基于改进粒子群算法的旋转弹气动系数快速辨识

Fast Identification of Aerodynamic Coefficients of Spin-stabilized Projectile Based on Improved Particle Swarm Optimization

为了准确、快速地辨识旋转弹主要气动系数,建立了一个基于改进粒子群算法的气动系数快速辨识模型。该辨识模型以四自由度修正质点弹道模型为基础,以最小标准欧几里德距离为辨识准则,利用弹丸自由飞行试验测得的速度和转速数据,可同时辨识出弹丸的零升阻力系数、升力系数导数、极阻尼力矩系数导数以及马格努斯力系数导数。利用某155 mm旋转弹仿真所得的弹道数据对提出的气动辨识模型进行验证。结果表明:与气动系数理论值相比,零升阻力系数、升力系数导数与极阻尼力矩系数辨识值的平均相对误差较小,当马赫数在0.8～1.25范围内马格努斯力系数导数相对误差约为30%～50%,但马赫数在1.25～2.7范围内其误差较大;根据气动系数辨识值计算出的弹道数据与仿真弹道数据相比,射程在26 km时相差约为8 m,速度变化完全一致;相比于标准粒子群算法,提出的改进粒子群算法具有更快的收敛速度。

In order to accurately and rapidly identify the main aerodynamic coefficients of spin-stabilized projectile,a fast identification model of aerodynamic coefficients based on an improved particle swarm optimization(PSO)was established.The identification model is based on modified particle trajectory model with four degrees of freedom.Taking the least standardized Euclidean distance as the identification criterion,the zero-lift drag coefficient,the lift coefficient,the Magnus force coefficient and the roll damping moment coefficient can be identified simultaneously by using the velocity and speed data measured by the projectile free-flight test.The proposed aerodynamic identification model was validated by using simulation trajectory data of a155mm spin-stabilized projectile.The results show that compared with the theoretical value of aerodynamic coefficients,the average relative error of the derivative values of zero-lift resistance coefficient,lift coefficient derivative and rolling damping moment coefficient was small.The relative error of the Magnus force coefficient derivative is about30%to50%while Mach number ranging from0.8to1.25,but the error is larger while Mach number ranging from1.25to2.7.The trajectory data calculated according to the aerodynamic coefficient identification value were compared with simulation trajectory data.The difference between the landing points was8m at a range of26km,and the speed change is the same.Compared with the standard PSO,the improved PSO has a faster convergence speed.