Power cables serve as critical hubs for electrical energy transmission

and their insulation performance directly affects the safety and stability of power systems. Regular partial discharge (PD) diagnosis effectively assesses insulation conditions and identifies early-stage defects

making it key to improving power supply reliability. This paper focuses on artificial intelligence (AI)-based diagnostic methods for partial discharge (PD) in power cables and systematically reviews recent research progress in this field. First

the fundamental detection methods for cable PD

including electrical

ultrasonic

and infrared techniques

are introduced

and their advantages and disadvantages are discussed. Next

the paper highlights the application of AI algorithms in cable PD pattern recognition

with a detailed analysis of traditional machine learning and deep learning approaches

covering their evolution

strengths

limitations

and applicability. Finally

the challenges in PD diagnosis are summarized

and future research directions are discussed from three key aspects: data acquisition

algorithm design

and the development of cable monitoring systems.