Abstract
Public traffic has a great influence, especially with the background of COVID-19. Solving simulation-based optimization (SO) problem is efficient to study how to improve the performance of public traffic. Global optimization based on Kriging (KGO) is an efficient method for SO; to this end, this paper proposes a Kriging-based global optimization using multi-point infill sampling criterion. This method uses an infill sampling criterion which obtains multiple new design points to update the Kriging model through solving the constructed multi-objective optimization problem in each iteration. Then, the typical low-dimensional and high-dimensional nonlinear functions, and a SO based on 445 bus line in Beijing city, are employed to test the performance of our algorithm. Moreover, compared with the KGO based on the famous single-point expected improvement (EI) criterion and the particle swarm algorithm (PSO), our method can obtain better solutions in the same amount or less time. Therefore, the proposed algorithm expresses better optimization performance, and may be more suitable for solving the tricky and expensive simulation problems in real-world traffic problems.
Highlights
Simulation optimization methods can be divided into three categories [1]: simulation-based optimization, optimization-based simulation and the optimization of simulation
Considering the disadvantages of both EGO and EGO-MO, this paper proposes a Kriging-based global optimization using multi-point infill sampling criterion
The results explain that the multi-point infill sampling criterion can get better optimization results, which is an increase of 7.2% compared to the best result of particle swarm algorithm (PSO), 17.5%, compared to the worst result of PSO, and even 45%, compared to the suboptimal value
Summary
Simulation optimization methods can be divided into three categories [1]: simulation-based optimization, optimization-based simulation and the optimization of simulation. The expected improvement (EI) criterion proposed by Jones et al [5] is the most well-known single-point criterion, which selects the point corresponding to the maximum value of the EI function in each iteration to update the Kriging model This criterion has been widely applied in other real-world engineering problems [6,7,8,9,10,11]. According to Sobester et al [27], the EI criterion sometimes cannot balance exploration and exploitation, which is important to the efficiency of algorithm To this end, Feng et al [28] proposed a method called EGO-MO, which used the multi-objective optimization to generate exploitation–exploration trade-off points.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have