摘要
基于CQ2系列气缸的优化模型,设计了一种改进遗传算法,此算法中采用交叉和变异操作点数随种群多样性函数而变化,在寻优后期,采用Powell法进行加速。高维复杂函数优化结果表明:该算法有效地解决了早熟收敛问题,同时在目标函数的寻优后期快速逼近最优值,达到全局最优。依据优化方案对CQ2系列气缸实施加速寿命试验,试验结果证明:此方案在不影响参数估计精度的前提下,有效地缩短了周期,降低了成本。
Based on optimization model of CQ2-series cylinder, an improved genetic algorithm was designed, in which the points of crossover and mutation in the algorithm vary with the population diversity function, and during later period of optimization, Powell method was employed for acceleration. The optimization results of the complex function with high dimensions show that the the problem of premature convergence is solved effectively and the optimal value was approached rapidly in later optimization, to get global optimum. Accelerated life test of CQ2-series cylinders was conducted based on the optimization scheme. The tested results verify that the scheme can shorten the period and reduce the cost effectively under the premise of without influence on the parameter estimation precision.
出处
《兵工学报》
EI
CAS
CSCD
北大核心
2009年第11期1493-1498,共6页
Acta Armamentarii
基金
北京航空航天大学与日本SMC国际合作项目资助
关键词
机械学
改进遗传算法
气缸
恒定应力加速寿命试验
优化模型
参数估计
mechanics
improved genetic algorithm
cylinder
constant stress accelerated life testing
optimum model
parameter estimation
