GNSS antenna placement for autonomous vehicles supported by Bayesian optimization. (15th March 2023)

• Record Type: Journal Article
• Authors: Cicconet, Franciele; Silva, Rui; Mendes, Paulo Mateus; Pereira, Tiago; Kaulmann, Stefan
• Is Part Of: Expert systems with applications. Volume 214(2023)
• Publication Date: 2023-03-15
• Subjects: Antenna placement -- Autonomous vehicles -- Optimization; Expert systems (Computer science) -- Periodicals; Systèmes experts (Informatique) -- Périodiques; Electronic journals; 006.33
• DOI: 10.1016/j.eswa.2022.119158
• Languages: English
• ISSNs: 0957-4174

Abstract

Abstract: Antenna positioning is critical to the accuracy of a car's location systems. However, specifying the best placement for an antenna in an autonomous vehicle is challenging due to the signal's non-linear dependency on position and the surrounding material. In this article, we introduce Bayesian Optimization (BO) as a solution to the problem of maximizing gain and minimizing axial ratio in GNSS antenna placement. We also demonstrate in simulations how it converges to the global optima (best antenna position) in 96% less objective function calls when compared to more traditional optimization methods — namely particle swarm optimization, genetic algorithm and differential evolution. Finally, we detail the use of BO in conjunction with CST Microwave Studio on a real-world scenario. Highlights: Bayesian Optimization is proposed to specify the best placement for a GNSS antenna. BO is well suited for global optimization of unknown objective functions. The method is compared to other well-known optimization algorithms. Tests in a real-world case validated the effectiveness of the proposal method.