Abstract: The discharge of contaminated porcelain insulators threatens the safe and stable operation of a power system, thus it is important to timely master the insulation state of contaminated porcelain insulators. This paper aims to investigate an evaluation method for the insulation state of contaminated porcelain insulators based on the solar-blind ultraviolet (UV) detection. At first, discharge tests were conducted on contaminated porcelain insulators on a 220-kV transmission line. On this basis and combined with discharge phenomena of contaminated porcelain insulators as well as characteristics of leakage current signals and UV discharge images, the insulation of contaminated porcelain insulators was graded into safe, warning, alarming, and dangerous states. Three characteristic parameters based on UV irradiance were selected by using the consistency method of UV discharge detection results and the method of statistical analysis. Then, the ambient humidity, instrument gain, and three characteristic parameters based on UV irradiance were taken as the input variables, and the insulation state was used as the output variable. The slime mold algorithm (SMA) was adopted to optimize the penalty factor and the parameter of the kernel function in the support vector machine (SVM), and a model for evaluating the insulation state of contaminated porcelain insulators was established. Finally, the verification results of the above evaluation model for the insulation state in the test set were compared with those of the SVM classification model optimized by the particle swarm optimization (PSO) and sparrow search algorithm (SSA). Results show that the classification accuracy of the proposed model is higher, reaching 93.85%. The established evaluation model for the insulation state of contaminated porcelain insulators is also applicable to other types of UV imagers and improves the practicability of UV detection for contaminated insulators.