Quantum-inspired Evolutionary Algorithm And Its Applications In Design Of Planar Wideband Matching Structure

Quantum-inspired evolutionary algorithm(QEA)combines the concept of quantum bit in the evolutionary algorithm,which can be deployed on the future quantum computer to achieve efficient optimization design.A multi-objective evolutionary algorithm for the design of microwave passive devices with discrete structure--multi-objective evolutionary algorithm based on decomposition combined with genetic operators(MOEA/D-GO)has typical qubit characteristics.This paper explores and implements the quantum algorithm architecture of MOEA/D-GO,and applies it to the design of planar wideband matching structure for verification.The main work of this paper includes:1.Introducing the concept and representation of quantum and quantum rotate gate in quantum-inspired evolutionary algorithm into MOEA/D-GO,and proposing quantum-inspired MOEA/D-GO(QMOEA/D-GO)algorithm for the design of single-bit discrete structure.Furthermore,for the multi-bit discrete structure,the multi-superposition state quantum bit is proposed,and the individual representation and individual update mode of QMOEA/D-GO after the introduction of multi-superposition state quantum bit is constructed.The multi-superposition states QMOEA/D-GO(MS-QMOEA/D-GO)algorithm is proposed.2.The efficiency of QMOEA/D-GO algorithm is analyzed by comparing with the multi-section impedance transformer optimized by HFSS.The validity of the QMOEA/D-GO algorithm is verified by the test of the actual matching structure.The proposed QMOEA/D-GO algorithm is applied to the design of planar multi-section impedance transformer and planar wideband compact matching structure,respectively.The proposed MS-QMOEA/D-GO algorithm is used to design high impedance transform ratio wideband matching structure.The simulated and measured results show that QMOEA/D-GO algorithm and MS-MOEA/D-GO algorithm can be used in the design of planar wideband matching structure.3.The proposed QMOEA/D-GO algorithm is used to design the output matching network structure of class-AB power amplifiers,and the proposed MS-QMOEA/D-GO algorithm is used to design the output matching network structure of base station power amplifier without lumped matching capacitance.The simulated and measured results show that QMOEA/D-GO algorithm and MS-MOEA/D-GO algorithm can be applied to the design of output matching network of base station power amplifier.In conclusion,the QMOEA/D-GO algorithm for the design of single-bit discrete structure design and the MS-QMOEA/D-GO algorithm for the design of multi-bit discrete structure can be well applied in the different design of planar wideband matching structures,and it is expected to achieve multi-objective optimization design of high-performance electromagnetic structures in the future quantum computer architecture.