The invention relates to the field of electrical automation, and discloses a power grid voltage out-of-limit cause diagnosis method based on a data mining technology, which comprises the following steps: A) extracting out-of-limit voltage, and collecting related characteristics of the out-of-limit voltage; B) deleting bad data, and carrying out data preprocessing to obtain sample data; C) constructing a K-means improved clustering model, and obtaining a voltage out-of-limit cause clustering result; D) carrying out classification identification on voltage out-of-limit cause clustering results;e) constructing an SVM multi-classifier model, and optimizing model parameters; and F) inputting a to-be-diagnosed data sample to the SVM multi-classification model, and taking the output of the SVM multi-classification model as the cause of voltage out-of-limit of the to-be-diagnosed data sample. The method is high in efficiency, reduces the influence of human subjective factors, quickly and effectively processes the power grid low-voltage mass data, improves the K-means clustering model, obtains a good clustering effect, carries out the parameter optimization of the SVM model, and improves the accuracy of the power grid voltage out-of-limit cause diagnosis.