A Cascade BP Neural Network Tuned PID Controller for a High-Voltage Cable-Stripping Robot

Abstract

A 10 kV distribution network is a crucial piece of infrastructure to guarantee enterprises' and households' access to electricity. Stripping cables is one of many power grid maintenance procedures that are now quickly expanding. However, typical cable-stripping procedures are manual and harmful to workers. Although numerous automated solutions for grid maintenance have been created, none of them focus on cable stripping, and most of them have large dimensions to guarantee multi-functions. In this paper, a new cable-stripping robot for the 10 kV power system is introduced. The design of a live working cable-stripping robot that is appropriate for installing insulating rods is introduced, taking into account the working environment of 10 kV overhead lines and the structural characteristics of overhead cables. The robot is managed by an auxiliary remote control device. A cascade PID control technology based on the back propagation neural network (BPNN) method was developed, as the stripper robot's whole system is nonlinear and the traditional PID controller lacked robustness and adaptability in complex circumstances. To validate the structural feasibility of the cable-stripping robot, as well as the working stability and adaptability of the BPNN-PID controller, a 95 mm2 cable-stripping experiment are carried out. A comparison of the BPNN-PID controller with the traditional PID method revealed that the BPNN-PID controller has a greater capacity for speed tracking and system stability. This robot demonstrated its ability to replace manual stripping procedures and will be used for practical routine power maintenance.