A Novel Temperature Prediction Method in Complex Indoor Environments Based on Parameter Estimation

This article proposes a novel temperature prediction technique for complex indoor environments based on parameter estimation. The method utilizes the Boussinesq approximation to establish a thermal model and estimate key parameters including pressure terms, heat transfer coefficients, and thermal diffusivity. An approximate approach is adopted to simulate the thermal dynamics of an indoor space. By comparing simulation results under an initial temperature with actual measurements, accurate parameter values are obtained. The estimated parameters can then be applied to achieve high-accuracy temperature prediction given varying initial conditions in the same environment. Moreover, the method can be formulated into a state-space model, enabling offline optimal temperature control. Through identifying the governing parameters via data-driven modeling, it provides an effective engineering tool for analyzing heat transfer and regulating temperature in intricate indoor settings. Rigorous derivation and sufficient validation demonstrate the reliability of this method. Given its potential in fields like architecture design and thermal engineering, this technique offers valuable impacts and insights for temperature management in enclosed environments.