Abstract
The Monash Epicyclic Parallel Manipulator (MEPaM) is a novel six-degree-of-freedom (dof) parallel mechanism with base mounted actuators. Closed form equations of the inverse dynamic model of MEPaM are obtained through two different methods, with simulation showing these models to be equivalent. The base inertial parameters for the dynamic model of MEPaM are determined, reducing the number of inertial parameters from 100 to 28. This significantly simplifies the dynamic calibration model and thus the number of computations required.
Similar content being viewed by others
References
Akbarzadeh, A., Enferadi, J., Sharifnia, M.: Dynamics analysis of a 3-RRP spherical parallel manipulator using the natural orthogonal complement. Multibody Syst. Dyn. 29(4), 361–380 (2013)
Angeles, J.: Fundamentals of Robotic Mechanical Systems: Theory, Methods and Algorithms, 3rd edn. Springer, Berlin (2007)
Angeles, J., Lee, S.: The formulation of dynamical equations of holonomic mechanical systems using a natural orthogonal complement. J. Appl. Mech. 55(1), 243–244 (1988)
Angeles, J., Ma, O.: Dynamic simulation of \(n\)-axis serial robotic manipulators using a natural orthogonal complement. Int. J. Robot. Res. 7(5), 32–47 (1988)
Chen, C., Gayral, T., Caro, S., Chablat, D., Moroz, G., Abeywardena, S.: A six degree of freedom epicyclic-parallel manipulator. J. Mech. Robot. 4(4), 041011 (2012)
Cleary, K., Brooks, T.: Kinematic analysis of a novel 6-dof parallel manipulator. In: IEEE International Conference on Robotics and Automation, pp. 708–713 (1993)
Craig, J.: Introduction to Robotics: Mechanics and Control, 3rd edn. Prentice Hall, New York (2005)
Dasgupta, B., Mruthyunjaya, T.: A Newton–Euler formulation for the inverse dynamics of the Stewart platform manipulator. Mech. Mach. Theory 33(8), 1135–1152 (1998)
Enferadi, J., Akbarzadeh, A.: Inverse dynamics analysis of a general spherical star-triangle parallel manipulator using principle of virtual work. Nonlinear Dyn. (2010)
Farhat, N., Mata, V., Page, A., Valero, F.: Identification of dynamic parameters of a 3-dof RPS parallel manipulator. Mech. Mach. Theory 43(1), 1–17 (2008)
Fattah, A., Kasaei, G.: Kinematics and dynamics of a parallel manipulator with a new architecture. Robotica 18, 535–543 (2000)
Gautier, M.: Numerical calculation of the base inertial parameters of robots. J. Robot. Syst. 8(4), 485–506 (1991)
Gautier, M., Khalil, W.: Direct Calculation of minimum set of inertial parameters of serial robots. IEEE Trans. Robot. Autom. 6(3), 368–373 (1990)
Geng, Z., Haynes, L.S., Lee, J.D., Carroll, R.L.: On the dynamic model and kinematic analysis of a class of Stewart platforms. Robot. Auton. Syst. 9(4), 237–254 (1992)
Gosselin, C.: Parallel computational algorithms for the kinematics and dynamics of planar and spatial parallel manipulators. J. Dyn. Syst. Meas. Control 118(1), 22–28 (1996)
Guegan, S., Khalil, W.: Dynamic modeling of the Orthoglide. In: Lenarčič, J., Thomas, F. (eds.) Advances in Robot Kinematics, pp. 387–396. Springer, Netherlands (2002)
Ibrahim, O., Khalil, W.: Kinematic and dynamic modeling of the 3-rps parallel manipulator. In: Proc. 12th IFToMM World Congress, vol. 19, pp. 95–100 (2007)
Khalil, W., Bennis, F.: Symbolic calculation of the base inertial parameters of closed-loop robots. Int. J. Robot. Res. 14(2), 112–128 (1995)
