Model description:
The motor dynamics are mapped by aset of five highly coupled nonlinear differential equations as given by
$$\begin{align*} \dfrac{{\mathrm d}i_{qs}^r}{{\mathrm d}t} &= \dfrac{1}{L_{\Sigma}} \left[-\dfrac{L_{RM}^2r_s+M^2r_r}{L_{RM}}i_{qs}^r + \dfrac{r_rM}{L_{RM}}\Psi_{qr}^r - (L_\Sigma i_{ds}^r + M \Psi_{dr}^r) \omega_r + L_{RM}u_{qs}^r\right] \\ \dfrac{{\mathrm d}i_{ds}^r}{{\mathrm d}t} &= \dfrac{1}{L_{\Sigma}} \left[-\dfrac{L_{RM}^2r_s+M^2r_r}{L_{RM}}i_{ds}^r + \dfrac{r_rM}{L_{RM}}\Psi_{dr}^r + (L_\Sigma i_{qs}^r + M \Psi_{qr}^r) \omega_r + L_{RM}u_{ds}^r\right] \\ \dfrac{{\mathrm d}\Psi_{qr}^r}{{\mathrm d}t} &= \dfrac{r_r M}{L_{RM}}i_{qs}^r - \dfrac{r_r}{L_{RM}}\Psi_{qr}^r \\ \dfrac{{\mathrm d}\Psi_{dr}^r}{{\mathrm d}t} &= \dfrac{r_r M}{L_{RM}}i_{ds}^r - \dfrac{r_r}{L_{RM}}\Psi_{dr}^r \\ \dfrac{{\mathrm d}\omega_r}{{\mathrm d}t} &= -\dfrac{B_m}{J}\omega_r + \dfrac{P}{2J}\left[\dfrac{P}{2}\dfrac{M}{L_{RM}}(i_{qs}^r\Psi_{dr}^r - i_{ds}^r-\Psi_{qr})-T_L\right], \end{align*}$$
where $L_{\Sigma} = L_{SM}L_{RM}-M^2$. For the load torque we assume the expression $T_L=c_2\omega_r^2 + c_3\omega_r^3$.
The state vector is given by
$x(t) = [i_{qs}^r, i_{ds}^r, \Psi_{qr}^r, \Psi_{dr}^r, \omega_r]^{\mathrm T} = [x_1, x_2, x_3, x_4, x_5]^{\mathrm T}.$
Type:
Form:
Model order:
Time domain:
Linearity:
Publication details:
| Title | Nonlinear identification of induction motor parameters |
| Publication Type | Conference Paper |
| Year of Publication | 1999 |
| Authors | Pappano, V., Lyshevski S.E., and Friedland B. |
| Conference Name | Proceedings of the 1999 American Control Conference, 1999. |
| Date Published | 01/1999 |
| Publisher | IEEE |
| Conference Location | San Diego, CA |
| ISBN Number | 0-7803-4990-3 |
| Accession Number | 6402981 |
| Keywords | dynamics, identification, multivariable systems, Nonlinear dynamical systems, squirrel cage motors, state-space methods, transients |
| Abstract | In this paper, a nonlinear mapping identification concept is applied to identify the unknown parameters of induction motors using transient dynamics. The developed identification algorithm has significant advantages due to computational efficiency, robustness and convergence, reliability and feasibility. The reported model-based state-space identification can be applied to a wide class of nonlinear multivariable continuous-time dynamic systems. To illustrate the analytical results and to demonstrate the practical capabilities, the unknown motor parameters are found for a squirrel-cage induction motor, under the assumption that all the state vector is available for measurement |
| DOI | 10.1109/ACC.1999.782431 |
