Development of a Novel Controller for Brushless DC Motor Drive Systems Combining Decision Tree and Sliding Mode Theory

This paper proposes a novel speed controller design for a brushless DC motor (BLDCM) under field-oriented control (FOC). The designed novel speed controller combines decision tree theory (DTT) and sliding mode theory (SMT). First, the regression algorithm of the classification and regression tree (CART) within decision tree theory is utilized to divide the speed error between the BLDCM's speed command and actual speed into 10 intervals. Based on this, three parameters of the existing exponential reaching law sliding mode controller (ERLSMC)—the sliding mode dynamic trajectory control gain, the exponential reaching gain, and the constant speed reaching gain—are configured. Next, the mean squared error (MSE) of each node after splitting is calculated to identify the root node. According to the selected split variable and splitting point, the data is divided into two subsets, and this process is repeated for each child node. Consequently, during the operation of the BLDCM, appropriate adjustments for the three gains can be provided to the sliding mode speed controller. Subsequently, a new sliding mode dynamic trajectory control gain is recalculated based on the rate of change of the speed error. This allows the overshoot in the system's speed response, caused by adopting the exponential reaching law (ERL), to be improved under different operating conditions through the modulation of its three gains. It also enables the speed response of the BLDCM drive system to rapidly track the speed command under various operating conditions. Therefore, the proposed control law involves no complex computations and does not require a massive amount of training data, making it easy to implement. Finally, Matlab/Simulink simulation software is used to simulate the application of the proposed control law to the BLDCM drive system. Its control performance is compared with sliding mode controllers (SMCs) utilizing three different reaching laws: the constant speed reaching law (CSRL), the ERL, and the extension theory combined with exponential reaching law (ETERL). The simulation results demonstrate that the proposed novel speed controller outperforms the SMCs with the other three reaching laws in terms of both speed command tracking and load regulation response.