Flight Control

(1) Multicopter Design and Control

The research on multicopter design and control has been summarized in book [1].

1) Multicopter Design

•  Based on Performance

Multicopters are becoming increasingly important both in civil and military fields. Currently, most multicopter propulsion systems are designed by experience and trial-and-error experiments, which are costly and ineffective. We have proposed a comprehensive offline evaluation algorithm for multicopter performance in [2], an analytical design optimization method for electric propulsion systems in [3], efficiency optimization and component selection in [4]. Based on the previous work, design automation and optimization methodology for electric multicopter is proposed in [5]. With these work, a website www.flyeval.com has been launched by our group since 2016.

•  Based on Controllability

It has long been known that control performance is an important property of a plant. However, few works consider the control performance when selecting the appropriate propulsion systems for multicopters. If the considered multicopter is not sufficiently controllable, then the propulsion system needs to be redesigned according to specific control requirements, which are a waste of time and money. The studies focus on selecting the propulsion system in [6] and configuration in [7] for multicopters by taking the controllability into consideration.

2) Multicopter Control

Attitude control of a multicopter subject to uncertainties and disturbances is studied [8],[9],[10],[11],[12], where some failures under control are also taken as disturbances. However, some failures will cause multicopters out of controllability. It is pointed out that classical controllability theories of linear systems are not sufficient to test the controllability of multicopters. Study [13] makes controllability analysis based on the theory of positive controllability [14]. Degraded control is further proposed for a class of hexacopters subject to rotor failures [15].

• [1] Quan Quan. Introduction to Multicopter Design and Control. Springer, Singapore, 2017.
• [2] Dongjie Shi, Xunhua Dai, Xiaowei Zhang, and Quan Quan. A practical performance evaluation method for electric multicopters. IEEE/ASME Transactions on Mechatronics. 2017, 22(3):13371348.
• [3] Xunhua Dai, Quan Quan, Jinrui Ren and Kai-Yuan Cai. An analytical design optimization method for electric propulsion systems of multicopter UAVs with desired hovering endurance. IEEE/ASME Transactions on Mechatronics, 2019, 24: 228-239.
• [4] Xunhua Dai, Quan Quan, Jinrui Ren and Kai-Yuan Cai. Efficiency optimization and component selection for propulsion systems of electric multicopters. IEEE Transactions on Industrial Electronics, 2019, 66(10): 7800–7809.
• [5] Xunhua Dai, Quan Quan, Kai-Yuan Cai. Design automation and optimization methodology for electric multicopter UAVs. Online: https://arxiv.org/abs/1908.06301
• [6] Guang-Xun Du and Quan Quan. Optimization of multicopter propulsion system based on degree of controllability. AIAA Journal of Aircraft, accepted, online: https://doi.org/10.2514/1.C035150
• [7] Binxian Yang, Guang-Xun Du, Quan Quan, Kai-Yuan Cai. The Degree of Controllability with Limited Input and an Application for Hexacopter Design. The 32nd Chinese control conference, Xi’an, 2013, 113-118. (in Chinese)
• [8] Ruifeng Zhang, Quan Quan, Kai-Yuan Cai. Attitude control of a quadrotor aircraft subject to a class of time-varying disturbances. IET Control Theory & Applications, 2011, 5(9): 1140-1146.
• [9] Quan Quan, Guang-Xun Du, Kai-Yuan Cai. Proportional-integral stabilizing control of a class of mimo systems subject to nonparametric uncertainties by additive-state-decomposition dynamic inversion design. IEEE/ASME Transactions on Mechatronics. 2016, 21(2):1092 – 1101.
• [10] Bai-Hui Du, Andrey Polyakov, Gang Zheng and Quan Quan. Quadrotor trajectory tracking by using fixed-time differentiator. International Journal of Control, 2019, 92(12): 2854–2868.
• [11] Guang-Xun Du, Quan Quan, Kai-Yuan Cai. Additive-state-decomposition-based dynamic inversion stabilized control of a hexacopter subject to unknown propeller damages. The 32nd Chinese control conference, Xi’an, 2013, 6231-6236.
• [12] Jing Zhang, Guang-Xun Du, Quan Quan. Initial research on vibration reduction for quadcopter attitude control: an additive-state-decomposition-based dynamic inversion method. 2017 Chinese Automation Congress (CAC), Oct. 20-22, 2017, Jinan, China.
• [13] Guang-Xun Du, Quan Quan, Binxian Yang and Kai-Yuan Cai. Controllability analysis for multirotor helicopter rotor degradation and failure. Journal of Guidance, Control, and Dynamics, 2015, 38(5): 978-985.doi: 10.2514/1.G000731
• [14]Guang-Xun Du, Quan Quan. Degree of Controllability and Its Application in Aircraft Flight Control. Journal of System Science and Mathematical Science, 2014 Vol. 34 (12): 1578-1594. (in Chinese)
• [15] Guang-Xun Du, Quan Quan, Kai-Yuan Cai. Controllability analysis and degraded control for a class of hexacopters subject to rotor failures. Journal of Intelligent & Robotic Systems, 2015, 78(1): 143-157.

