Introduction to Satellite Attitude Determination and Control System (ADCS)

by Sam Lee
2021.07.18

     Every spacecraft has 6 degrees of freedom (DOF). Three of them are displacement, and the rest of them are rotation. The attitude determination and control system (ADCS) is responsible for the 3 DOF rotational control of the spacecraft.

     Let’s use the image-taking process for remote sensing satellites as an example to illustrate this attitude determination and control process. First, we need to know where the camera is pointing right now. This process is called attitude determination. Then, we apply torque to rotate the satellite to the desired orientation. This process is called attitude control.

The Attitude Determination

     The first thing we should notice in attitude determination is the “reference frame.” When we talk about the spacecraft’s attitude, we refer to its rotational relationship to some other reference frames adhered on an inertial frame like Earth-Centered Earth-Fixed Frame (ECEF) or Earth-Centered Inertial Frame (ECI). The following discussion will introduce you to common attitude sensors we use for satellites orbiting the earth. These include (Coarse/Fine) Sun sensor, Earth sensor, IMU, Star tracker, and Geomagnetic sensor.

Sun sensors

     Coarse Sun Sensor (CSS) consists of a single photodiode with a proper amplifier and filter optimized for measuring the sunlight’s incident angle. Sometimes four CSSs form a pyramid configuration to have a wider field of view (FOV). The attitude determination is based on the measured light intensity on each CSS.

Figure 1. CSS100- Coarse Sun Sensor made by Tensor Tech

     Fine Sun Sensor (FSS) is composed of a segmented photodiode or photodiode array. The attitude determination is based on the incident light angle on two axes. The angle accuracy can reach  or even 1 arcminute. Since the preprocessing involves complex computation, some FSS embedded with DSP or MCU for these calculations. Similar to CSS, the satellite requires multiple FSSs to widen its FOV of the FSS array. For more details, there is an excellent article about FSS written by Luke.

Figure 2. FSS100- Nano Fine Sun Sensor made by Tensor Tech

Star tracker

     Star tracker has the best attitude determination accuracy among all other attitude sensors. It majorly consists of CCD/CMOS camera and a powerful processor. The attitude determination based on the star catalog can provide accuracy close to 1 arcsecond [1]. However, they often consume more power and have smaller bandwidth compared to other non-image-based attitude sensors.

Figure 3. Star Tracker on Alphasat [2]

Geomagnetic sensor

     The Geomagnetic sensor achieves attitude determination with the help of GNSS information and the IGRF model. However, the readings of Geomagnetic sensors are vulnerable to magnetic disturbances, and the error in GNSS/IGRF information are also unneglectable.

Figure 4. Geomagnetic sensor from Honeywell [3]

     For better reliability, we often apply multiple kinds of attitude sensors on a satellite. The processor should apply post-processing and filtering to these data, such as the TRIAD method or Kalman filter. The final output is the estimated attitude of the satellite.

The Attitude Control

     The reaction wheel, Control Moment Gyro (CMG), thruster, and magnetorquer are the most popular actuators on the market. One of the purposes of these actuators is to provide torque to compensate the disturbance torques on-orbit for stabilization, such as gravity-gradient torque, radiation pressure torque, aerodynamic drag torque etc. The other is to apply yaw, roll, and pitch angle control for a satellite.

     Before choosing the actuator, the system engineer should specify the mass properties of the satellite, desired slew rate, and the pointing accuracy of the satellite. We wrote another blog about picking a suitable ADCS and how you can obtain mass properties information.

Reaction wheel (RW)

     The RW serves as an angular momentum exchange device on a satellite. Torque and maximum angular momentum are the major metrics to evaluate an RW. When maximum angular momentum/maximum speed is reached, the RW can no longer provide torque for the satellite in its negative direction to the angular velocity vector of the rotor, called “saturation.” At saturation, the RW reaches its maximum speed and required external torque for desaturation. The control algorithm should minimize the times of the occurrence of the saturation events. Despite the saturation issue, an RW can achieve arcsecond-level attitude control precision under proper configuration.

Figure 5. RW from Space Inventor [4]

Figure 6. RW cluster to achieve 3-axis control. This produt is made by NanoAvionics [4]

Magnetorquer

     The magnetorquer is a comparatively cheap attitude actuator that exerts external torque on the satellite. We sometimes just call them “torquer” and they are separated into two major categories, “Torquer Rod” and the “Air Coil.” Both operate similarly, but the former contains ferrite materials for intensifying its magnetic dipole moment.

     Since the torque source is based on Lorentz force, there is an uncontrollable axis parallel to the Geomagnetic flux density vector. Moreover, magnetorquers can only function in environments that contain enough local magnetic fields like Low earth orbit (LEO). The magnetorquers furtherly require a local magnetic flux density vector information for proceeding with its attitude control logic. A straightforward method for obtaining such data is using a magnetometer; propagating using the IGRF model is applicable, too. Torquer-based control is probably the simplest active attitude control methodology; however, its control precision is only around 1~5 degrees.

Figure 7. Magnetorquer from ISISPACE [5]

Single-Gimbal Control Moment Gyro (CMG)

     The CMG is an angular momentum exchange device consisting of a high-speed flywheel and a gimbal. Unlike RW, the CMG exerts gyroscopic torque via gimballing instead of accelerating or decelerating the wheel. This gyroscopic torque makes CMG generally more power-efficient than RW. However, this benefit is more significant on larger satellites than smallsats [6].

     CMGs usually appear with a combination of 4 to provide 3-axis controllability, called “clusters.” Although a single CMG doesn’t suffer from saturation problem, a cluster does, like all other angular-momentum-exchange-based attitude actuators. However, its singularity characteristics make its control algorithm much more complicated. The additional gimbal also makes the volume and mass larger than the RW. To sum up, the CMG can achieve similar control precision to the RW while having better torque to power consumption ratio but cost more.

Figure 8. CMG from AIRBUS [7]

Reaction Sphere (RS)

     The RS is a novel product from Tensor Tech, serving as a simple reaction wheel, dual-axis reaction wheel, or a single gimbal CMG. Being compactly designed, the RS helps simplify and minify the multi-axis actuator system. The RS provides flexibility to switch between the reaction wheel and CMG, such as exert gyroscopic torque when saturation. As the only existing CMG compatible product in Cubesat market, the RS is one of the best solutions for leveraging the efficiency of your ADCS.

Figure 9. RS100 from TensorTech [8]

     Since attitude actuators often occupied a large volume, mass, and power budget, we recommend considering the spec of ADCS at the earliest phase of mission planning. To help you get it through, Tensor Tech provides such free of charge consulting service for every potential customer. After all, it’s all about finding the “fit” into your mission requirements while balancing the budget.

     There are still some other kinds of attitude actuators, such as thrusters or fluid loop actuators [9]. However, we won’t make an intro for them here since they are not yet a popular offer.

Future of ADCS? An integrated solution

     The satellite market is booming majorly because of the participation of commercial players. These people calculate to balance between performance, reliability, lead time, and cost. That’s why the “integrated ADCS” solution becomes a charming option. The specification of integrated ADCS can satisfy most of the mission requirements while providing reliability and shorten up customer’s development time. Furthermore, users don’t have to worry about how to get those sensors and actuators to work. The users only need to sprcify desired attitude, the integrated ADCS will do the rest for you. If extra requirements were made, optional components are often feasible for extending the system’s capabilities.

Figure 10. ADCS100 from TensorTech [8]

References