A Unified Battery Discharge Prediction Framework Based on Electro-Thermal Coupling Model and Markov-Modulated Monte Carlo Algorithm

Abstract

This paper addresses the modeling of dynamic characteristics during the discharge process of smart device batteries by constructing a continuous-time equivalent model that integrates electrical and thermal coupling mechanisms. At the electrical level, an improved first-order Thevenin equivalent circuit is introduced to uniformly model open-circuit voltage, ohmic internal resistance, and polarization effects. Combined with a constant-power load constraint, this model characterizes the nonlinear feedback behavior where voltage drop triggers current amplification; At the thermal level, a temperature evolution model is established based on the equilibrium between Joule heating and convective cooling, while a temperature-dependent internal resistance expression based on the Arrhenius equation is introduced to achieve a bidirectional electro-thermal coupling description. On this basis, the system is transformed into a state-space form to uniformly characterize the cooperative evolution of the charge state, polarization voltage, and temperature. Furthermore, to address random load fluctuations and performance degradation in real-world usage scenarios, a Monte Carlo simulation framework based on continuous-time Markov processes is proposed. This framework incorporates Gaussian perturbations to model load uncertainty and introduces a health state parameter to unify the modeling of capacity degradation and internal resistance increase. Simulation results demonstrate that this method effectively captures discharge behavior and lifespan trends under multi-factor coupling, exhibiting good generalization capability and predictive stability. It provides a unified modeling foundation for battery performance evaluation and energy management.