为提高模态密集大型星栽天线在轨模态参数辨识精度,提出基于均匀设计的遗传算法对传感器数量和位置进行优化部署。根据模态空间H2范数确定参振模态阶数;分别以测量信息正交性和能量最大为优化目标,引入均匀设计方法对传感器数量、种群规模、交叉概率和变异概率4个组合参数进行选取;采用特征实现算法进行在轨模态参数辨识研究。数学仿真表明:所提出的方法可以有效解决传感器数量和位置的优化部署问题,避免了以往人为经验选取参数时存在的主观性和盲目性问题。
In order to improve the precision of modal parameter identification for large close mode satellite antenna reflector, the genetic algorithm based on uniform design method was proposed to optimize the numbers and locations of sensors. The order of observed modal shape was selected by the energy criterion based on H2 norm in modal space. According to the largest orthogonality and the highest energy of measured information as the optimal objectives respectively, the uniform design method was introduced to optimize the four compounding parameters in genetic algorithm, including., sensor number, population size, probability of crossover and mutation. The eigensystem realization algorithm (ERA) was utilized in on-orbit modal identification. The numerical simulation results show that the proposed method can solve the problem of optimizing the numbers and locations of the sensors effectively, and avoid the blindness in parameters selection.