论文部分内容阅读
讨论能拦戴高空高速目标、并带有使用卡尔曼滤波器组合的自适应实时估算电路的学习网络。卡尔曼滤波器组合中有一个内部学习程序,能在飞行中进行天线罩误差校正,而天线罩误差补偿在倾斜转弯(BTT)导弹的控制和制导系统的回路中进行,这样可以得到满意的脱靶距离性能指数。在制导系统和自动驾驶仪指令中包含有天线罩误差斜率的估算值,这样可以为修正的比例导航系统和最佳控制器提供最佳俯仰速率补偿。已经为修正比例导航系统推导出一个解析表达式,在理论上该式是按最佳控制规律引入俯仰速率补偿。上述的天线罩误差补偿取决于制导指令产生的导弹俯仰角速度,同时又考虑了天线罩误差瞄准线速率。初步结果表明,天线罩误差的补偿可以使脱靶距离和俯仰加速度的时间特性得到明显的改善。卡尔曼滤波器组合用来提高天线罩斜率估算时的动态响应的时间特性。和其他自适应方案相比,它的主要和独特的一点是在计算临界的后验概率的加权系数时,使用了测量-预测-测量技术,一个随机过程的半马尔可夫统计法通常用来进行中间预测。对一个误差斜率变化较大的天线罩的数字模拟结果表明,该方法十分成功。本研究的目的是尽快将天线罩误差校正补偿方案应用于三维拦截仿真的滚动通道中。
Discusses learning networks that can hold high-altitude, high-speed targets with an adaptive real-time estimation circuit using a combination of Kalman filters. An internal learning process in the Kalman filter combination enables radome error correction in flight, while radome error compensation is performed in the control and guidance system loops of BTT missiles so that a satisfactory off-target can be obtained Distance performance index. The guidance system and autopilot commands include estimates of the radome error slope to provide optimum pitch rate compensation for the modified proportional navigation system and optimal controller. An analytical expression has been derived for a modified proportional navigation system, which theoretically introduces pitch rate compensation according to the optimal control law. The above-mentioned compensation of the radome depends on the pitch velocity of the missile generated by the guidance instruction, and at the same time, the linear error of the radome is considered. The preliminary results show that the compensation of radome error can obviously improve the time characteristics of off-target range and pitch acceleration. The Kalman filter combination is used to improve the time response of the dynamic response when the radome’s slope is estimated. Compared with other adaptive schemes, its main and unique point is that the measurement-prediction-measurement technique is used in calculating the weighted coefficients of the critical posteriori probabilities. A semi-Markov statistical method for stochastic processes is usually used Interim forecast. A numerical simulation of a radome with a large change in slope of error shows that the method is very successful. The purpose of this study is to apply the radome error correction compensation scheme to the rolling channel of 3D interception simulation as soon as possible.