摘要
基于弹性流体动力润滑理论,以额定动载荷最大和膜厚比最大为目标函数,通过对圆柱滚子轴承多目标优化数学模型的研究,应用退火罚函数方法将其转化为无约束单目标优化问题.采用遗传算法求解含连续及离散变量的优化设计问题,同时针对遗传算法的局限性,采用了实数编码,调整了适应函数,设计了基于确定性准则的杂交算子和变异算子,借鉴了优解保留策略,结合模拟退火算法,形成了遗传模拟退火算法,该算法发挥了模拟退火与遗传算法各自的特点.算例表明了该优化方法的有效性.
Based on the theory of elasto-hydrodynamic lubrication （EHL）, multi-objective optimization model of cylindrical roller bearing is established, in which the maximum load rating and the maximum film thickness ratio are taken as objective functions,and the problem is transformed to unconstrained single-ob- jective optimization problem by annealing dynamic function. The model consists of continuous discrete ar- guments, and aimed at the limitation of genetic algorithm. Combining with the simulated annealing algo- rithm and using the retention strategy of optimal solution for reference, the genetic simulated annealing al- gorithm which combines the advantages of improved genetic algorithm and simulated annealing algorithm is formed by adopting the real coding, adjusting the fitness function and designing the crossover operator and mutation operator according to the deterministic criteria. The numerical example shows that the algorithm is efficient.
出处
《兰州交通大学学报》
CAS
2013年第6期141-145,共5页
Journal of Lanzhou Jiaotong University
基金
铁道车辆热工教育部重点实验室开放基金(TE011006)
甘肃省自然科学基金(1112RJZA045)
