When an abnormal event such as islanding or fault occurs in the distribution network

the distributed generation systems should be disconnected from the distribution network by the interconnection protection. This paper starts with the illustration of the functions and characteristics of the interconnection protection. Then an intelligent interconnection protection scheme is proposed based on multi-class support vector machine (SVM). Based on the traditional binary SVM classifier

the proposed scheme is used to build a probability-based multi-class classifier

to realize fault and islanding detection. To improve the generalizing ability of the classifier

SVM-recursive feature elimination (SVM-RFE) algorithm is utilized to identify the critical features. Finally

detailed simulation results are given to demonstrate the effectiveness of the proposed protection scheme over the conventional protections in terms of reliability and security. This work is supported by National Key Technologies R&D Program (No. 2013BAA01B04).