ABSTRACT
Based on previous research, it can be concluded that the traditional power supply reliability forecasting methods usually require detailed distribution network structure and a lot of accurate equipment reliability parameters, such as cable failure rate, mean time to repair each line, transformer failure rate, contact switching time and the number of fuses, etc. Therefore, the traditional methods are only applied to regional power system with simple structure and complete equipment reliability parameters. When it comes to largescale grid with complex structure and missing equipment reliability parameters, researchers have proposed novel supply reliability prediction algorithms based on statistical methods which greatly reduce the dependence on the device parameters, and have broad application prospects. Song et al. [4] proposed a reliability prediction method for
1 INTRODUCTION
Nowadays electricity, used widely in all walks of life, plays an increasingly important role in social development and economic growth. Social production and living are inseparable from secure, highquality, and efficient power supply. Specifically, power enterprises tend to be customer-oriented, in order to better understand and meet customers’ demand; “customer satisfaction” is regarded as the ultimate goal and the pursuit of improving service levels never stops [1]. Naturally, the power outage has a detrimental influence on the normal production, lowers living quality and directly causes user dissatisfaction. Therefore, it has become an inevitable trend that power supply reliability evaluation and predictive analytics are integrated into power supply system management.