(2) Probe and Drogue Autonomous Aerial Refueling

Probe-and-drogue refueling is widely adopted owing to its simple requirement of equipment and flexibility, but it has an apparent drawback that the drogue position is susceptible to disturbances. There are three types of disturbances: atmospheric turbulence, trailing vortex of the tanker, and bow wave effect caused by the receiver. The former two disturbances are independent of the receiver, whereas the bow wave effect, which depends on the state of the receiver, greatly influences the docking within a close distance [1].

To deal with this problem, we start the study from modeling, especially the bow wave effect on docking control [2],[3]. On the whole, there still exist some challenges to designing a docking controller for probe-and drogue refueling. First, the probe-and-drogue refueling system is complicated. Moreover, as one of the main disturbances in the docking stage, the bow wave effect is a repetitive nonlinear disturbance and highly related to the states of the receiver and the drogue, which also makes docking control difficult. Second, the problem of the “slow dynamics” to track the “fast dynamics” becomes together if the bow wave effect is taken into account. Some feedback control methods may result in a chasing process between the receiver and the drogue, which may cause over control. Besides, the chasing action may lead to impact and damage to the drogue and the probe, which is very dangerous and needs to be avoided according to ATP-56(B) issued by NATO (North Atlantic Treaty Organization). Third, influenced by the environment, some unexpected sensor delay may happen. We studied vision-based robust position estimation [4].

Based on the model obtained, iterative learning control is applied to make docking control reliable [5], [6], [7]. Here,“reliability”means a small docking error, certain robustness against disturbances and uncertainties, and little dependence on the system model and sensors.

Autonomous aerial refueling is vulnerable to various failures and involves co-operation among autonomous receivers, tankers and remote pilots. Dangerous flight maneuvers may be executed when unexpected failures or command conflicts happen. To solve this problem, a failsafe mechanism based on state tree structure is proposed [8]. The failsafe mechanism is a control logic that guides what subsequent actions the autonomous receiver should take, by observing real-time information of internal low-level subsystems such as guidance and drogue & probe and external instructions from tankers and pilots.

