Novel Adaptive Controller for Buck Converter with High Resource Efficiency and Low Computational Complexity

Abstract

Power converters are used in a wide range of industrial processes. Computational complexity, tracking ability, and calculation accuracy are the main parameters that affect the switching performance of power converters. One of the major parts of switch-mode power converters is the controllers which are essential for proper operation. A new adaptive controller is proposed to reduce the computational complexity, the algorithm is presented based on Improved Variable Forgetting Factor (IVFF), Leading Dichotomous Coordinate Descent (DCD), and Exponentially-weighted Recursive Least Square (ERLS). The proposed method estimates the system coefficients with 98% accuracy. The settling time of the output voltage is 0.008[Formula: see text]ms which is faster than other algorithms. According to Leading DCD, this structure needs no multiplier and divider blocks. VFF leads to the improvement of the tracking ability and convergence rate in the system variations. This structure can be implemented on any application that needs an optimal controller. The Vedic mathematics as a multiplier operation is used in the structure of the improved VFF for reducing the calculation delay and area. The error of the proposed method converges to zero with lower than 60 iterations. In other words, the proposed algorithm calculates the optimal coefficients with lower than 50 iterations and is faster than another algorithm.