Khalil, W., Dombre, E.: Modeling, Identification and Control of Robots. Butterworth-Heinemann, Oxford (2004)
Khalil, W., Guegan, S.: Inverse and direct dynamic modelling of Gough–Stewart robots. IEEE Trans. Robot. Autom. 20(4), 754–762 (2004)
Khalil, W., Ibrahim, O.: General solution for the dynamic modelling of parallel robots. J. Intell. Robot. Syst. 49(1), 19–37 (2007)
Khalil, W., Kleinfinger, J.: A new geometric notation for open and closed-loop robots. In: Proceedings of IEEE International Conference on Robotics and Automation, vol. 3, pp. 1174–1179 (1986)
Khalil, W., Kleinfinger, J.F.: Minimum operations and minimum parameters of the dynamic models of tree structure robots. IEEE J. Robot. Autom. 3(6), 517–526 (1987)
Khosla, P., Kanade, T.: Parameter identification of robot dynamics. In: 24th IEEE Conference on Decision and Control, vol. 24, pp. 1754–1760 (1985). 1985
Lebret, G., Liu, K., Lewis, F.L.: Dynamic analysis and control of a Stewart platform manipulator. J. Robot. Syst. 10(5), 629–655 (1993)
Lee, S.U., Kim, S.: Analysis and optimal design of a new 6-dof parallel type haptic device. In: IEEE/RSJ International Conference on Intelligent Robots and Systems, pp. 460–465 (2006)
Liu, M.J., Li, C.X., Li, C.N.: Dynamics analysis of the Gough–Stewart platform manipulator. IEEE Trans. Robot. Autom. 16(1), 94–98 (2000)
Ma, O.: Mechanical analysis of parallel manipulators with simulation, design, and control applications. PhD thesis, McGill University (1991)
Mayeda, H., Yoshida, K., Osuka, K.: Base parameters of manipulator dynamic models. IEEE Trans. Robot. Autom. 6(3), 312–321 (1990)
Monsarrat, B., Gosselin, C.: Workspace analysis and optimal design of a 3-leg 6-dof parallel platform mechanism. IEEE Trans. Robot. Autom. 19(6), 954–966 (2003)
Rao, A., Saha, S., Rao, P.: Dynamics modelling of hexaslides using the decoupled natural orthogonal complement matrices. Multibody Syst. Dyn. 15(2), 159–180 (2006)
Tsai, L.: Solving the inverse dynamics of a Stewart–Gough manipulator by the principle of virtual work. J. Mech. Des. 122, 3–9 (2000)
Wang, J., Gosselin, C.: A new approach for the dynamic analysis of parallel manipulators. Multibody Syst. Dyn. 2(3), 317–334 (1998)
Xi, F., Sinatra, F.: Inverse dynamics of hexapods using the natural orthogonal complement method. J. Manuf. Syst. 21(2), 73–82 (2002)
Zanganeh, K.E., Sinatra, R., Angeles, J.: Kinematics and dynamics of a six-degree-of-freedom parallel manipulator with revolute legs. Robotica 15, 385–394 (1997)
Author information
Authors and Affiliations
Corresponding author
Appendices
Appendix A: Recursive energy relations
Letting the transformation matrix \({^{B}\mathbf{T}_{j}} = \bigl[{\scriptsize\begin{matrix}{}\mathbf{i} & \mathbf{j} & \mathbf{k} & {^{B}\mathbf{p}_{j}} \cr 0 & 0 & 0 & 1\end{matrix}}\bigr]\) with \(\mathbf{i}\), \(\mathbf{j}\) and \(\mathbf{k}\) being \(3\times1\) vectors which describe the orientation of link \(j\) with respect to the base frame, the angular velocity of link \(j\) be \(\boldsymbol{\omega}_{j} = [\omega_{xj} \ \omega_{yj} \ \omega_{zj}] ^{T}\) and the linear velocity of link \(j\) be \(\mathbf{v}_{j} = [v_{xj} \ v_{yj} \ v_{zj}]^{T}\), the recursive energy relations associated with the link’s inertial parameter vector \(\boldsymbol{\chi}_{j}\) are [19, 20]
