The rate of mutation has deep effects on the performance of
genetic algorithm (GA). Current mechanisms to control mutation rate (MR)
utilize the mathematical functions which usually are monotonic. These mechanisms are too rigid and inflexible. These methods change the MR without
enough attending the position of GA. For instance these methods don’t attend whether GA is in trap or not. This research proposes a novel mechanism
which controls MR by an algorithm which uses a concept of defined local trap.
This algorithm probes whether GA is in the local trap or is not. In case of
local trap, it changes the MR. This methodology is named MRCA (Mutation
Rate Control Algorithm). To evaluate performance of MRCA, it is applied to
multimodal continuous optimization functions and also a type of combinatorial optimization problem. The results show that MRCA outperforms other
state-of-the-art strategy in term of accuracy and speed.