Ensemble of Two New Mutation Strategies for History-Based Parameter Adaptation Differential Evolution

History-based parameter adaptation differential evolution (DE) algorithms, including adaptive differential evolution with optional external archive (JADE) and successful history-based parameter adaptation differential evolution with linear population size reduction (L-SHADE) variants, have exhibited excellent performance in solving many optimization problems. But they are still exposing their weakness in solving some complex problems. To enhance the performance of the history-based parameter adaptation DE algorithm, this paper presents a DE with an ensemble scheme of two new mutation strategies (ESTMDE). In the proposed ESTMDE, two new mutation strategies, DE/current-to-Aqmean/1 and DE/current-to-pqbest, are used to balance the exploration ability and the exploitation ability. The proposed ESTMDE is tested using CEC2017 benchmark functions. Experimental results indicate that the proposed ESTMDE has better performance than the compared L-SHADE variants in terms of solution quality.