自适应遗传PID算法在风洞风速控制中的应用  

Application of PID based on adaptive genetic algorithms in wind velocity control system of wind tunnels

作　　者：尼文斌[1] 董金刚[1] 刘书伟[1] 贺丽慧[1] 付增良[1] 

机构地区：[1]中国航天空气动力技术研究院,北京100074

出　　处：《实验流体力学》2015年第5期84-89,共6页Journal of Experiments in Fluid Mechanics

基　　金：国家自然科学基金(11302214)

摘　　要：风速控制是风洞的核心控制部分,风速控制系统的优劣直接影响风洞性能指标,为了完成FDxx风洞的风速控制系统,设计了一种基于自适应在线遗传算法的PID参数整定方法,在风洞气源资源有限的情况下,快速建立流场,确保流场稳定时间。首先对控制参数进行联合编码,在种群个体进化前期采用锦标赛精英保留策略,后期采用基于轮盘赌非线性选择方法,加快算法收敛速度,同时避免了算法过早陷入局部最优,交叉选用单点交叉,变异采用均匀取反法,动态调整过程为了减小甚至避免超调,采用误差绝对值及误差和误差变化率加权方式设计目标函数,并采取了惩罚措施,即一旦产生超调,将超调量作为最优指标的一项,现场测试验证了算法的可靠性及实用性。In order to accomplish the wind speed control system for a FDxx wind tunnel, a PID parameter setting method based on adaptive online genetic algorithm was designed. In the case of the limited gas source for a wind tunnel test, the algorithm can quickly establish the flow field and ensure the stability time of the flow. First, joint coding is done with these parameters. With elitist strategy adopted in early genetic evolution stage, a non-linear selection method based on roulette selection is introduced in the last period. So optimum individuals are preserved and the 15opulation diversity is increased. The fixed value thus would not be exceeded in the dynamic adiustment process. The objective function is set up with the error absolutes and variation rate accompanied by penalty functions. When the fixed value is exceeded, the excess is taken as the optimal. This method has been applied in a field debugging of FDXX wind tunnel. The result shows that it is fast, effective and reliable to use online adaptive genetic algorithms to set the PID control parameters.

关 键 词：风洞 风速控制 参数调整 PID算法 遗传算法 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]