A hybrid approach based on differential evolution and tissue membrane systems for solving constrained manufacturing parameter optimization problems

This paper presents a hybrid approach based on appropriately combining Differential Evolution algorithms and Tissue P Systems (DETPS for short), used for solving a class of constrained manufacturing parameter optimization problems. DETPS uses a network membrane structure, evolution and communication rules like in a tissue P system to specify five widely used DE variants respectively put inside five cells of the tissue membrane system. Each DE variant independently evolves in a cell according to its own evolutionary mechanism and its parameters are dynamically adjusted in the process of evolution. DETPS applies the channels connecting the five cells of the tissue membrane system to implement communication in the process of evolution. Twenty-one benchmark problems taken from the specialized literature related to constrained manufacturing parameter optimization are used to test the DETPS performance. Experimental results show that DETPS is superior or competitive to twenty-two optimization algorithms recently reported in the literature.

Figure optionsDownload as PowerPoint slideHighlights
► A method hybridizing differential evolution and tissue P systems is proposed.
► First attempt to use a tissue P system to organize different types of algorithms.
► First investigation regarding the use of P system-based technique to optimize manufacturing parameters.
► 21 benchmark constrained manufacturing parameter optimization problems are involved.
► 22 methods are considered as benchmark algorithms.