本篇論文中，我們提出了一種改良式的粒子群演算法，名為改良式模糊粒子群演算法AWFPSO (Adaptive Weighted Fuzzy Particle Swarm Optimization)，將使用模糊規則調適加速因子(C_(1 )、C_2)的AFPSO (Adaptive Fuzzy Particle Swarm Optimization)以及使用歸屬函數當作粒子權重的FPSO (Fuzzy Particle Swarm Optimization)作結合，並且應用在射頻積體電路(Radio Frequency Integrated Circuits，RFIC)的螺旋(Spiral)電感之最佳化設計。藉由模糊理論來調適粒子群演算法的加速因子，目的是希望能不斷的拓展搜索範圍以及找尋新的最佳解。並且參考最佳粒子以外第二好、第三好等粒子的搜尋經驗納入考慮，使其方法能改善傳統的粒子群演算法提早落入區域最佳解的缺點。另外，為了能分析所提出的演算法效能以及適用性，我們使用16種標準測試函數來模擬並且與數十種不同的粒子群演算法做比較。經由模擬結果顯示，本論文提出的方法確實能有效的改善原始PSO (Particle Swarm Optimization)的性能，並且應用於螺旋電感上能有效提升品質因數。;In this thesis, we propose a variant algorithm for Particle Swarm Optimization (PSO) which is called Adaptive Weighted Fuzzy Particle Swarm Optimization (AWFPSO). The algorithm combines two methods: AFPSO which uses fuzzy rules to adjust the acceleration parameters of PSO, and FPSO which manipulates membership function values to obtain weights. We also take the second best particle into consideration to prevent the AWFPSO from falling into local optimum too earlier. The performance of AWFPSO is compared with several PSO algorithms in the literature by utilizing sixteen benchmark functions. Finally, we apply AWFPSO to optimizing the design of the spiral inductor of Radio Frequency Integrated Circuits (RFIC). From experimental results, the proposed method improves the performance of RFIC by enhancing the quality factor of the designed spiral inductor.
