High performance algorithms for the control and load identification of boost DC-DC converters

This paper proposes two high performance control methods and an on-line load identification algorithm for boost DC-DC converters. The first is a short-horizon model predictive control method whose solution can be computed analytically, thus rendering an algorithm that can be implemented on-line. Further, we show how to find a suitable performance index so as to achieve good response. The idea is to use the inductance current as the penalized output, instead of the output voltage. The second is a 3-level hysteresis control method which is a natural extension of the conventional hysteresis control method and has three layers in the hysteresis band. The key is to set the duty ratio in the central layer as the steady-state duty ratio. Compared with the well-known PID-PWM method, the proposed methods are able to achieve a better trade-off between rise-time and overshoot in the response of output voltage. Finally, a simple on-line identification algorithm is proposed for the load resistance. Combination of the proposed control methods and identification algorithm yields high performance robust control for boost converters.