The flight sensing module of unmanned aerial vehicles has two main purposes. One is to provide flight control systems. Due to the main function of the flight control system being to control the drone to achieve the required attitude and spatial position, this sensing technology mainly measures physical quantities related to the state. The modules involved include gyroscopes, accelerometers, magnetic compasses, barometers, global navigation satellite system modules, and optical flow modules.

Another purpose is to provide autonomous navigation systems for unmanned aerial vehicles, namely path and obstacle avoidance planning systems. Therefore, it needs to perceive the state of the surrounding environment, such as the location of obstacles. The relevant modules include ranging module, object detection and tracking module, etc...

Body motion sensing module
gyroscope
At present, gyroscopes based on MEMS technology are widely used in commercial drones due to their small size and low cost. They can be packaged in the form of integrated circuits. MEMS gyroscopes are used to measure the angular rate at which an object rotates around its own axis. The commonly used models are 6050A (Invernse) and ADXRS290 (ADI). The indicators for measuring gyroscope performance include measurement range, sensitivity, stability (drift), and signal-to-noise ratio.
The test environment was heated from 25 ° C to 50 ° C. The gyroscope remains stationary throughout the entire process, and the precise output of the gyroscope should be a fixed value. But from the results, it can be seen that the actual output of these two sensors is affected by temperature changes. In contrast, the output value of ADXRS290 (ADI) has a small range of variation, basically around 0.5.

Accelerometer
Accelerometers measure the linear acceleration of human motion. Due to the gravity of the Earth, the measured value will also include the gravitational acceleration component, which needs to be subtracted in certain usage cases. The commonly used MEMS accelerometer sensor models include 6050A (Inversese) and ADXL350 (ADI). In order to improve chip integration, some sensor manufacturers package gyroscopes and accelerometers together, known as six axis sensors, such as 6050A (Inversense).

Magnetic compass
The physical quantity measured by a magnetic compass is the component of the Earth's magnetic field intensity along the axis of the body, based on which the heading angle of the computer body is calculated. The commonly used MEMS magnetic compass sensor models are HMC5983L (Honeywell) and QMC5883L (silicon core), with similar performance. The former has already ceased production. The main performance parameters of a magnetic compass include sensitivity, stability (drift), etc.

baromete
The physical quantity measured by a barometer is the atmospheric pressure value, from which absolute height can be calculated. Commonly used barometer sensor models include MS5611 (MEAS), MS5607 (MEAS), and BMP180 (Bosch) [View detailed information on this product]. The problem with the barometer is that when flying near the ground, the "ground effect" will cause the air pressure distribution around the aircraft to be different from the static atmosphere, so that the barometer cannot be used to measure altitude. The usual solution is to use other sensors, such as ultrasonic sensors or laser rangefinder, during takeoff or landing.

Global Navigation Satellite System Module
The GNSS module measures a relatively rich range of physical quantities, mainly including geographic coordinates (latitude and longitude), altitude, line velocity, and heading angle (RTK system). Commonly used GNSS module manufacturers include U-BLOX from Switzerland and NOVATEL from Canada. When using the global navigation satellite system module, the placement of the satellite signal receiving antenna needs to avoid electromagnetic interference shielding. Some powerful aircraft manufacturers customize satellite signal receiving antennas based on aircraft models.

Optical flow module
The optical flow module is a special type of module that can be used to perceive the motion status of objects, such as measuring displacement velocity in the horizontal direction, and also to perceive the surrounding environment to achieve obstacle avoidance. A more common optical flow module is the open source PX4FLOW. The optical flow module is usually used indoors to solve the problem of poor indoor satellite signal. In addition, the ground to be photographed requires a certain texture pattern.
Surrounding environment sensing module
Ranging module
The following are five commonly used ranging modules: ultrasonic, infrared TOF, LiDAR, millimeter wave radar sensors, and depth sensing cameras.
Ultrasonic measurement, infrared ToF laser measurement, depth sensing camera, millimeter wave radar
Measurement distance short short long medium long
High, high, and low measurement accuracy
Weak anti-interference
Module size small small large medium large
Ultrasonic and infrared TOF have similar performance in all aspects. For example, the measurement distance is relatively close. The distance measured by ultrasound is usually about 4 meters. In addition, the range of use of these two sensors is easily limited by the actual environment. For example, infrared TOF emits red light onto the surface of the measured object and reflects it. If you encounter an object with low red light reflectivity like glass, it will fail. But one of the biggest advantages of these two types of sensors is their low cost and relatively small module size, making them widely used in consumer drones.

The measurement distance of LiDAR ranging is far enough. Most products can reach over 100 meters, but heavy rain and fog can affect their measurement results. Another drawback is the relatively high cost: Velodyne is the strongest in the LiDAR industry. Its VLP-16 is a miniaturized product suitable for drones, and its price also exceeds $1000. The cost is still relatively high.

Depth sensing cameras can be divided into three types based on measurement technology. ① Stereo cameras are also known as binocular vision technology. The representative product of this technology is Wizard 4 from DJI Structured light technology. It represents Microsoft's Kinect Time of Flight (TOF) technology. Due to fewer manufacturers and higher costs, the application of drones is rare.

Depth sensing cameras also have limitations when used. The disadvantage of binocular vision technology is that it cannot work normally in weak light environment, while structured light technology, on the contrary, cannot work normally in strong light. Therefore, some manufacturers combine these two technologies to compensate for each other's shortcomings and expand their range of applicable environments.

Methods for improving measurement accuracy
Sensor calibration
Sensor calibration, including fine calibration and rough calibration. Fine calibration is better, but requires expensive calibration equipment; Rough calibration does not rely on external devices, only the sensor itself can be operated.
Taking the rough calibration of a magnetic compass as an example, since the intensity of the geomagnetic field at any position on Earth can be considered constant for a long period of time when the magnetic compass rotates, it can be assumed that the magnetic compass is fixed based on relative motion. The geomagnetic field vector rotates accordingly, and the trajectory of the vector endpoint in space should be a standard sphere. However, due to sensor errors, the actual measured data is not strictly on the surface of the sphere. At this point, it needs to be based on measurement data. Numerical values and known precise values are used to calculate the conversion relationship between the two, namely the error model of a magnetic compass. In the future, when using this magnetic compass, the measured values can be processed based on the error model obtained from rough calibration, thereby reducing the error of the measured values.

Multi sensor data fusion
There are many different types of sensor data fusion methods. The most commonly used method in the industry is EKF, which is an extended Kalman filter.
Taking the fusion method for calculating aircraft attitude angles as an example, the EKF update process is mainly divided into two parts: prediction update and measurement update. Predictive updates mainly use gyroscopes to update the predicted state variables, while calculating the covariance matrix of the state variables. In measurement updates, the filter gain is first calculated, and then the predicted state variables, accelerometer, and magnetic compass data are fused into the fused state variables using the filter gain. At the same time, calculate the covariance matrix of the fusion state variables, which will be calculated in the next update cycle.

Sensor redundancy design
Sensor redundancy design mainly involves combining multiple sensors of the same type together. The processing method is to first remove sensors with abnormal data, and then perform sensor fusion. Redundant design can not only improve measurement accuracy, but also enhance the reliability of the entire system. When a sensor malfunctions, the entire system can continue to function normally.