Predictive Control of Power Converters and Electrical Drives

6.4 Predictive Power Control

By using the rectifier model and considering the instantaneous power theory [7], it is possible to predict the behavior of the input active power and reactive power at the input of the converter. Then, by defining an appropriate cost function it is possible to directly control the power flow between the converter and the grid [8].

The switching state of the converter is changed at equidistant time instants and is constant during a whole sampling interval. On each sampling interval, the control strategy selects the switching state that will be applied by minimization of a cost function.

Predictive power control has no internal control loops and does not need external modulators. The currents are forced by directly controlling the active and reactive power.

6.4.1 Cost Function and Control Scheme

The block diagram of the control strategy is shown in Figure 6.9. The input currents i_s(k) are measured and the future current i_s(k + 1) is calculated using the applied converter voltage v_afe(k). Predictions of the future current i_s(k + 1) are generated for each one of the seven possible voltage vectors v_afe generated by the AFE. These predictions are used to calculate the future input active and reactive power, P_in(k + 1) and Q_in(k + 1), using (6.13) and (6.14). Each prediction of P_in(k + 1) and Q_in(k + 1) is evaluated by the cost function g_afe.

Figure 6.9 Predictive power control scheme for the rectifier (Cortes et al., 2008 © IEEE)

The cost function g_{afe ...}