Controlling vehicle attitude requires sensors to measure vehicle orientation, actuators to apply the torques needed to re-orient the vehicle to a desired attitude, and algorithms to command the actuators based on (1) sensor measurements of the current attitude and (2) specification of a desired attitude. The integrated field that studies the combination of sensors, actuators and algorithms is called "Guidance, Navigation and Control" (GNC).
A spacecraft's attitude must typically be stabilized and controlled for a variety of reasons. It is oftentimes needed so that the spacecraft high-gain antenna may be accurately pointed to Earth for communications, so that onboard experiments may accomplish precise pointing for accurate collection and subsequent interpretation of data, so that the heating and cooling effects of sunlight and shadow may be used intelligently for thermal control, and also for guidance: short propulsive maneuvers must be executed in the right direction.
There are two principal approaches to stabilizing attitude control on spacecraft:
There are advantages and disadvantages to both spin stabilization and three-axis stabilization. Spin-stabilized craft provide a continuous sweeping motion that is desirable for fields and particles instruments, as well as some optical scanning instruments, but they may require complicated systems to de-spin antennas or optical instruments that must be pointed at targets for science observations or communications with Earth. Three-axis controlled craft can point optical instruments and antennas without having to de-spin them, but they may have to carry out special rotating maneuvers to best utilize their fields and particle instruments. If thrusters are used for routine stabilization, optical observations such as imaging must be designed knowing that the spacecraft is always slowly rocking back and forth, and not always exactly predictably. Reaction wheels provide a much steadier spacecraft from which to make observations, but they add mass to the spacecraft, they have a limited mechanical lifetime, and they require frequent momentum desaturation maneuvers, which can perturb navigation solutions because of accelerations imparted by the use of thrusters.
Many spacecraft have components that require articulation. Voyager and Galileo, for example, were designed with scan platforms for pointing optical instruments at their targets largely independently of spacecraft orientation. Many spacecraft, such as Mars orbiters, have solar panels that must track the Sun so they can provide electrical power to the spacecraft. Cassini's main engine nozzles are steerable. Knowing where to point a solar panel, or scan platform, or a nozzle — that is, how to articulate it — requires knowledge of the spacecraft's attitude. Because AACS keeps track of the spacecraft's attitude, the Sun's location, and Earth's location, it can compute the proper direction to point the appendages. It logically falls to one subsystem, then, to manage both attitude and articulation. The name AACS may even be carried over to a spacecraft even if it has no appendages to articulate.
Many sensors generate outputs that reflect the rate of change in attitude. These require a known initial attitude, or external information to use them to determine attitude. Many of this class of sensor have some noise, leading to inaccuracies if not corrected by absolute attitude sensors.
Gyroscopes are devices that sense rotation in three-dimensional space without reliance on the observation of external objects. Classically, a gyroscope consists of a spinning mass, but there are also "ring laser gyros" utilizing coherent light reflected around a closed path. Another type of "gyro" is a hemispherical resonator gyro where a crystal cup shaped like a wine glass can be driven into oscillation just as a wine glass "sings" as a finger is rubbed around its rim. The orientation of the oscillation is fixed in inertial space, so measuring the orientation of the oscillation relative to the spacecraft can be used to sense the motion of the spacecraft with respect to inertial space.
Motion reference units are a kind of inertial measurement unit with single- or multi-axis motion sensors. They utilize MEMS gyroscopes. Some multi-axis MRUs are capable of measuring roll, pitch, yaw and heave. They have applications outside the aeronautical field, such as:
This class of sensors sense the position or orientation of fields, objects or other phenomena outside the spacecraft.
A horizon sensor is an optical instrument that detects light from the 'limb' of Earth's atmosphere, i.e., at the horizon. Thermal infrared sensing is often used, which senses the comparative warmth of the atmosphere, compared to the much colder cosmic background. This sensor provides orientation with respect to Earth about two orthogonal axes. It tends to be less precise than sensors based on stellar observation. Sometimes referred to as an Earth sensor.