• [1] Quan Quan, Zi-Bo Wei, Jun Gao, Ruifeng Zhang, Kai-Yuan Cai. A survey on modeling and control problems for probe and drogue autonomous aerial refueling at docking stage. Acta Aeronautica ET Astronautica Sinica, 2014, 35(9): 2390-2410. (in Chinese)
• [2] Zi-Bo Wei, Xunhua Dai, Quan Quan, Kai-Yuan Cai. Drogue dynamic model under bow wave effect in probe and drogue aerial refueling. IEEE Transactions on Aerospace and Electronic Systems, 2016, 52(4): 1728-1742.
• [3] Xunhua Dai, Zi-Bo Wei, Quan Quan. Modeling and simulation of bow wave effect in probe and drogue aerial refueling. Chinese Journal of Aeronautics, 2016, 29(2): 448-461.
• [4] Yan Gao, Xunhua Dai, Quan Quan. Vision-based robust position estimation in probe-and-drogue autonomous aerial refueling. IEEE/CSAA Guidance, Navigation and Control Conference, Xiamen, 2018.
• [5] Xunhua Dai, Quan Quan, Jinrui Ren, Zhiyu Xi and Kai-Yuan Cai. Terminal iterative learning control for autonomous aerial refueling under aerodynamic disturbances. AIAA Journal of Guidance, Control, and Dynamics, 2018, 41(7):1576-1583.
• [6] Xunhua Dai, Quan Quan, Jinrui Ren, Kai-Yuan Cai. Iterative learning control and initial value estimation for probe–drogue autonomous aerial refueling of UAVs. Aerospace Science and Technology, 2018, 82: 583-593.
• [7] Jinrui Ren, Xunhua Dai, Quan Quan, Zi-Bo Wei and Kai-Yuan Cai. Reliable docking control scheme for probe–drogue refueling. AIAA Journal of Guidance, Control, and Dynamics, accepted, DOI: 10.2514/1.G003708
• [8] Ke Dong, Quan Quan and W. Murray Wonham. Failsafe mechanism design for autonomous aerial refueling using state tree structures. Unmanned Systems, 2019, 07(04): 261–279. Code: https://github.com/KevinDong0810/Failsafe-Design-for-AAR-using-STS

Control Theory and Methods

(1) Additive Decomposition Based Tracking

A commonly-used decomposition in the control field is to decompose a system into two or more lower-order subsystems, called lower-order subsystem decomposition here. In contrast, additive ( state ) decomposition [1] is to decompose a system into two or more subsystems with the same dimension as that of the original system. Taking a system $P$ for example, it is decomposed into two subsystems: $P_{p}$ and $P_{s}$ , where dim $P_{p}=n_{p}$ and dim $P_{s}=n_{s}$ ; respectively. The lower-order subsystem decomposition satisfies $$n=n_p+n_s \text { and } P=P_{p} \oplus P_{s}$$         By contrast, the additive state decomposition satisfies $$n=n_p=n_s \text { and } P=P_{p}+P_{s}$$

The additive decomposition is applicable not only to differential dynamic systems, but also to almost any dynamic or static system. The essential idea of the additive decomposition can be expressed in the following equation,

where $A$ is the original system, $B$ is the primary system and $C=A-B$ represents the secondary system. The symbols “$+$” and “$-$” in the equation are the plus and minus signs in the elementary arithmetic. It is obvious that the equality holds. The additive state decomposition is an extension of the superposition principle for linear systems. What is more, it can be applied to nonlinear systems. You can refer to Wiki for more information.

https://en.wikipedia.org/wiki/Additive_State_Decomposition for more information

The additive decomposition can be used for linearization, called compensation linearization. As shown in Figure 2, the principle behind is to use nonlinear System $B$ to compensate for the original nonlinear System $A$ resulting in a linear System $C$. With the compensation linearization, the idea of filter and frequency domain analysis can be apply to improve the robustness of the system.

The flowing applications of additive decomposition are for your reference.
•  Additive decomposition for iterative learning control [2],[3],[4],[5]
•  Additive decomposition for general trajectory tracking [6],[7],[8],[9],[10]
•  Additive decomposition for stabilizing control [11],[12]
•  Additive decomposition for system analysis[1],[13]
•  Furthermore, additive state decomposition is extended to additive output decomposition [14].

