An Electro-Thermal Battery Runtime Prediction Framework Based on Markov Chain and Monte Carlo Simulation

Abstract

This paper proposes an electrothermal coupling modeling and stochastic simulation method for predicting battery runtime in mobile devices. An electrical model of lithium-ion batteries is constructed based on Thevenin equivalent circuits, combined with SOC dynamics and polarization voltage evolution to establish a continuous-time state-space model. Concurrently, thermal equilibrium equations describe Joule heating generation and convective heat dissipation, forming an electro-thermal coupled kinetic model to characterize the coupling relationships among temperature, current, and internal resistance. For load modeling, a Markov chain describes the random switching of device operating states, overlaid with Gaussian perturbations to construct a stochastic power model. Based on this, Monte Carlo simulations generate multiple load sequences, iteratively updating SOC, voltage, and temperature states to derive the statistical distribution of the device's remaining operational time. Results demonstrate that this method effectively reflects the impact of multiple factors on battery endurance performance, exhibiting good applicability and potential for broader implementation.