Partial Feedback Linearization (PFL)
Partial Feedback Linearization (PFL) [1-3] is a classical method from control theory. With PFL it is possible to provoke a linear response in both joints of the double pendulum even if operated as a pendubot or acrobot. For an intuition of its functionality consider the manipulator equation for the acrobot (\(u_1 \equiv 0\))
The unactuated upper part of the vector equation can be solved for the acceleration \(\ddot{q}_1\) and then plugged into the lower part of the equation. The control input \(u_2\) can now be designed as PD control with an energy term
with the desired configuration \(q_2^{d}\) of the second link, the total energy \(E\), the desired total energy \(E^{d}\) and the gain parameters \(k_p, k_d\) and \(k_e\). The above described method is called collocated PFL. Similarly, it is also possible to eliminate \(\ddot{q}_2\) instead of \(\ddot{q}_1\) from the equations which is than called non-collocated PFL. Partial feedback linearization for the pendubot can be done on the same way. The collocated control law in this case reads
References
[1] M. W. Spong, “Swing up control of the acrobot using partial feedback linearization” IFAC Proceedings Volumes, vol. 27, no. 14, pp. 833–838, Sep. 1994, doi: 10.1016/S1474-6670(17)47404-0. url: https://www.sciencedirect.com/science/article/pii/S1474667017474040?via%3Dihub
[2] M. W. Spong, “The swing up control problem for the Acrobot,” IEEE Control Systems Magazine, vol. 15, no. 1, pp. 49–55, Feb. 1995, doi: 10.1109/37.341864. url: https://ieeexplore.ieee.org/document/341864
[3] M. W. Spong, “Energy Based Control of a Class of Underactuated Mechanical Systems,” IFAC Proceedings Volumes, vol. 29, no. 1, pp. 2828–2832, Jun. 1996, doi: 10.1016/S1474-6670(17)58105-7. url: https://www.sciencedirect.com/science/article/pii/S1474667017581057?via%3Dihub