7.5 Typical Technology
249
ensure accurate altitude relative to lunar surface after the end of main deceleration,
ranging correction was used in the later stage of main deceleration; after the quick
attitude adjustment phase, the lander remained relatively stable attitude to the Moon.
Velocity accuracy requirements were getting higher and higher. At this time, the
velocity correction was introduced. In the slow descending, due to the low height
of the lunar lander to lunar surface, the engine plume would cause lunar dust to
adversely affect the range and velocity sensors. Therefore, pure inertial navigation
was restored in this phase.
1. Inertial Navigation
The basic navigation method used in the powered descent process was inertial navigation. The attitude, position, and velocity of the lunar lander could be extrapolated
based on the angular velocity and non-gravity acceleration measurement information provided by IMU on the lunar lander (including gyros and accelerometers) and
integration of translational and rotational kinematic equations of the lunar lander.
The major error sources in the inertial navigation system included:
(1) Initial reference error: The initial position and velocity were provided by
ground tracking system, the position accuracy was about 1 km, and the velocity
accuracy was about 1 m/s. The initial attitude was given by the star sensor and
the error was about tens of arc-second.
(2) IMU error. There were IMU errors including IMU installation errors and
measurement errors. During LTO, on-orbit calibration of various IMU errors
was performed.
(3) Gravitational field model error. Gravitational field model used a two-body
model. For the brief 12 min of powered descent phase, the altitude calculation
error caused by the error of two-body gravitational field model did not exceed
60 m, which was within an acceptable range.
(4) The error caused by the integration step. From the perspective of improving
the accuracy of inertial navigation, it was desirable that the sampling period of
the IMU should be as short as possible. Due to the ability of the lunar lander
components, it was impossible to achieve a very high sampling frequency, so
a four-sample algorithm was used for the attitude updating.
2. Range Correction and Multi-Beam Combination
The Chang’E-3 lunar lander was equipped with a Laser Rangefinder and a MRVS.
Both sensors could measure the distance from the lunar lander to lunar surface to
correct the altitude information of the inertial navigation system.
For a certain beam, the kinematics equation of the vertical channel of the lunar
lander and the ranging sensor measurement equation were established, and the
Kalman filter method was used to calculate the gain matrix. The preset gain matrix
function on the control computer could be obtained by calculating the steady-state
gain matrix for different altitudes based on inertial navigation performance, ranging
sensor performance, and lunar surface fluctuations, as well as fitting the trend with
curve function of altitude.
249
ensure accurate altitude relative to lunar surface after the end of main deceleration,
ranging correction was used in the later stage of main deceleration; after the quick
attitude adjustment phase, the lander remained relatively stable attitude to the Moon.
Velocity accuracy requirements were getting higher and higher. At this time, the
velocity correction was introduced. In the slow descending, due to the low height
of the lunar lander to lunar surface, the engine plume would cause lunar dust to
adversely affect the range and velocity sensors. Therefore, pure inertial navigation
was restored in this phase.
1. Inertial Navigation
The basic navigation method used in the powered descent process was inertial navigation. The attitude, position, and velocity of the lunar lander could be extrapolated
based on the angular velocity and non-gravity acceleration measurement information provided by IMU on the lunar lander (including gyros and accelerometers) and
integration of translational and rotational kinematic equations of the lunar lander.
The major error sources in the inertial navigation system included:
(1) Initial reference error: The initial position and velocity were provided by
ground tracking system, the position accuracy was about 1 km, and the velocity
accuracy was about 1 m/s. The initial attitude was given by the star sensor and
the error was about tens of arc-second.
(2) IMU error. There were IMU errors including IMU installation errors and
measurement errors. During LTO, on-orbit calibration of various IMU errors
was performed.
(3) Gravitational field model error. Gravitational field model used a two-body
model. For the brief 12 min of powered descent phase, the altitude calculation
error caused by the error of two-body gravitational field model did not exceed
60 m, which was within an acceptable range.
(4) The error caused by the integration step. From the perspective of improving
the accuracy of inertial navigation, it was desirable that the sampling period of
the IMU should be as short as possible. Due to the ability of the lunar lander
components, it was impossible to achieve a very high sampling frequency, so
a four-sample algorithm was used for the attitude updating.
2. Range Correction and Multi-Beam Combination
The Chang’E-3 lunar lander was equipped with a Laser Rangefinder and a MRVS.
Both sensors could measure the distance from the lunar lander to lunar surface to
correct the altitude information of the inertial navigation system.
For a certain beam, the kinematics equation of the vertical channel of the lunar
lander and the ranging sensor measurement equation were established, and the
Kalman filter method was used to calculate the gain matrix. The preset gain matrix
function on the control computer could be obtained by calculating the steady-state
gain matrix for different altitudes based on inertial navigation performance, ranging
sensor performance, and lunar surface fluctuations, as well as fitting the trend with
curve function of altitude.