Ross' π lemma, named after I. Michael Ross,[1][2][3] is a result in computational optimal control. Based on generating Carathéodory-π solutions for feedback control, Ross' π-lemma states that there is fundamental time constant within which a control solution must be computed for controllability and stability. This time constant, known as Ross' time constant,[4][5] is proportional to the inverse of the Lipschitz constant of the vector field that governs the dynamics of a nonlinear control system.[6][7]

Theoretical implications edit

The proportionality factor in the definition of Ross' time constant is dependent upon the magnitude of the disturbance on the plant and the specifications for feedback control. When there are no disturbances, Ross' π-lemma shows that the open-loop optimal solution is the same as the closed-loop one. In the presence of disturbances, the proportionality factor can be written in terms of the Lambert W-function.

Practical applications edit

In practical applications, Ross' time constant can be found by numerical experimentation using DIDO. Ross et al showed that this time constant is connected to the practical implementation of a Caratheodory-π solution.[6] That is, Ross et al showed that if feedback solutions are obtained by zero-order holds only, then a significantly faster sampling rate is needed to achieve controllability and stability. On the other hand, if a feedback solution is implemented by way of a Caratheodory-π technique, then a larger sampling rate can be accommodated. This implies that the computational burden on generating feedback solutions is significantly less than the standard implementations. These concepts have been used to generate collision-avoidance maneuvers in robotics in the presence of uncertain and incomplete information of the static and dynamic obstacles.[8]

See also edit

References edit

  1. ^ Mordukhovich, Boris S. (2006). Variational Analysis and Generalized Differentiation I: Basic Theory. Springer Science & Business Media. ISBN 978-3-540-31247-5.[page needed]
  2. ^ Kang, Wei (November 2010). "Rate of convergence for the Legendre pseudospectral optimal control of feedback linearizable systems". Journal of Control Theory and Applications. 8 (4): 391–405. doi:10.1007/s11768-010-9104-0. S2CID 122945121.
  3. ^ Li, Jr-Shin; Ruths, Justin; Yu, Tsyr-Yan; Arthanari, Haribabu; Wagner, Gerhard (1 February 2011). "Optimal pulse design in quantum control: A unified computational method". Proceedings of the National Academy of Sciences. 108 (5): 1879–1884. Bibcode:2011PNAS..108.1879L. doi:10.1073/pnas.1009797108. PMC 3033291. PMID 21245345.
  4. ^ Bedrossian, Nazareth; Karpenko, Mark; Bhatt, Sagar (November 2012). "Overclock my satellite, sophisticated algorithms boost performance on the cheap". Naval Postgraduate School. hdl:10945/41128.
  5. ^ Stevens, Robert; Wiesel, William (November 2008). "Large Time Scale Optimal Control of an Electrodynamic Tether Satellite". Journal of Guidance, Control, and Dynamics. 31 (6): 1716–1727. Bibcode:2008JGCD...31.1716S. doi:10.2514/1.34897.
  6. ^ a b Ross, I. Michael; Sekhavat, Pooya; Fleming, Andrew; Gong, Qi (March 2008). "Optimal Feedback Control: Foundations, Examples, and Experimental Results for a New Approach". Journal of Guidance, Control, and Dynamics. 31 (2): 307–321. Bibcode:2008JGCD...31..307R. doi:10.2514/1.29532.
  7. ^ Ross, I.M.; Qi Gong; Fahroo, F.; Wei Kang (2006). "Practical stabilization through real-time optimal control". 2006 American Control Conference. pp. 6 pp. doi:10.1109/ACC.2006.1655372. ISBN 1-4244-0209-3. S2CID 16726351.
  8. ^ Hurni, Michael A.; Sekhavat, Pooya; Ross, I. Michael (2010). "An Info-Centric Trajectory Planner for Unmanned Ground Vehicles". Dynamics of Information Systems. Springer Optimization and Its Applications. Vol. 40. pp. 213–232. doi:10.1007/978-1-4419-5689-7_11. ISBN 978-1-4419-5688-0.