• [1] Quan, Quan and Kai-Yuan Cai. Additive decomposition and its applications to internal-model-based tracking. Joint 48th IEEE Conference on Decision and Control and 28th Chinese Control Conference, Shanghai, China, 2009. 817-822.
• [2] Zi-Bo Wei, Quan Quan, Kai-Yuan Cai. Output feedback ILC for a class of nonminimum phase nonlinear systems with input saturation: an additive-state-decomposition-based method. IEEE Transactions on Automatic Control. 2017, 62(1):502-508.
• [3] Quan Quan, Kai-Yuan Cai. Additive-state-decomposition-based tracking control for tora benchmark. Journal of Sound and Vibration, 2013, 332(20), 4829-4841
• [4]Quan Quan, Kai-Yuan Cai. Repetitive control for TORA benchmark: an additive-state-decomposition-based approach. International Journal of Automation and Computing, 2015, 12 (3): 289-296
• [5] Quan Quan, Lu Jiang, Kai-Yuan Cai. Discrete-time output-feedback robust repetitive control for a class of nonlinear systems by additive state decomposition. Online: http://arxiv.org/abs/1401.1577
• [6] Quan Quan, Hai Lin, Kai-Yuan Cai. Output feedback tracking control by additive state decomposition for a class of uncertain systems. International Journal of Systems Science, 2014, 45(9): 1799–1813.
• [7] Quan Quan, Kai-Yuan Cai, Hai Lin. Additive-state-decomposition-based tracking control framework for a class of nonminimum phase systems with measurable nonlinearities and unknown disturbances. International Journal of Robust and Nonlinear Control, 2015, 25(2):163-178.
• [8] Jinrui Ren, Zhiyu Xi, Li-Bing Zhao and Quan Quan. Transient tracking performance improvement for nonlinear nonminimum phase systems: an additive-state-decomposition-based control method. International Journal of Systems Science, 2017, 48(10): 2157–2167.
• [9] Jinrui Ren, Quan Quan, Li-Bing Zhao, Xunhua Dai and Kai-Yuan Cai. Two-degree-of-freedom attitude tracking control for bank-to-turn aerial vehicles: an additive-state-decomposition-based method. Aerospace Science and Technology, 2018, 77: 409-418.
• [10] Zi-Bo Wei, Jin-Rui Ren, Quan Quan. Further results on additive-state-decomposition-based output feedback tracking control for a class of uncertain nonminimum phase nonlinear systems Control and Decision Conference (CCDC), 2016 Chinese. IEEE, 2016: 6557-6562.
• [11] Quan Quan, Guang-Xun Du, Kai-Yuan Cai. Proportional-integral stabilizing control of a class of mimo systems subject to nonparametric uncertainties by additive-state-decomposition dynamic inversion design. IEEE/ASME Transactions on Mechatronics. 2016, 21(2) :1092-1101.
• [12] Jin-Rui Ren, Quan Quan. Initial research on stability margin of nonlinear systems under additive-state-decomposition-based control framework. Control and Decision Conference (CCDC), 2016 Chinese. IEEE, 2016: 5766-5771
• [13] Quan Quan, Kai-Yuan Cai. A filtered repetitive controller for a class of nonlinear systems. IEEE Transaction on Automatic Control, 2011, 56(2): 399-405.
• [14] Quan Quan, Kai-Yuan Cai. Additive-output-decomposition-based dynamic inversion tracking control for a class of uncertain linear time-invariant systems. The 51st IEEE Conference on Decision and Control, 2012, Maui, Hawaii,USA, 2866-2871.

(2) Filtered Repetitive Control

In practice, many control tasks are often considered to exhibit periodic behavior. High-precision control performance can be realized for such periodic control tasks using repetitive control (RC)[1]. A linear RC system is a neutral type system in a critical case [2]. The characteristic equation of a neutral type system includes an infinite sequence of roots with negative real parts approaching zero; i.e. sup$\{Re(s)|F(s)=0\}=0$, where $F(s)$ is the characteristic equation. This implies that a sufficiently small uncertainty may lead to sup$\{Re(s)|F(s)= 0\}>0$, namely instability. For a nonlinear RC system, [3] shows that it will lose its stability when subject to a small input delay. Therefore, it is important to design robust RC to achieve a tradeoff between precision and robustness. Filtered RC is an improved RC, where a filter is ntroduced into the internal model of RC to reshape the internal model for a tradeoff between precision and robustness. Filtered RC has been widely-used in linear-invariant systems. But, recent developments concerning RC have not been consistent, with limited research on RC in nonlinear systems [1].