Appendix B: Grouping relations for serial legs
A complete derivation and overview of the grouping relations for serial and tree robots can be found in [13, 18, 19]. The rules and expressions used to determine the base inertial parameters of the serial legs of MEPaM are considered below. Using the rules developed in [13, 18, 19], where possible, the parameters of link \(3i\) were grouped with the parameters of link \(2i\) which were subsequently grouped with the parameters of link \(1i\). With reference to Table 1, \(\gamma_{j} = 0\) and \(b_{j} = 0\) for \(j = 2\) and 3, hence the rules which pertain to serial robots suffice. These are (in terms of the Khalil–Kleinfinger parameters):
-
If joint \(j\) is revolute, the parameters \(YY_{j}\), \(MZ_{j}\) and \(m_{j}\) can be grouped with the parameters of link \(j\) and link \(j-1\). The resulting parameters are:
$$\begin{aligned} XX_{Rj} &= XX_{j} - YY_{j}, \\ XX_{Rj-1} &= XX_{j-1} + YY_{j} + 2d_{j}MZ_{j} + {d_{j}}^{2}m_{j}, \\ XY_{Rj-1} &= XY_{j-1} + a_{j}s_{\alpha_{j}}MZ_{j}+a_{j}d_{j}s_{\alpha_{j}}m_{j}, \\ XZ_{Rj-1} &= XZ_{j-1} - a_{j}c_{\alpha_{j}}MZ_{j}-a_{j}d_{j}c_{\alpha_{j}}m_{j}, \\ YY_{Rj-1} &= YY_{j-1} + c_{\alpha_{j}}^{2}YY_{j} + 2d_{j}c_{\alpha_{j}}^{2}MZ_{j} + \bigl(a_{j}^{2}+d_{j}^{2}{c_{\alpha_{j}}}^{2} \bigr)m_{j}, \\ YZ_{Rj-1} &= YZ_{j-1} + c_{\alpha_{j}}s_{\alpha_{j}}YY_{j} + 2d_{j}c_{\alpha _{j}}s_{\alpha_{j}}MZ_{j} + {d_{j}}^{2}c_{\alpha_{j}}s_{\alpha_{j}}m_{j}, \\ ZZ_{Rj-1} &= ZZ_{j-1} + s_{\alpha_{j}}^{2}YY_{j} + 2d_{j}s_{\alpha_{j}}^{2}MZ_{j} + \bigl(a_{j}^{2}+d_{j}^{2}{s_{\alpha_{j}}}^{2} \bigr)m_{j}, \\ MX_{Rj-1} &= MX_{j-1} + a_{j}M_{j}, \\ MY_{Rj-1} &= MY_{j-1} - s_{\alpha_{j}}MZ_{j} - d_{j}s_{\alpha_{j}}m_{j}, \\ MZ_{Rj-1} &= MZ_{j-1} + c_{\alpha_{j}}MZ_{j} + d_{j}c_{\alpha_{j}}m_{j}. \\ m_{Rj-1} &= m_{j-1} + m_{j}. \end{aligned}$$ -
If joint \(j\) is prismatic, the parameters of the inertia tensor of link \(j\) (\(XX_{j}, XY_{j}, XZ_{j}, YY_{j}, YZ_{j}, ZZ_{j}\)) can be grouped with those of link \(j-1\). The resulting parameters are:
$$\begin{aligned} XX_{Rj-1} &= XX_{j-1} + {c_{\theta_{j}}}^{2}XX_{j} - 2c_{\theta_{j}}s_{\theta _{j}}XY_{j} + {s_{\theta_{j}}}^{2}YY_{j}, \\ XY_{Rj-1} &= XY_{j-1} + c_{\theta_{j}}s_{\theta_{j}}c_{\alpha_{j}}XX_{j} + \bigl(c_{\theta_{j}}^{2}-s_{\theta_{j}}^{2} \bigr)c_{\alpha_{j}}XY_{j}-c_{\theta_{j}}s_{\alpha _{j}}XZ_{j} \\ &\quad{}-c_{\theta_{j}}s_{\theta_{j}}c_{\alpha_{j}}YY_{j}+s_{\theta _{j}}s_{\alpha_{j}}YZ_{j}, \\ XZ_{Rj-1} &= XZ_{j-1} + c_{\theta_{j}}s_{\theta_{j}}s_{\alpha_{j}}XX_{j} + \bigl(c_{\theta_{j}}^{2}-s_{\theta_{j}}^{2} \bigr)s_{\alpha_{j}}XY_{j}+c_{\theta_{j}}c_{\alpha _{j}}XZ_{j} \\ &\quad{}-c_{\theta_{j}}s_{\theta_{j}}s_{\alpha_{j}}YY_{j}-s_{\theta _{j}}c_{\alpha_{j}}YZ_{j}, \\ YY_{Rj-1} &= YY_{j-1} + s_{\theta_{j}}^{2}c_{\alpha_{j}}^{2}XX_{j} + 2c_{\theta _{j}}s_{\theta_{j}}c_{\alpha_{j}}^{2}XY_{j}-2s_{\theta_{j}}c_{\alpha_{j}}s_{\alpha _{j}}XZ_{j}+c_{\theta_{j}}^{2}c_{\alpha_{j}}^{2}YY_{j} \\ &\quad{}-2c_{\theta_{j}}c_{\alpha _{j}}s_{\alpha_{j}}YZ_{j} + s_{\alpha_{j}}^{2}ZZ_{j}, \\ YZ_{Rj-1}& = YZ_{j-1} + s_{\theta_{j}}^{2}c_{\alpha_{j}}s_{\alpha_{j}}XX_{j} + 2c_{\theta_{j}}s_{\theta_{j}}c_{\alpha_{j}}s_{\alpha_{j}}XY_{j}+s_{\theta _{j}} \bigl(c_{\alpha_{j}}^{2}-s_{\alpha_{j}}^{2} \bigr)XZ_{j} \\ &\quad{}+c_{\theta_{j}}^{2}c_{\alpha _{j}}s_{\alpha_{j}}YY_{j}+2c_{\theta_{j}} \bigl(c_{\alpha_{j}}^{2}-s_{\alpha_{j}}^{2} \bigr)YZ_{j} - c_{\alpha_{j}}s_{\alpha_{j}}ZZ_{j}, \\ ZZ_{Rj-1} &= ZZ_{j-1} + s_{\theta_{j}}^{2}s_{\alpha_{j}}^{2}XX_{j} + 2c_{\theta _{j}}s_{\theta_{j}}s_{\alpha_{j}}^{2}XY_{j}+2s_{\theta_{j}}c_{\alpha_{j}}s_{\alpha _{j}}XZ_{j}+c_{\theta_{j}}^{2}s_{\alpha_{j}}^{2}YY_{j} \\ &\quad{}+2c_{\theta_{j}}c_{\alpha _{j}}s_{\alpha_{j}}YZ_{j} + c_{\alpha_{j}}^{2}ZZ_{j}. \end{aligned}$$ -
If the axis of the prismatic joint \(j\) is parallel to the nearest revolute joint axis \(i\) (\(i\) is not necessarily \(j-1\)), then \(MZ_{j}\) has no effect on the dynamic model and the parameters \(MX_{j}\) and \(MY_{j}\) can be grouped as follows:
$$\begin{aligned} MX_{Rj-1}& = MX_{j-1} + c_{\theta_{j}}MX_{j} - s_{\theta_{j}}MY_{j}, \\ MY_{Rj-1} &= MY_{j-1} + s_{\theta_{j}}c_{\alpha_{j}}MX_{j} + c_{\theta _{j}}c_{\alpha_{j}}MY_{j}, \\ MZ_{Rj-1}& = MZ_{j-1} + s_{\theta_{j}}s_{\alpha_{j}}MX_{j} + c_{\theta _{j}}s_{\alpha_{j}}MY_{j}, \\ ZZ_{Ri} &= ZZ_{i} + 2a_{j}c_{\theta_{j}}MX_{j} - 2a_{j}s_{\theta_{j}}MY_{j}. \end{aligned}$$ -
If joint \(j\) is revolute and articulated on the base, the parameters \(XX_{j}\), \(XY_{j}\), \(XZ_{j}\), \(YY_{j}\), \(YZ_{j}\), \(MZ_{j}\) and \(m_{j}\) have no effect on the dynamic model and can be eliminated.
Appendix C: Grouping leg parameters with platform
For the parameters \(m_{3i}\) to be grouped with the parameters of the platform, the energy function \(e_{m_{3i}}\) needs to be in a form that is a linear combination of the platform energy functions, i.e.
This is possible by use of the velocity of the connection point of the legs to the platform, i.e.
where \({^{P}\mathbf{v}_{P}} = [^{P}v_{xP} \ ^{P}v_{yP} \ ^{P}v_{zP}]^{T}\), \({^{P}\boldsymbol{\omega}_{P}} = [^{P}\omega_{xP} \ ^{P}\omega_{yP} \ ^{P}\omega_{zP}]^{T}\) and \({^{P}\boldsymbol{\rho}_{i}} = [^{P}b_{xi} \ ^{P}b_{yi} \ ^{P}b_{zi}]^{T}\).
Let the transformation of the platform frame to the base frame be
where \(\mathbf{i}\), \(\mathbf{j}\) and \(\mathbf{k}\) are \(3\times1\) vectors which describe the orientation of the platform with respect to the base frame and \({^{B}\mathbf{p}_{P}}\) is the translation between the origin of the two frames. Then the co-ordinates of the connection points in the base frame are
Therefore, the energy function for the masses \(m_{3i}\) can be determined using Eq. (49) for \(e_{m_{j}}\), Eq. (51) for the velocity of link \(3i\) and Eq. (53) for the connection point expressed in \(\mathcal{F}_{B}\), i.e.
With aid of Eq. (49), Eq. (54) is rearranged to
As a result of Eq. (55), it is possible to group the \(m_{3i}\) parameters with the inertial parameters of the platform. Due to the assignment of the platform frame, \(\mathcal{F}_{P}\), the co-ordinates of the platform vertices in \(\mathcal{F}_{P}\) are:
Hence, the grouping relations are
Rights and permissions
About this article
Cite this article
Abeywardena, S., Chen, C. Inverse dynamic modelling of a three-legged six-degree-of-freedom parallel mechanism. Multibody Syst Dyn 41, 1–24 (2017). https://doi.org/10.1007/s11044-016-9506-y
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11044-016-9506-y