【摘要】 根据杆长约束条件,建立求6自由度一般6-SPS并联机构位置正解的无约束优化模型,再应用差分进化(Differential evolution,DE)算法求解该问题。针对基本DE算法可能出现进化停滞或陷入局部极值区域的缺点,提出一种引入新个体的自适应策略,以增强算法全局优化性能。将引入新个体的自适应策略融入DE算法,并使用混合变异算子及基于三角函数扰动的缩放因子和交叉因子,形成自适应差分进化(Adaptive DE,ADE)算法。数值结果表明,对于一般6-SPS并联机构正运动学分析问题,ADE算法能以较少计算开销求出全部高精度位置正解。通过与基本DE算法、自适应变异粒子群算法和改进人工蜂群算法比较,验证了ADE算法的收敛精度和计算稳健性指标优于对比算法。更多还原

【Abstract】 According to the constrained length of the bars,the unconstrained optimization model is constructed to formulate the forward position analysis for a 6 degrees-of-freedom general 6-SPS parallel mechanism,and the differential evolution( DE) algorithm is employed to solve this problem. Aiming at the basic DE algorithm has some weaknesses,such as evolutionary stagnation and falling into local extremum domains,an adaptive strategy for introducing new individuals is presented to improve its optimization performance. This strategy is dynamically incorporated into DE algorithm to form an adaptive DE( ADE) algorithm,which adopts hybrid mutant operators,and scale and cross factors with trigonometric function perturbations. The numerical results of the forward kinematic analysis for a general 6-SPS parallel mechanism show that ADE algorithm can obtain all high accuracy solutions with lighter computational cost. Furthermore,the ADE algorithm is compared with basic DE,the adaptive mutation particle swarm optimization,and improved artificial bee colony algorithms,and it is verified that the former outperforms the compared algorithms in term of convergence precision and calculated robustness.更多还原