Recursive Model Predictive Control for Fast Varying Dynamic Systems - Intelligent Computing for Sustainable Energy and Environment - page 111

Tabl e 2. Specifications of the PMSM

RMPC MPC

M P Time(s) M P Time(s)

5 20 0.00071 5 20 0.074

10 20 0.00075 10 20 0.19

20 20 0.00092 20 20 0. 097

10 35 0.0015 10 5 > 5

10 50 0.0022 10 50 > 5

9 1

Iq²Î¿¼Öµ
Iqʵ¼ÊÖµ

Id²Î¿¼Öµ
Idʵ¼ÊÖµ

8

7

0.5

6

5

0

4

3

−0.5

2

1

0 −1

0 0.02 0.04 0.06 0.08 0.1 0 0.02 0.04 0.06 0.08 0.1

ʱ¼ä (t/s) ʱ¼ä (t/s)

(a) (b)

Fig. 4. The performance of RMPC controller. (a) q-axis current, (b) d-axis current.

The reason is that the approach is completed in Matlab. In practice, C lan-
guage is used and controller is often DSP. So it is expectable that the approach
can operate faster in practice.
The full picture of the performance of RMPC is shown in Fig.4. The set point
of i d is 0. To provide setpoints for i q , a PI controller is used in PMSM speed
control. The set point of PMSM speed is 100 rad/s . Fig.4 shows that actual
values can tract setpoints quickly and precisely.

5Con lu on

In this paper, the RMPC suitable for fast varying dynamic systems has been
proposed. While maintaining the general structure of a conventional MPC, the
proposed approach reduces the computational time significantly. Simulation re-
sults have confirmed the effectiveness of the proposed RMPC with a comparison
with the conventional MPC.

References

1. Chen, W.H., Ballance, D.J., Gawthrop, P.J.: Optimal control of nonlinear systems:
a predictive control approach. Automatica 39, 633-641 (2003)
2. Geyer, T., Papafotiou, G., Morari, M.: Model predictive direct torque control Part
I: Concept, algorithm, and analysis. IEEE Transactions on Industrial Electron-
ics 56, 1894-1905 (2009)