论文部分内容阅读
“阿波罗”飞船数字计算机的设计程序是:当位置和速度坐标增值（相对估计值）超过某定值时,宇航员开始注意计算机工作过程,并将需要大修正量的坏原始参数取掉。问题在于,什么样的修正量可以认为是合理的或不合理的。 在登月舱和指挥舱计算机内为了对观测（测量）加工,远用卡尔曼滤波器（指示K）以保证改善飞船状态可变向量估计值（以P表示）。 当登月舱和指挥舱相遇时,进行无线电测距及其改变速率,并测量圆锥角和转动角。指挥舱计算机借助六分仪测出的距离、圆锥角和转动角值改善状态向量估计值。计算机的信息由卡尔曼滤波器加工。当线性模型适用时卡尔曼滤波器的估算结果与最小二乘法（以WLS表示）的结果相同。这一点将在下节证明。
The Apollo spacecraft digital computer was programmed to: As the position and velocity coordinate values increase (relative estimates) exceed a certain value, the astronauts begin to notice the computer’s work process and remove bad primitive parameters that require large corrections. The question is, what kind of correction can be considered reasonable or unreasonable. In the lunar module and the command module computer, the Kalman filter (K) is used far in order to process the observations (measurements) to ensure that the spacecraft state variable vector estimates (in P) are improved. When the lunar module meets the command module, the radio range and its rate of change are measured and the cone angle and the angle of rotation are measured. The command chamber computer improves the state vector estimates with the aid of the distance, cone angle, and rotation angle values measured by the sextant. The computer’s information is processed by a Kalman filter. The Kalman filter’s estimation results are the same as the least-squares method (expressed in WLS) when the linear model is applied. This will be proved in the next section.