In the domain of intelligent manufacturing and industrial optimization,evolutionary multitasking algorithms typically rely on manually set key parameters(e.g.,offline/online learning rates for random mating probabilit...In the domain of intelligent manufacturing and industrial optimization,evolutionary multitasking algorithms typically rely on manually set key parameters(e.g.,offline/online learning rates for random mating probability)to guide evolutionary exchanges among populations,thereby enhancing algorithmic performance.However,trait segregation,a wellrecognized phenomenon in biological evolution,can naturally guide the exchange of genetic information without the need for manually predefined parameters.Inspired by this principle,this paper proposes a mutagenic multifactorial evolutionary algorithm based on trait segregation(M-MFEA),which thoroughly investigates how trait expression can naturally guide population evolution.First,the trait expression(dominant or recessive)of individuals,based on the trait segregation mechanism in a unified multitasking search space,is defined.Next,a mutagenic genetic information interaction strategy based on trait segregation is designed to enhance information transfer within and across tasks,enabling individuals to guide evolution according to their trait expressions spontaneously.Finally,an adaptive mutagenic gene inheritance mechanism is developed to drive continuous task convergence,and the complete framework and key components of M-MFEA are presented.Validation on benchmark suites and an industrial planar kinematic arm control problem shows M-MFEA has significant competitive advantages over state-of-the-art methods,establishing a novel paradigm for collaboratively solving multitasking optimization problems in complex industrial scenarios.展开更多
基金supported by the National Key Research and Development Program of China(Grant No.2023YFB3308002)the National Natural Science Foundation of China(Grant Nos.62473186,62573076)the Science and Technology Research Program of Chongqing Municipal Education Commission(Grant No.KJZD-K202400513)。
文摘In the domain of intelligent manufacturing and industrial optimization,evolutionary multitasking algorithms typically rely on manually set key parameters(e.g.,offline/online learning rates for random mating probability)to guide evolutionary exchanges among populations,thereby enhancing algorithmic performance.However,trait segregation,a wellrecognized phenomenon in biological evolution,can naturally guide the exchange of genetic information without the need for manually predefined parameters.Inspired by this principle,this paper proposes a mutagenic multifactorial evolutionary algorithm based on trait segregation(M-MFEA),which thoroughly investigates how trait expression can naturally guide population evolution.First,the trait expression(dominant or recessive)of individuals,based on the trait segregation mechanism in a unified multitasking search space,is defined.Next,a mutagenic genetic information interaction strategy based on trait segregation is designed to enhance information transfer within and across tasks,enabling individuals to guide evolution according to their trait expressions spontaneously.Finally,an adaptive mutagenic gene inheritance mechanism is developed to drive continuous task convergence,and the complete framework and key components of M-MFEA are presented.Validation on benchmark suites and an industrial planar kinematic arm control problem shows M-MFEA has significant competitive advantages over state-of-the-art methods,establishing a novel paradigm for collaboratively solving multitasking optimization problems in complex industrial scenarios.
