In order to address the insufficient consideration of wheel polygonal wear and inter-process dependencies in existing reliability analyses of locomotive wheels, which mainly focus on natural degradation and random shocks, a competing-failure reliability model for locomotive wheels considering polygonal wear and mutual dependencies among multiple processes was proposed. Flange thickness and tread diameter were selected as the key performance indicators, and a generalized nonlinear Wiener process was employed to characterize their nonlinear stochastic degradation, while a shared frailty factor was introduced to describe the dependence between the two degradation processes. A multi-state Markov process was used to model the evolution of wheel polygonal wear states. Polygonal wear was incorporated into the natural degradation rate and random shock process, and the mutual dependencies among natural degradation, random shocks, and polygonal wear evolution were further considered. The EM algorithm was applied to estimate model parameters based on measured wheel degradation data, and the Monte Carlo method was employed to calculate system reliability. The results show that the wheel reliability decreases fastest when all inter-process dependencies are considered. Neglecting either polygonal wear or inter-process dependencies leads to an increase in the estimated reliability to varying degrees, with a more pronounced increase in reliability when polygonal wear is ignored. When the reliability levels are 0.9 and 0.5, the corresponding life mileages are 14.03×10^4 km and 20.28×10^4km, respectively. Compared with the model without considering polygonal wear, the reliability of the wheel decreases more rapidly when polygonal wear is taken into account, and the mutual dependencies of polygonal wear with natural degradation and random shocks further promote damage accumulation. It is concluded that incorporating polygonal wear and inter-process dependencies into the competing-failure model can improve the rationality of locomotive wheel reliability assessment and provide a reference for maintenance decision-making.