基于卡尔曼滤波提高陀螺传感器测量精度的研究

发布时间：2018-08-08 16:06

【摘要】：航空光电稳定平台凭借图像探测装置实现对目标的自动跟踪和识别,被广泛应用于军事领域以及公安、环境监测等民用环境中。其中,陀螺传感器是航空光电稳定平台的核心敏感传感器,它被安装在运动载体上,敏感载体速度的变化从而隔离扰动,使航空光电稳定平台尽可能的保持稳定,进而保证安装在平台内部载荷的视轴具有稳定指向。由于受到外界环境的干扰以及自身结构误差等因素的影响,航空光电稳定平台中的陀螺传感器在敏感平台速度变化时,往往会引入随机漂移和测量噪声,导致直流驱动电机的输出力矩无法完全精准地驱动平台的框架运动,如此一来,将会造成视轴不能稳定在目标理想位置。所以,想要提高航空光电稳定平台的视轴稳定精度,必须对陀螺传感器所输出的信号进行有效的滤波处理。卡尔曼滤波根据状态方程对动态系统进行描述,以线性最小方差作为估计准则,利用预测和修正的方法,对系统的状态进行最优估计。卡尔曼滤波以其实时性好、存储空间小、易实现等优点而得到广泛应用。考虑到实际的工程需求,本论文在两轴四框架航空光电稳定平台的环境下,利用强跟踪卡尔曼滤波算法,来为航空光电稳定平台中陀螺传感器所输出的信号滤除噪声干扰,从而提升航空光电稳定平台的伺服系统速度环的控制精度,本文研究的主要内容如下:(1)首先根据两轴四框架航空稳定平台的工作原理,对系统中存在的各类干扰因素进行综合分析,着重分析了陀螺传感器的测量噪声为视轴稳定精度带来的不利影响。(2)为了对陀螺传感器的输出信号进行有效滤波,对陀螺信号建立数学模型。时间序列法是本文所采用的建模方法,利用系统输出的数据对系统内部的特性和外部因素进行一个推测,同时对系统的未来状况做出预测。考虑到算法的复杂程度以及实际应用中的计算量,采用AR(1)模型作为陀螺传感器的数学模型。该模型可以为卡尔曼滤波中的参数设置提供参考。(3)在陀螺传感器数学模型的基础上,构建卡尔曼滤波的状态方程和量测方程,针对陀螺信号的测量噪声,采用卡尔曼滤波技术对其进行抑制。同时引入强跟踪算法,将卡尔曼滤波中的状态预测方差矩阵不断调整,强迫输出误差序列保持不相关的特性,构造强跟踪卡尔曼滤波器,从而提高系统的鲁棒性。(4)针对本文提出的强跟踪卡尔曼滤波器,与工程上常用的二阶巴特沃斯滤波器进行对照试验,从而验证强跟踪卡尔曼滤波器的滤波效果与鲁棒性。实验结果表明:相较于二阶巴特沃斯滤波器,强跟踪卡尔曼滤波器的输出信号在幅值和方差上均有明显的改善,优化了控制系统的动态性能,保证了信号的实时性。同时,本文所提出的强跟踪卡尔曼滤波器具有更好的鲁棒稳定性。
[Abstract]:The aeronautical photoelectric stabilization platform is widely used in the military field, public security, environmental monitoring and other civil environments because of the automatic tracking and recognition of the target by the image detection device. Among them, gyroscope sensor is the core sensitive sensor of aeronautical optoelectronic stabilization platform. It is installed on the motion carrier. The change of the speed of the sensitive carrier can isolate the disturbance and make the aero-photoelectric stabilization platform as stable as possible. Furthermore, the view axis installed in the platform interior load has a stable direction. Because of the influence of the external environment and its own structure error, the gyroscope sensor in the aero-optoelectronic stabilization platform will introduce random drift and measurement noise when the speed of the sensitive platform changes. As a result, the output torque of DC drive motor can not drive the frame motion of the platform accurately, which will lead to the instability of the axis of view in the ideal position of the target. Therefore, in order to improve the accuracy of optical axis stabilization of aero-optoelectronic stabilization platform, the signal output from gyroscope must be filtered and processed effectively. Kalman filter is used to describe the dynamic system according to the equation of state. The linear minimum variance is used as the estimation criterion and the method of prediction and correction is used to estimate the optimal state of the system. Kalman filter is widely used because of its advantages of good real-time, small storage space and easy to be realized. Considering the actual engineering requirements, in this paper, the strong tracking Kalman filter algorithm is used to filter the noise interference for the signal output from the gyroscope sensor in the two-axis and four-frame aeronautical photovoltaic stabilization platform by using the strong tracking Kalman filter algorithm. The main contents of this paper are as follows: (1) according to the working principle of two-axis four-frame aeronautical stabilization platform, Based on the comprehensive analysis of all kinds of interference factors in the system, the adverse effects of the measurement noise of the gyroscope sensor on the stability accuracy of the optic axis are analyzed. (2) in order to effectively filter the output signal of the gyroscope sensor, The mathematical model of gyroscope signal is established. The time series method is the modeling method used in this paper. The system output data is used to predict the internal characteristics and external factors of the system and to predict the future status of the system. Considering the complexity of the algorithm and the computational complexity in practical application, AR (1) model is used as the mathematical model of gyroscope sensor. The model can provide reference for parameter setting in Kalman filter. (3) based on the mathematical model of gyroscope sensor, the state equation and measurement equation of Kalman filter are constructed, and the measurement noise of gyro signal is analyzed. Kalman filter technique is used to suppress it. At the same time, the strong tracking algorithm is introduced, the state prediction variance matrix in Kalman filter is adjusted constantly, and the output error sequence is forced to remain irrelevant, and a strong tracking Kalman filter is constructed. In order to improve the robustness of the system. (4) the strong tracking Kalman filter proposed in this paper is compared with the second order Butterworth filter which is commonly used in engineering to verify the filtering effect and robustness of the strong tracking Kalman filter. The experimental results show that compared with the second order Butterworth filter, the output signal of the strong tracking Kalman filter is obviously improved in amplitude and variance. The dynamic performance of the control system is optimized and the real-time performance of the signal is guaranteed. At the same time, the strong tracking Kalman filter proposed in this paper has better robust stability.
【学位授予单位】：中国科学院长春光学精密机械与物理研究所
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TN713;TP212

【相似文献】