Improving the accuracy of insulation resistance measurements of a single arm unbalanced bridge using a parameter matching design and filtering algorithm for battery packs used in electric ships

Abstract

In the DC system of electric ships, insulation drops and grounding will pose a threat to stable ship operation and personal safety. Therefore, it is necessary to monitor the insulation status of the battery pack between high voltage and ground in real time. However, traditional measurement methods have problems such as low accuracy, long measurement time, and poor reliability. In this paper, an insulation detection method based on an unbalanced bridge was proposed and the principle of insulation resistance detection was analyzed. The influence of different resistance values of insulation detection circuits on the measurement error of insulation resistance was studied by simulation analysis and experimental tests to obtain the optimal set of bridge resistance matching designs. Digital filter and Kalman filter were used to filter out noise in the voltage signal. The results show that measurement error of insulation resistance after optimization was less than 3% and drops to approximately 0.46% after processing by the Kalman filter algorithm, which proves that the insulation detection method designed in this paper has high measurement accuracy.