直接侧向力控制火箭弹的微机电系统速率陀螺滤波方法

Filtering Algorithm for Micro-electro-mechanical Gyroscope of Rocket with Lateral Thrust Control

火箭弹采用直接侧向力轨道控制时会使弹体姿态产生较高频率的振动,其对微机电系统(MEMS)速率陀螺测量会产生不利影响。目前的滤波算法通常假设弹体姿态在短时间内是缓变的,因此其对高频运动的弹体姿态滤波效果较差。根据气动力学分析可知,弹体在短暂冲击后的高频振荡属于自然频率下的阻尼振荡,而弹体的自然频率和阻尼是可预先估计的,因此可应用基于摆动Markov模型的交互多模型(IMM)来描述弹体姿态的不确定变化,并利用点火时间动态调整模型概率转移矩阵。仿真实验表明,基于摆动Markov的IMM滤波算法可对MEMS速率陀螺输出有较好的滤波效果。

The attitude of rocket oscillates in high frequency when it is controlled by lateral thrust,and its micro-electro-mechanical systems（MEMS）gyroscope will be affected.Most of current filtering algorithms assume that the roket attitude changes slowly in a short interval,and it will generate unacceptable modeling error.With aerodynamic analysis,it can be seen that the attitude oscillation is a natural frequency damping oscillation and the natural frequency and damp ratio can be estimated in advance.Thus,a novel filtering algorithm is proposed,which applies the interacting multiple models（IMM） with swing Markov model to describe the uncertain roket attitude motions.Furthermore,the dynamic adjustment mechanism of state transferring probability matrix with the lateral thrust control command can be introduced.Finally,the numerical simulation indicates that IMM filtering algorithm with swing Markov model can accurately trace the attitude vibration.