Adaptive Control of DC Power Distribution Systems: Applying Pseudo-Random Sequences and Fourier Techniques

Interconnected AC and DC power-distribution systems are prone to interactions that can reduce system stability. Since the system parameters usually vary over time and with operating conditions, adaptive controllers are required to guarantee desired system operation. Recent studies have presented adaptive methods based on online measurements of converter small-signal transfer functions. This paper proposes a computationally efficient strategy for adaptive control that enables each converter in a system to adjust its own local control. In the paper, a pseudo-random-binary-sequence (PRBS) injection is applied, and the converter control-to-output frequency response is measured online. The controller loop gain is quickly computed based on the measured control-to-output response. A loop-shaping technique is then utilized for adaptive control design to obtain desired stability margins. Advantages of the presented method compared to previously presented techniques are that it does not require a parametric system model for control design, and that the frequency-response measurements do not require opening the feedback loop. The online adaptive control method is fully implementable in the converter digital controller. Experimental results are presented and used to demonstrate the effectiveness of the proposed method.