In order to improve the evaluation accuracy of the relay protection system and the adaptability of the evaluation method

a risk evaluation model of the relay protection system based on the semi-supervised Mahalanobis distance machine learning algorithm is proposed.First，according to the network topology of the smart substation， the evaluation indicators of the relay protection system are analyzed; second， on the basis of the analytic hierarchy process，the evaluation results are used as the training set for fuzzy comprehensive evaluation. The performance of supervised multiple regression analysis algorithm and unsupervised k-means algorithm is compared with that of the new algorithm proposed in this paper in terms of accuracy

processing time and feasibility.The comparison results show that the proposed algorithm has better evaluation accuracy and stronger adaptability compared to the other two algorithms.