CAJ | 학술논문

为了提高粒子群算法全局寻优能力，提出一种远邻粒子群算法，该算法引入邻域算子概念，每个粒子选择与自身欧氏距离较远的粒子建立邻域，邻域中粒子的数目用邻域算子表示。测试函数实验结果表明，该算法在一定程度上消除了标准粒子群算法容易陷入局部最优的缺点。应用远邻粒子群算法对Delta机器人进行优化设计，结果证实：所提出的远邻粒子群算法较标准粒子群算法具有更好的寻优能力，比邻居递增粒子群算法搜索精度更高。
위료제고입자군산법전국심우능력，제출일충원린입자군산법，해산법인입린역산자개념，매개입자선택여자신구씨거리교원적입자건립린역，린역중입자적수목용린역산자표시。측시함수실험결과표명，해산법재일정정도상소제료표준입자군산법용역함입국부최우적결점。응용원린입자군산법대Delta궤기인진행우화설계，결과증실：소제출적원린입자군산법교표준입자군산법구유경호적심우능력，비린거체증입자군산법수색정도경고。
In order to improve the global search capability ,a distant neighborhood PSO was pres-ented .In the algorithm ,each particle had its own neighborhood by selecting particles far away from it in Euclidean distance .Neighborhood operator was adopted to control the number of parti-cles in its neighborhood .Experimental results of benchmark functions indicated that the algo-rithm eliminated the standard PSO's weakness of easily converging to local optimal value .The distant neighborhood PSO was applied to optimization design of Delta robot .It turns out that the distant neighborhood PSO has better search capability than the standard PSO and higher search precision than neighbors increasing PSO .