快速平衡粒子群最佳化方法(Swiftly balanced particle swarm optimization, SBSPO)是一種改良的粒子最佳化方法(Particle swarm optimization, PSO)，利用改變加速係數來平衡個體經驗及群體經驗，改善粒子最佳化方法易落入區域最佳解的缺點。利用粒子群收斂狀況決定加速係數大小，加速係數大小被設定為三段線性直線，一旦得知粒子群收斂狀況，則可求得一組適合的加速係數。因為能利用粒子群收斂狀況快速求得一組加速係數大小，又因這組加速係數能平衡個體經驗與群體經驗，因此名為快速平衡粒子群最佳化方法。本文也將二次內插演算法(Quadratic interpolation)與快速平衡粒子群最佳化方法(SBPSO)做結合，名為SBPSO-QI。另外提出考慮兩個群體最佳解來改良粒子最佳化方法，讓PSO在處理複雜問題時，能跳出區域最佳解，求得全域最佳解，並將此方法與SBPSO做結合，名為SBPSO-2G。並將提出的SBPSO、SBPSO-QI與SBPSO-2G與8種不同的粒子最佳化方法做比較。經模擬結果顯示，提出的方法對於多數的測詴函數均有較優越的表現。本文所提出的快速平衡粒子群最佳化方法保有粒子最佳化方法容易實現的特性，同時改良粒子最佳化方法易落入區域最佳解的缺點。 Swiftly balanced particle swarm optimization (SBPSO) is a new variant of particle swarm optimization which can quickly balanced the personal and social experience. A new strategy of the acceleration coefficients makes SBPSO more effective, because the swarm can efficiently adjust the velocity by changing the acceleration coefficients. The acceleration coefficients of SBPSO are obtained by three segment line dependent on the swarm convergence. The advantage is that SBPSO become more accurate and also easy to implement. The acceleration coefficients of SBPSO can be applied to many variants of PSO. In this paper, incorporating the acceleration coefficients of SBPSO and The quadratic interpolation PSO, named SBPSO-QI. In the result section, compared the proposed SBPSO and SBPSO-QI with standard PSO (SPSO), quadratic interpolation PSO (QIPSO), unified PSO (UPSO), fully informed particle swarm (FIPS), dynamic multi-swarm PSO (DMSPSO), adaptive fuzzy PSO (AFPSO), and PSO with time-varying acceleration coefficients (PSO-TVAC) across sixteen benchmark functions.
