Research On Electromagnetic Optimization Based On Particle Swarm Optimization And HFSS

Optimization is often met in engineering technology, scientific research, economic management, and many other fields. When optimizing some complicated electromagnetic models, the calculations always refer to lots of complicated mathematic or physics questions, including irregular structures, nonuniform composite materials, nonlinear dynamical systems, active boundaries and constraint conditions. It is still a problem to conduct large-scale and high accuracy calculations with complicated nonlinear mathematic or physics equations. HFSS (High Frequency Structure Simulator) is an industrystandard for three-dimensional electromagnetic field design and analysis. It can calculate the S parameter and get the full-wave analysis of electromagnetic field of three-dimensional passive structures with any shape. However, HFSS can't be used for optimaization because it can obtain optimal values only by iteration, which is very time-consuming and large workload.PSO (Particle Swarm Optimization) is a stochastic global optimal method based on swarm intelligence. PSO algorithm has some advantages, such as simple principle, few parameters, fast convergence, few requirements to the solved problem, and high universality. Therefore, it is often used to solve the optimal issues that are complex, large-scale, and non-linear. By researching the algorithm convergence and limitation and analyzing the interface of HFSS, this thesis worked at how to improve the PSO algorithm and combined it with HFSS to be used in the electromagnetic field. The main contents are as following:(1) Elaborated the PSO algorithm progressively, analyzed the deficiencies of algorithm and introduced some familiar improvement method.(2) Elaborated the advantages and disadvantages of Quantum-behaved PSO algorithm and proposed an improved algorithm called CQPSO-DE by introducing the idea of DE (Differential Evolution) operator and chaos disturbance. Then some benchmark functions were used to test the performance of the algorithm. Computing results demonstrated the effectiveness.(3) Researched the VBscript language used to control HFSS, combined HFSS and PSO algorithm by MATLAB and VBscript to design a new optimization method making most use of PSO's optimization skills and HFSS's analysis skills to optimize electromagnetic structures.(4) Optimized a wideband microstrip antenna and a compact multiband folded loop chip antenna for small-size mobile phone, designed a double-layer EBG (Electromagnetic Band Gap) structure filter successfully based on PSO algorithm and HFSS. This proved the effectiveness and value of the optimization method.