About filtered RC, filtered RC for a class of time-delay systems is studied in [6], filtered RC for a class of nonlinear systems is also proposed by following the idea of adaptive control [3],[4],[5]. However, for nonlinear systems, feedback linearization or error dynamics derived required by adaptive control is often difficult to perform. To remedy this, three new methods parallel to the two methods mentioned above are proposed. They are

•  Additive-state-decomposition based method [7],[8],will bridge the linear time-invariant systems and nonlinear systems so that the linear RC methods can be used in nonlinear systems.
•  Actuator-focused design method stems from another viewpoint of the internal model principle proposed by us [9].
•  Contraction mapping method solves the RC problem for nonlinear systems without requiring the corresponding Lyapunov functions[10]

Because the filtered RC system is a time-delay system, we also have studied the ultimate bound and convergence rate for perturbed time-delay systems [11].

• [1] Quan Quan, Kai-Yuan Cai. A survey of repetitive control for nonlinear systems. Science Foundation in China, 2010, 18(2): 45-53.
• [2] Quan Quan, Dedong Yang, Kai-Yuan Cai. Linear matrix inequality approach for stability analysis of linear neutral systems in a critical case. IET Control Theory & Applications, 2010, 4(7):1290-1297.
• [3] Quan Quan, Kai-Yuan Cai. A filtered repetitive controller for a class of nonlinear systems. IEEE Transaction on Automatic Control, 2011, 56(2): 399-405.
• [4] Quan Quan, Kai-Yuan Cai. Filtered repetitive control of robot manipulators. International Journal of Innovative Computing, Information and Control, 2011, Vol. 7, No. 5(A): 2405-2415.
• [5] Quan Quan, Kai-Yuan Cai. Internal-model-based control to reject an external signal generated by a class of infinite-dimensional systems. International Journal of Adaptive Control And Signal Processing, 2013, 27(5): 400-412.
• [6] Quan Quan, Dedong Yang, Kai-Yuan Cai and Jiang Jun. Repetitive control by output error for a class of uncertain time-delay systems. IET Control Theory & Applications, 2009, 3(9): 1283-1292.
• [7]Quan Quan, Kai-Yuan Cai. Repetitive control for TORA benchmark: an additive-statedecomposition-based approach. International Journal of Automation and Computing, 2015, 12 (3): 289-296.
• [8] Quan Quan, Lu Jiang, Kai-Yuan Cai. Discrete-time output-feedback robust repetitive control for a class of nonlinear systems by additive state decomposition. Online: http://arxiv.org/abs/1401.1577.
• [9] Quan Quan, Kai-Yuan Cai. Repetitive control for nonlinear systems: an actuator-focussed design method. International Journal of Control, online: https://doi.org/10.1080/00207179.2019.1639077.
• [10] Quan Quan, Kai-Yuan Cai. Saturated repetitive control for a class of nonlinear systems: A contraction mapping method. Systems & Control Letters, 2018, 122: 93-100.
• [11] Quan Quan, Kai-Yuan Cai. A new method to obtain ultimate bounds and convergence rates for perturbed time-delay systems. International Journal of Robust and Nonlinear Control, 2012, 22(16):1873-1880.

