In this paper, a multi-agent type-2 FLC method optimized by DE
for a traffic signal control application has been proposed. Type-2
FLC preserves the fuzziness and provides better approximation
than type-1 FLC as its three-dimensional membership functions.
This makes type-2 FLC more appropriate for traffic signal control.
Then DE was used to fine-tune the parameters of type-2 fuzzy system
and the expert rule base. DE was much more simple and
straightforward to implement compared with other evolutionary
algorithms and had low space complexity, which made DE more
suitable for the optimization of type-2 FLC. As the computational
complexity would increase exponentially if MF parameters and
the rule base are evolved simultaneously, the expert rule bases
and MF parameters were optimized by turns in order to avoid
the computational complexity in this paper. Aiming at the complexity
of traffic network, multi-agent systems were employed to
communicate the information among the adjacent intersections,
thus reducing the computational complexity. This architecture
improved the efficient application of expert system