High voltage underground cable bonding optimisation to prevent cable termination faults in mixed high-voltage lines

Abstract

Cable termination fault (CTF) generally occurs due to increasing of electrical force (EF) and thermal effect (TE) in mixed line. EF increases due to sheath voltage (SV), and TE increases due to harmonic distortion (HD) on cable terminations (CTs). Also, SV and HD increase due to zero-sequence current (ZC). In the literature, different bonding methods are used to prevent CTF, but these methods are not sufficient to prevent CTF that is based on ZC. In this study, the modified sectional solid bonding (MSSB) is suggested to prevent CTF. The aims of MSSB method are to restrict EF and TE on CT. Thus, MSSB parameters are optimised to restrict SV and HD, but SV and HD on CT should be known for optimisation of MSSB parameters. Thus, SV and HD are forecasted by using adaptive neuro-fuzzy inference system (Anfis), artificial neural networks (ANNs) and hybrid ANN. Training and forecasting errors of hybrid ANN are less than Anfis and the other ANN methods. Thus, hybrid ANN methods are used as fitness function of optimisation algorithms in MSSB optimisation. When the optimised MSSB method is used for cable bonding, HD and SV on CT are restricted the determined limits to prevent CTF in long mixed line.