Similar to the way that a terrestrial gyrocompass uses a pendulum to sense local gravity and force its gyro into alignment with Earth's spin vector, and therefore point north, an orbital gyrocompass uses a horizon sensor to sense the direction to Earth's center, and a gyro to sense rotation about an axis normal to the orbit plane. Thus, the horizon sensor provides pitch and roll measurements, and the gyro provides yaw. See Tait-Bryan angles.
An Earth sensor is a device that senses the direction to Earth. It is usually an infrared camera; nowadays the main method to detect attitude is the star tracker, but Earth sensors are still integrated in satellites for their low cost and reliability.
A magnetometer is a device that senses magnetic field strength and, when used in a three-axis triad, magnetic field direction. As a spacecraft navigational aid, sensed field strength and direction is compared to a map of Earth's magnetic field stored in the memory of an on-board or ground-based guidance computer. If spacecraft position is known then attitude can be inferred.
Control algorithms are computer programs that receive data from vehicle sensors and derive the appropriate commands to the actuators to rotate the vehicle to the desired attitude. The algorithms range from very simple, e.g. proportional control, to complex nonlinear estimators or many in-between types, depending on mission requirements. Typically, the attitude control algorithms are part of the software running on the hardware, which receives commands from the ground and formats vehicle data telemetry for transmission to a ground station.
The attitude control algorithms are written and implemented based on requirement for a particular attitude maneuver. Asides the implementation of passive attitude control such as the gravity-gradient stabilization, most spacecrafts make use of active control which exhibits a typical attitude control loop. The design of the control algorithm depends on the actuator to be used for the specific attitude maneuver although using a simple proportional–integral–derivative controller (PID controller) satisfies most control needs.
The appropriate commands to the actuators are obtained based on error signals described as the difference between the measured and desired attitude. The error signals are commonly measured as euler angles (Φ, θ, Ψ), however an alternative to this could be described in terms of direction cosine matrix or error quaternions. The PID controller which is most common reacts to an error signal (deviation) based on attitude as follows
where is the control torque, is the attitude deviation signal, and are the PID controller parameters.
A simple implementation of this can be the application of the proportional control for nadir pointing making use of either momentum or reaction wheels as actuators. Based on the change in momentum of the wheels, the control law can be defined in 3-axes x, y, z as
This control algorithm also affects momentum dumping.
Another important and common control algorithm involves the concept of detumbling, which is attenuating the angular momentum of the spacecraft. The need to detumble the spacecraft arises from the uncontrollable state after release from the launch vehicle. Most spacecraft in low earth orbit (LEO) makes use of magnetic detumbling concept which utilizes the effect of the earth's magnetic field. The control algorithm is called the B-Dot controller and relies on magnetic coils or torque rods as control actuators. The control law is based on the measurement of the rate of change of body-fixed magnetometer signals.
where is the commanded magnetic dipole moment of the magnetic torquer and is the proportional gain and is the rate of change of the Earth's magnetic field.
Attitude control can be obtained by several mechanisms, specifically:
Vernier thrusters are the most common actuators, as they may be used for station keeping as well. Thrusters must be organized as a system to provide stabilization about all three axes, and at least two thrusters are generally used in each axis to provide torque as a couple in order to prevent imparting a translation to the vehicle. Their limitations are fuel usage, engine wear, and cycles of the control valves. The fuel efficiency of an attitude control system is determined by its specific impulse (proportional to exhaust velocity) and the smallest torque impulse it can provide (which determines how often the thrusters must fire to provide precise control). Thrusters must be fired in one direction to start rotation, and again in the opposing direction if a new orientation is to be held. Thruster systems have been used on most manned space vehicles, including Vostok, Mercury, Gemini, Apollo, Soyuz, and the Space Shuttle.
To minimize the fuel limitation on mission duration, auxiliary attitude control systems may be used to reduce vehicle rotation to lower levels, such as small ion thrusters that accelerate ionized gases electrically to extreme velocities, using power from solar cells.
