Contact Resistance Volatility Prediction of Pantograph-Catenary Based on IPOA-ELM

Abstract

The volatility of contact resistance directly reflects the fluctuation of friction pairs in the pantograph-catenary system, so this paper proposes a new prediction model of contact resistance volatility based on an improved Pelican optimization algorithm and optimized extreme learning machine to judge the magnitude of contact resistance volatility. The paper first examines the volatility of contact resistance of pantograph-catenary under fluctuating loads. It concludes that it has an inverse relationship with sliding velocity and pressure fluctuation frequency but a positive correlation with the given current and pressure fluctuation amplitude. Secondly, the Pelican optimization algorithm (POA) is enhanced to improve its stability, accuracy, and speed. The simulation results demonstrate that the improved Pelican optimization algorithm (IPOA) outperforms the standard POA. Finally, the IPOA-ELM model is developed by optimizing the extreme learning machine (ELM) parameters with IPOA. The experimental results show that the IPOA-ELM model is highly effective in predicting contact resistance volatility.

Keywords

• Fluctuating load
• contact resistance volatility
• pelican optimization algorithm
• extreme learning machine