The invention discloses an abnormal state automatic identification model and a method for voltage measurement data. The model comprises: a data preprocessing module for excluding noise data; an optimal parameter matching module for matching an optimal combined input parameter of a density-based clustering algorithm DBSCAN; an abnormal state boundary solving module for solving a relative error maximum value of a normal data sample and serving as a determination value of the abnormal state boundary; an outlier voltage measurement data determination module for using an optimal combined parameterprovided by the optimal parameter matching module as an input condition of the density clustering algorithm DBSCAN, and separating a point that is clearly far from the central area, that is, separating an abnormal state point; and a voltage measurement data abnormal state determination module for using the abnormal state boundary provided by the abnormal state boundary solving module to complete the voltage measurement data abnormal state determination. According to the abnormal state automatic identification model and the method for the voltage measurement data, the technical effect that thestate of the voltage measurement data is quickly and accurately analyzed online, the recognition precision is high, the speed is high, and the operability is strong is realized.