(3) Stability of Neutral Type Systems in a Critical Case

A linear repetitive control system is a neutral type system in a critical case. The principal neutral term usually plays an important role in the stability analysis of a neutral system. For clarity, we first introduce a class of linear neutral systems $$\begin{equation} \dot{x}\left(t\right)-H\dot{x}\left(t-\tau\right)=F\left(x_{t}\right) \label{neutral system} \end{equation}$$ where $\tau > 0$ is a constant delay, $F\left(\cdot\right)$ is a linear functional and $x_{t}\triangleq x\left(t+\theta\right)$, $\theta\in\left[-\tau,0\right].$ According to the spectral radius of matrix $H$, the neutral system (\ref{neutral system}) can be classified into three cases: $\rho\left(H\right) < 1$, $\rho\left(H\right) > 1$ and $\rho\left(H\right)=1$. The case $\rho\left(H\right) < 1$, namely matrix $H$ is Schur stable, is a necessary condition for exponential stability of the linear neutral system (\ref{neutral system}). To the best knowledge of us, the case $\rho\left( H\right) > 1$ usually leads the linear neutral system (\ref{neutral system}) to instability, for there exist characteristic roots ith positive real parts. The last case $\rho\left( H\right) =1$ is the critical case. The characteristic equation of a neutral type system in the critical case includes an infinite sequence of roots with negative real parts approaching zero.

We have made stability analysis on this type of systems obtaining some criteria [1]. Further results on it are robustness criteria [2] and a criteria with a relaxed condition on the system [3]. Also, we proposed a new model transformation method to relax stability criteria [4]. A framework was proposed further to extend a class of existing stability criteria to determine the stability of a class of neutral type systems in a critical case [5]. For general nonlinear systems, a stability theorem of the direct Lyapunov's method for neutral type systems in a critical case was proposed based on the previous results. The proposed criteria can help to determine the stability of the case where H has multiple eigenvalues of modulus 1 without Jordan chains. This gives an alternative to handle the ‘open problem’ according to J. Differ. Equ., 2005, 214, (2), pp. 391– 428.

• [1] Quan Quan, Dedong Yang, Kai-Yuan Cai. Linear matrix inequality approach for stability analysis of linear neutral systems in a critical case. IET Control Theory & Applications, 2010, 4(7):1290-1297.
• [2] Quan Quan, Kai-Yuan Cai. Robustness analysis of a class of linear neutral systems in a critical case. IET Control Theory & Applications, 2010, 4(9):1807-1816.
• [3] Quan Quan, Kai-Yuan Cai. Stability analysis of a class of neutral type systems in a critical case without restriction on the principal neutral. Asia Journal of Control, 2013, 15(6), 1858-1861.
• [4] Quan Quan, Dedong Yang, Hai Hu, Kai-Yuan Cai. A new model transformation method and its application to extending a class of stability criteria of neutral type systems. Nonlinear Analysis: Real World Applications, 2010, 11(5): 3752-3762.
• [5] Quan Quan, Kai-Yuan Cai. A framework for the stability of a class of neutral type systems in a critical case. Proceedings of the 29th Chinese Control Conference, Beijing, 2010, 90- 95.
• [6] Quan Quan, Kai-Yuan Cai. A stability theorem of the direct Lyapunov's method for neutral type systems in a critical case. International Journal of Systems Science, 2012,43(4):641-646.

Evaluation Theory and Methods

(1) Online Health Evaluation

Online health evaluation can predict a system’s health status and avoid dangerous accidents in advance. Health evaluation can be used for condition-based maintenance to ensure reliability and extend the whole life cycle of a system. According to the fault and understanding degree of flight control software, four models, namely fault tree model, stochastic hybrid dynamic system model, controllability model and data model, have been proposed to evaluate a system’s health.

•  Stochastic hybrid dynamic system model[1],[2]

Health performance prediction of a stochastic hybrid dynamic system aims at determining the probability or possibility that the system state will remain in a permitted area (safe set) or reach a forbidden area (unsafe set) at a future time instance.

•  Fault tree model [3],[4]

Profust reliability theory extends the traditional binary state space into a fuzzy state space, which is therefore suitable to characterize a gradual physical degradation. Based on the fault tree model a profust reliability based health evaluation approach is proposed, where the profust reliability is employed as a health indicator to evaluate the real time system performance.

•  Controllability model [5],[6],[7]

By taking the failure as a disturbance, the degree of controllability of the perturbation is taken as a health indicator.

