Indicator-Based Multi-objective Bacterial Foraging Algorithm with Adaptive Searching Mechanism.

Lianbo Ma,Xu Li,Tianhan Gao,Qiang He,Guangming Yang,Ying Liu
DOI: https://doi.org/10.1007/978-981-10-3614-9_33
2016-01-01
Abstract:Derived from the social foraging behavior of E. coli bacteria and the general adaptive concentration searching strategy, this paper proposes and develops a novel indicator-based multi-objective bacterial colony foraging algorithm (I-MOBCA) for complex multi-objective or many-objective optimization problems. The main idea of I-MOBCA is to develop an adaptive and cooperative model by combining bacterial foraging, adaptive searching, cell-to-cell communication and preference indicator-based measure strategies. In this algorithm, each bacteria can adopt its run-length unit to appropriately balance exploitation and exploration states, and the quality of position or solution is calculated on the basis of the binary quality indicator to determine the Pareto dominance relation. Our algorithm uses Pareto concept and preference indicator-based measure to determine the non-dominated solutions in each generation, which can essentially reduce the computation complexity. With several mathematical benchmark functions, I-MOBCA is proved to have significantly better performance over compared algorithms for solving some complex multi-objective optimization problems.
What problem does this paper attempt to address?