A Novel Cultural Algorithm and Its Application to the Constrained Optimization in Ammonia Synthesis

A novel cultural differential evolution algorithm with multiple populations (MCDE) is proposed. The single individual in each population is affected by the situational and normative knowledge from belief space simultaneously. The populations communicate with each other following a rule of knowledge exchange, which helps to enhance the search rate of evolution. The concept of culture fusion is introduced to develop an adaptive mechanism of preserving the population diversity. The mechanism ensures that populations are diverse along the whole evolution and excellent candidate solutions are not rejected. The performance of MCDE algorithm is validated by typical constrained optimization problems. Finally, MCDE is applied to maximizing the net value of ammonia in an ammonia synthesis loop. The results indicate that the proposed algorithm has the potential to be used in other problems.