n this paper, the least squares support vector machines (LS‐SVMs) based design of superconducting magnetic energy storage (SMES) controller is proposed for wide area stability control. The LS‐SVMs for SMES controllers are trained by local and inter‐area data based on synchronized phasor measurements considering time delay. A large amount of training data set of a multi‐machine power system is reduced by the measurement of similarity among samples. The LS‐SVM parameters and the similarity threshold are optimized by a particle swarm optimization. Subsequently, the redundant data in the training set can be discarded while the reduced data are the optimal support vectors in the LS‐SVM model. The LS‐SVM control signals can be adapted by various operating conditions and different disturbances. Simulation results in a six‐area West Japan interconnected power system demonstrate that the proposed LS‐SVM for SMES controller is robust to various disturbances under wide range of operating conditions in comparison to the conventional SMES. Copyright © 2011 John Wiley & Sons, Ltd.