Active Disturbance Rejection Control of Robots
The research paper published in IEEE focuses on the importance, methodology, and underlying mathematics of an active disturbance rejection controller. The use of active disturbance rejection control in industrial robots and mini robots is promising and has proven to provide great precision and efficiency as compared to other complex control algorithms.
Control of robots has always been a center of attention for scientists and researchers. From classical feedback controls to advanced control algorithms, the control theory has evolved. The control of different robots has sparked the fascination of researchers in the development of computationally inexpensive and efficient control algorithms. These algorithms, sometimes, can even help solve problems that have to be solved by altering the design of robots. In control systems, feedback is used to adjust the system’s dynamics (typically to make the response stable and rapid) and to lower the system’s sensitivity to signal uncertainty (disturbances) and model uncertainty. The simplest type of classical feedback control is PID.
The PID control, being simple and effective, has been used in industrial robotics, the aerospace industry, and even process control; however, the rising requirements for precision and efficiency have driven engineers to seek alternative modern control algorithms. The advanced control methodologies, however, contain complex mathematical algorithms and have a much higher computational cost as compared to PID. But active disturbance rejection control is quite simple compared to advance control algorithms and might provide better precision and efficiency than many advance control schemes.
What is an Active Disturbance Rejection Control?
Active disturbance rejection control (ADRC), first proposed by Han in 1995, is a model-independent, robust, and disturbance observer-based non-linear control scheme. Its main idea is to estimate and reject disturbances and make the robot follow the desired trajectory. It consists of a non-linear extended state observer (NLESO), a transient profile generator (TPG), and a non-linear state error feedback system (NLSEF).
A NLESO simplifies the plant description and groups all the uncertainties, disturbances (foreign and internal), and unmodeled dynamics as an equivalent disturbance term. Now, with the equivalent disturbance term and the states being estimated by NLESO, the NLSEF drives the error to zero, keeping the robot on a desired path despite all foreign and internal disturbances. The TPG helps eliminate overshoot-related issues and smooths aggressive input, such as step input. There are a few tuning parameters involved in the ADRC scheme, which can be tuned intuitively.
Active Disturbance Rejection Control Applications in Mobile Robots
Active disturbance rejection control, being efficient and computationally inexpensive, has wide applications in the field of mobile robots. The quadcopters, hexacopters, hexapods, nanorobots, and industrial robots, etc., have coupled and non-linear dynamics. In order to control the above-mentioned robots, feedback control is required. Since, the inertial and physical parameters of robots are not precisely known and their environment disturbances are unforeseen and uncertain, there is a need for a robust control algorithm that can handle the underlying situation. These uncertainties include the variation in mass, inertia, and gravity, and the disturbances can include wind gusts and the sudden grasping of some items as robots usually do. Since ADRC incorporates the use of an extended state-observer, the disturbances and the uncertainties can be actively estimated and rejected.
In order to achieve accurate navigation of mobile robots, the lateral control of a mobile robot has to be robust, which means it should adapt well to the uncertainties associated with its lateral motions, such as longitudinal velocity varieties and disturbances. The active disturbance rejection control proves to be of great advantage here.
For aerial systems such as quadrotors and mini drones, the disturbances are significant and the environment is uncertain. Usually for aerial systems, the adaptive control strategy, model predictive control and gain scheduling control is used. These control algorithms are quite difficult to tune and are computationally quite expensive. But, on the contrast, ADRC makes these types of systems easy to control since, the accurate mathematical modeling, and tuning of the gains over the wide range of trajectory is not needed. The ADRC will simply estimate these disturbances when applied and reject it immediately using a non-linear state error feedback law.
Industrial robots have many problems associated with them. These robots must operate in quite rough and tough environment. They are usually designed to work in areas of varying environments where the pressure is sometimes too high and sometimes too low like vacuum chambers, in extremely dirty areas and in areas where there is a threat of explosion, radiations and extreme hazards to humans. Since these robots must be very precise, efficient and accurate, the control algorithm must be very robust. The controller will have to operate the robot in the forementioned environments, and it should be designed in such a way that all these extreme conditions can be well handled. ADRC has been used many industrial robots and has provided excellent results.
Limitations
Control by observer-based feedback will inevitably lead to transient energy growth in response to disturbances for the closed-loop system if the uncontrolled system exhibits transient energy growth in response to optimal disturbances. This shows that observer-based feedback is a poor candidate for controller synthesis in the setting of transient energy growth suppression and transition delay. Hence, observer-based feedback can never satisfy the performance goal of transient energy growth suppression and transition delay.
References
CHENG, X., TU, X., ZHOU, Y. & ZHOU, R. 2019. Active Disturbance Rejection Control of Multi-Joint Industrial Robots Based on Dynamic Feedforward. Electronics, 8. Available at: https://doi.org/10.3390/electronics8050591
DING, L., HE, Q., WANG, C., QI, R. & JIAQIANG, E. 2021. Disturbance Rejection Attitude Control for a Quadrotor: Theory and Experiment. International Journal of Aerospace Engineering, 2021, 1–15. Available at:
https://doi.org/10.1155/2021/8850071
HAN, J. 2009. From PID to Active Disturbance Rejection Control. IEEE Transactions on Industrial Electronics, 56, 900–906. Available at:
https://doi.org/10.1109/TIE.2008.2011621
HEMATI, M. S. & YAO, H. 2018. Performance Limitations of Observer-Based Feedback for Transient Energy Growth Suppression. AIAA Journal, 56, 2119–2123. Available at:
https://doi.org/10.2514/1.J056877
KANG, C., WANG, S., REN, W., LU, Y. & WANG, B. 2019. Optimization Design and Application of Active Disturbance Rejection Controller Based on Intelligent Algorithm. IEEE Access, 7, 59862–59870. Available at: