Multicopters Step by Step：How to Develop More Full-Stack Engineers based on Multicopters
Nowadays, the micro-small UAV industries have entered a new, orderly, and more opportunities development phase. With the rise of 5G, micro-small UAVs will become big data collection platforms at low altitudes, which is expected to bring seven to ten times more business opportunities to the industries. Meanwhile, electric and hybrid-electric powered vertical takeoff and landing (eVTOL) aircraft as “flying cars” are also promising and in progress.
The rapid development of UAVs or eVTOL aircraft in the industry is inseparable from the support of open-source flight control systems, like APM and PX4. Whether based on open-source flight control systems or fully independent development, developers need to have a comprehensive grasp in multicopter design, modeling, perception, control, and decision-making, to make them with outstanding performance, especially safety. Furthermore, developers must have practical experience in embedded hardware, operating-system development, software debugging, and flight tests. However, it is difficult to find engineers knowing all of them, namely chief and full-stack engineers. For more engineers interested in this, the big question is how to get started and how to find the right balance between theoretical depth and practical pragmatism. One important reason is that the development of UAVs or eVTOL is too fast, but the corresponding textbooks and curriculum lacks. The other reason is that the autopilot is too complex to lack new tools and new tutorials to guide developers through the jungle of embedded hardware and software.
To solve the two concerns above, we start with multicopters, which are simpler than fixed-wing aircraft, helicopters, and their hybrid. Although the house sparrow is small, but be fullies equipped . Multicopter systems are integrated systems that require interdisciplinary knowledge, which makes the perfect research object and touchstone for engineering education and practice.
Multicopters can be controlled autonomously by onboard computers or remotely through wireless communication (involving knowledge in Information and Communication Engineering) by the ground control station or the radio control system.Multicopter systems are composed of many electronic components (involving knowledge in Electronics Science and Technology).Multicopter systems require operating systems to run flight control algorithms (involving knowledge in Computer Science).For multicopter designs, the selection of materials, configuration, and structure (involving knowledge in Mechanics and Mechanical Engineering) must be considered; additionally, the propulsion system (involving knowledge in Mechanics and Electrical Engineering) must be taken into account.For the state estimation of multicopters, it is necessary to consider the problems of signal non-synchronization, sampling time difference, data delay, and sensor failures (e.g., GPS, gyroscope, accelerometer, magnetometer, barometer, ultrasonic rangefinder, and photoelectric sensor), which make the state estimation of the multicopter robust and high-performance (involving knowledge in Instruments Science and Technology).The multicopter system is a typical closed-loop control system(involving knowledge in Control Science and Engineering) with many interesting features that are unstable, nonlinear, underactuated and control-direction-limited (the propeller can only generate positive thrust perpendicular to the plane of the fuselage).Multicopters are also cheap and easy to pilot, thereby making it easy to obtain flight data, which provides a basis for the health assessment of multicopters.
Motivated by these above, we first published the book “Introduction to Multicopter Design and Control" by Springer in 2017. This book gives a complete picture of multicopter systems rather than a single method or technique, covering design, modeling, perception, control, and decision of multicopters. Furthermore, we published its sister book " Multicopter Design and Control Practice" by Springer in 2020. The new book offers new tools, RflySim, and new tutorials, so that readers can apply the theory to practice (see rflysim.com).
The new tools adopt:
The most widely-used flight platform - multicopters, as a flight platform;The most widely-used autopilot hardware - Pixhawk, as a control platform;One of the most widely-used programming languages in the field of control engineering - MATLAB + Simulink, as a programming language.
What is more, based on the current advanced development concept "Model-Based Design" process, the above three are closely linked. In addition to the advancement of software, hardware, and development concepts, the tutorials consist of eight tasks corresponding to the first book “Introduction to Multicopter Design and Control." Each task consists of three step-by-step experiments from shallow to deep so that developers with different backgrounds can benefit from them. Each experiment needs to be implemented in MATLAB + Simulink, and the simulation test is carried out in a built simulation platform. After that, the developers need to upload the controller to the Pixhawk autopilot through the automatic code generation technology and form a closed loop with a given real-time simulator for hardware-in-the-loop simulation test. The design task also includes outdoor flight experiments, so that developers would experience the full development of a multicopter. Throughout the process, the developers would be familiar with the basic process of the model-based design, also including the composition, mathematical model, and control of a multicopter. Advanced development tools and processes allow one engineer or a small team to realize new ideas rapidly. Besides, by lowering the learning threshold, more people with less early-stage preparations have opportunities to enter the aviation and robotics field.
The book
About the author
Quan Quan received his B.S. and Ph.D. degrees from Beihang University (Beijing University of Aeronautics and Astronautics), China in 2004 and 2010, respectively. He has been an associate professor at Beihang University since 2013. He is also the director of reliable flight control group at Beihang University. His main research interests include reliable flight control and health managment. To date he has published more than 50 peer-reviewed journal papers and three books, where “Introduction to Multicopter Design and Control” is the first introductory book on multicopters. In addition, he launched Rflysim platfrom (rflysim.com) for unmanned systems.
Line