•  Data model [8]

A normal and abnormal system’s vibration signals are used to train an artificial neural network. Then, new data will feed into the trained artificial neural network to get a health indicator.

• [1] Zhiyao Zhao, Quan Quan, Kai-Yuan Cai. A health performance prediction method of large-scale stochastic linear hybrid systems with small failure probability. Reliability Engineering and System Safety, 2017, 165: 74–88.
• [2] Zhiyao Zhao, Quan Quan, Kai-Yuan Cai. A health evaluation method of multicopters modeled by stochastic hybrid system. Aerospace Science and Technology 2017, 68: 149–162.
• [3] Zhiyao Zhao, Quan Quan, Kai-Yuan Cai. A profust reliability based approach to prognostics and health management. IEEE Transaction on Reliability, 2014, 63(1), 26-41.
• [4] Zhiyao Zhao, Quan Quan, Kai-Yuan Cai. A modified profust-performance-reliability algorithm and its application to dynamic systems. Journal of Intelligent and Fuzzy Systems 2017, 32(1): 643660.
• [5]Guang-Xun Du, Quan Quan, Binxian Yang and Kai-Yuan Cai. Controllability analysis for multirotor helicopter rotor degradation and failure. Journal of Guidance, Control, and Dynamics, 2015, 38(5): 978-985.doi: 10.2514/1.G000731
• [6]Guang-Xun Du, Quan Quan, Zhiyu Xi, Yang Liu, Kai-Yuan Cai. A Control Performance Index for Multicopters Under Off-nominal Conditions. Online: https://arxiv.org/abs/1705.08775, [Video]
• [7] Guang-Xun Du, Quan Quan. Degree of Controllability and Its Application in Aircraft Flight Control. Journal of System Science and Mathematical Science, 2014 Vol. 34 (12): 1578-1594. (in Chinese)
• [8] Jiang Yan, Zhiyao Zhao, Haoxiang Liu, Quan Quan. Fault detection and identification for quadrotor based on airframe vibration signals: a data-driven method. Proceedings of the 34th Chinese Control Conference, July 28-30, 2015, Hangzhou, China

(2) Offline Performance Evaluation

•  Basic Design Performance

In the design phase, designers and users wonder if an assembled multicopter can meet their performance requirements, such as hovering endurance, system efficiency, maximum load, maximum pitch, and maximum flight distance. However, in practice, they used to evaluate the performance of a multicopter through lots of flight experiments or by experience, which are normally inefficient and costly. We have proposed a comprehensive offline evaluation algorithm of ulticopter performance and launched a website www.flyeval.com. Reliability evaluation of different multicopter configurations is also proposed in [2].

•  Degree of Handling Performance

A multicopter is a kind of aircraft with simple structure and has the outstanding agility and handling performance. These made it wildly applicable in many fields and developed rapidly. In recent years, lots of functional multicopter products have been made. For different multicopter products, their flight performance and the difficulty degree of handling, which are called handling qualities, are different from each other. However, as far as we know, there is no existing methods and unified standards to assess multicopters so far. The handling quality assessment standards can offer a ‘ruler’ to compare for choosing multicopters, and it also can offer references to multicopter designers.

• [1] Dongjie Shi, Xunhua Dai, Xiaowei Zhang, and Quan Quan. A practical performance evaluation method for electric multicopters. IEEE/ASME Transactions on Mechatronics. 2017, 22(3):13371348.
• [2] Dongjie Shi,Binxian Yang, Quan Quan. Reliability analysis of multicopter configurations based on controllability theory. Control Conference (CCC), 2016 35th Chinese. IEEE, 2016: 6740-6745.

Model-based Rapid Design Platform

Model-Based Design (MBD) is an efficient and cost-effective way to develop complex embedded systems in aerospace, automotive, communications, and other industries. With MBD methods, we have developed a multicopter experimental platform for the rapid control algorithm development based on the Pixhawk autopilot and MATLAB/Simulink, with a series of experiment courses. In this platform, users can use automatic design, code generation and hardware-in-the-loop simulation methods to improve the development speed and the safety and reliability of UAV systems.

