rules, we encode different rule sets in different chromosomes. Chromosome fitness is then determined according to a fitness function defined in terms of the number of track losses and the prediction accuracy when the set of rules it encodes is tested against training data. The rules encoded in the fittest chromosome at the end of the evolutionary process are taken to be the best possible set of fuzzy rules. For experimentation, several sets of real radar data, collected jointly by the defense departments of Canada and the United States, were used, and the performance of an SSGA-based fuzzy tracker was compared against a fuzzy tracker that uses expert-supplied rules. The experimental results showed that the SSGA-based tracker performed better in terms of its ability to minimize track losses and prediction error. The fuzzy rules obtained through the SSGA are also found to be more meaningful, in some cases, than those obtained from the experts. To further evaluate the effectiveness of the SSGA-based fuzzy tracker, we also compared it with some popular trackers including: i) a standard Kaiman tracker with maneuver model, ii) a two-stage Kaiman tracker, iii) an interacting multiple model (IMM) tracker, iv) an SSGA-based standard Kaiman tracker with maneuver model (to ensure that the optimal or near-optimal sets of parameters be used), v) an SSGA-based two-stage Kaiman tracker, and vi) an SSGAbased IMM tracker. Among these trackers, the SSGA-based fuzzy tracker was found to be the most effective in different tracking tasks involving a variety of targets moving in different environments.