The support vector machine is a learning algorithm that uses fewer
training samples to achieve the classification and generalization ability. It has some
advantages, like a few adjustable parameters, computing speed and small-time cost,
and it can have nothing to do with the data dimension. It has also a good scalability.
This paper introduces the basic principle and summarizes the successful applications
in load forecast, soft measurement, electrical equipment fault diagnosis and the
stability analysis of power system. The results show that the support vector machine
has overcome the limitations of the traditional algorithm, such as the difficulty
of controlling the convergence and the structural design. All in all, it has broad
application prospects.