Two optimal approaches of two-stage fuzzy controller for traffic signals at isolated intersections are presented. Firstly

from the perspective of structural optimization

a two-stage combination fuzzy controller is designed

for the consideration of multi-dimensional traffic status variables leading to the inefficiency of traffic states at intersections under low traffic flow. The controller introduces 0-1 combination and determines the combination of traffic status variables of fuzzy controller's inputs. Secondly

aiming at the problems of fuzzy controller parameter empirical settings and functional disability of learning

a two-stage fuzzy logic traffic signal controller with online optimization is proposed

for the insufficiency of fuzzy controller parameter settings and functional disability of learning. This controller introduces the rolling horizon framework and optimizes the parameters of membership functions and fuzzy rules of fuzzy controllers by using hybrid genetic algorithm. Finally

the performance of the two proposed models is validated via online Paramics-based simulation platform

and extensive relative simulation tests have demonstrated the potential of the developed controllers for adaptive traffic signal control.