The entire space vehicle itself can be spun up to stabilize the orientation of a single vehicle axis. This method is widely used to stabilize the final stage of a launch vehicle. The entire spacecraft and an attached solid rocket motor are spun up about the rocket's thrust axis, on a "spin table" oriented by the attitude control system of the lower stage on which the spin table is mounted. When final orbit is achieved, the satellite may be de-spun by various means, or left spinning. Spin stabilization of satellites is only applicable to those missions with a primary axis of orientation that need not change dramatically over the lifetime of the satellite and no need for extremely high precision pointing. It is also useful for missions with instruments that must scan the star field or Earth's surface or atmosphere. See spin-stabilized satellite.
These are electric motor driven rotors made to spin in the direction opposite to that required to re-orient the vehicle. Because momentum wheels make up a small fraction of the spacecraft's mass and are computer controlled, they give precise control. Momentum wheels are generally suspended on magnetic bearings to avoid bearing friction and breakdown problems. To maintain orientation in three dimensional space a minimum of three must be used, with additional units providing single failure protection. See Euler angles.
These are rotors spun at constant speed, mounted on gimbals to provide attitude control. Although a CMG provides control about the two axes orthogonal to the gyro spin axis, triaxial control still requires two units. A CMG is a bit more expensive in terms of cost and mass, because gimbals and their drive motors must be provided. The maximum torque (but not the maximum angular momentum change) exerted by a CMG is greater than for a momentum wheel, making it better suited to large spacecraft. A major drawback is the additional complexity, which increases the number of failure points. For this reason, the International Space Station uses a set of four CMGs to provide dual failure tolerance.
Small solar sails (devices that produce thrust as a reaction force induced by reflecting incident light) may be used to make small attitude control and velocity adjustments. This application can save large amounts of fuel on a long-duration mission by producing control moments without fuel expenditure. For example, Mariner 10 adjusted its attitude using its solar cells and antennas as small solar sails.
In orbit, a spacecraft with one axis much longer than the other two will spontaneously orient so that its long axis points at the planet's center of mass. This system has the virtue of needing no active control system or expenditure of fuel. The effect is caused by a tidal force. The upper end of the vehicle feels less gravitational pull than the lower end. This provides a restoring torque whenever the long axis is not co-linear with the direction of gravity. Unless some means of damping is provided, the spacecraft will oscillate about the local vertical. Sometimes tethers are used to connect two parts of a satellite, to increase the stabilizing torque. A problem with such tethers is that meteoroids as small as a grain of sand can part them.
Coils or (on very small satellites) permanent magnets exert a moment against the local magnetic field. This method works only where there is a magnetic field against which to react. One classic field "coil" is actually in the form of a conductive tether in a planetary magnetic field. Such a conductive tether can also generate electrical power, at the expense of orbital decay. Conversely, by inducing a counter-current, using solar cell power, the orbit may be raised. Due to massive variability in Earth's magnetic field from an ideal radial field, control laws based on torques coupling to this field will be highly non-linear. Moreover, only two-axis control is available at any given time meaning that a vehicle reorient may be necessary to null all rates.
There exist two main passive control types for satellites. The first one uses gravity gradient, and it leads to four stable states with the long axis (axis with smallest moment of inertia) pointing towards Earth. As this system has four stable states, if the satellite has a preferred orientation, e.g. a camera pointed at the planet, some way to flip the satellite and its tether end-for-end is needed. The other passive system orients the satellite along Earth's magnetic field thanks to a magnet. These purely passive attitude control systems have limited pointing accuracy, because the spacecraft will oscillate around energy minima. This drawback is overcome by adding damper, which can be hysteretic materials or a viscous damper. The viscous damper is a small can or tank of fluid mounted in the spacecraft, possibly with internal baffles to increase internal friction. Friction within the damper will gradually convert oscillation energy into heat dissipated within the viscous damper.