YouTube   Youku Polyv

Automatic Safety Testing Platform for UAV Systems

YouTube   Youku Polyv

Automatic Safety Testing Platform comprised of the master computer, real-time simulation computer, and vehicle body to test. With FPGA simulating real sensors, the platform achieves real-time hardware-in-the-loop simulation without other expensive hardware. The testing platform includes reliability testing and safety testing. The algorithms of the vehicle, such as multicopter, fixed-wing, VTOL, helicopter, ship, and car, can be validated and verified by the HIL test on the platform.

Multicopter Evaluation Website

flyeval.com is a performance evaluation website for multicopters. Users can obtain the detail evaluations after providing the body frame, environment and propulsion system parameters. flyeval.com/recalc.html is an online toolbox for multicopter design optimization. By simply inputting the design requirements, the toolbox outputs the optimal multicopter design including the size, weight, payload capability, and propulsion component selection.

Aerial Refueling Platform

A MATLAB/SIMULINK based simulation environment has been developed to simulate the capture stage of autonomous aerial refueling procedure. The hose-drogue dynamic model used in this simulation is a 20-links-connected model. The tanker is a Boeing-707, which is assumed to fly straight and level with constant speed and direction. The receiver is an F-16 nonlinear model, which is a high fidelity model that can simulate the response of an actual F-16 using the high-precision aircraft data. In this platform, the modeling, vision-based navigation, safety analysis, control method and control strategy for aerial refueling can be studied..

YouTube   Youku Polyv

Visual Sensor Network

Visual Sensor Network (VSN) is a sensor-based networks with spatially distributed smart cameras capable of processing image data locally and fusing images of a scene from different viewpoints. It has emerged as an important class of multi-camera distributed intelligent systems, with unique performance, complexity and quality of service challenges. Current optical motion capture systems are a successful example of how useful the multi-camera system can be. We have designed a comprehensive multi-camera-based testbed for 3D tracking and control of multiple unmanned aerial vehicles (UAVs). In the testbed, some reflective markers can be detected by smart cameras, and their positions are easily reconstructed by triangulation algorithms and Kalman filters. The proposed testbed is a comprehensive and complete platform with good scalability applicable for research on a variety of advanced guidance, navigation, and control algorithms.

Research Directions:

➤   Design and implementation of a wireless visual sensor network for cooperative navigation and localization of UAVs.
➤   Accurate and flexible multi-camera calibration method of a visual sensor network.
➤   Optimal camera arrangement of a visual sensor network to obtain maximum coverage.
➤   Research on the calibration and localization problems caused by asynchronous cameras.
➤   Design and implementation of a simulation environment of a visual sensor network.

Multicopter Swarm Ground-Based Control Platform

YouTube   Youku Polyv

Multicopter Swarm Ground-Based Control Platform is proposed to implement multicopter swarm tests based on a central control computer combined with navigation provided by external optical positioning systems. It can provide a safe and effective environment for researchers to verify and test advanced UAV swarm algorithms. We have employed a commercial small-size quadcopter DJI Tello. In the platform, the navigation data is obtained by the external optical position systems (i.e., the commercial OptiTrack or our designed multi-camera testbed), and then transferred to the Simulink model on a central control computer via User Datagram Protocol (UDP). Then, the main swarm control rate in the Simulink model generates the command controls to the Tello swarm.

Research Directions:

➤   Design and implementation of accurate, fast, and robust swarm control rates.
➤   Model-based design of any specific multicopter swarm control.
➤   Swarm control of other unmanned vehicles, such as mobile robots.

① Multicopter navigation acquisition
② Multicopter trajectory generation
③ Multicopter swarm control rate
④ Control commands sending
⑤ Navigation data plotting
⑥ Trajectory plotting
⑦ Data logging
⑧ Battery status
⑨ Real-time control module