The SF₆ auxiliary circuit-breaker (ACB) plays a role in the synthetic test circuit to isolate the current source, carry the higher transient recovery voltage of the voltage circuit, protect the current source, so timely grasping the state of the ACB is of great significance for test plan and safety. In this paper, the dynamic contact resistance of the ACB during the opening process after different times of arc erosion in a life cycle is measured using the four-wire method, and the dynamic contact resistance curve is plotted. According to the contact situation of the contacts during the opening process, the dynamic contact resistance curve can be divided into two stages: the main-contact stage and the arc-contact stage. With the deepening of the ablation degree of the ACB contacts, the average dynamic contact resistance of the arc-contact stage shows an increasing trend and the arc contact separation time shows a decreasing trend. The average dynamic contact resistance of the main contact and its separation time have the same trend as those of the arc-contact stage, but the amplitude is smaller than those of the arc-contact stage. After further analysis of the measurement results, the evaluation parameters are determined, and the state evaluation model of the whole life cycle of the ACB is established. The evaluation model is trained by Back Propagation Neural Network optimized by Mind Evolutionary Algorithm (MEA-BP neural network). After the evaluation model is trained, it can be used to evaluate the state of the ACB. In this paper, the accuracy of the model is verified by using the dynamic contact resistance measurement results in other different life cycles.