<p>Conventional methods for estimating the residual capacity of lead-acid batteries often overlook the variations in available capacity across different environments and usage scenarios throughout the life cycle of batteries, as well as the natural aging and degradation processes. The oversight results in inaccurate capacity estimations, subsequently shortening battery lifespan and diminishing system reliability. This paper introduces a model to predict the available residual capacity of batteries throughout their life cycle, taking into account various disturbances such as temperature, discharge rate, and degradation. The changes of state-of-health are characterized using internal resistance, while the battery is modeled as a nonlinear system affected by disturbances. Variational modal decomposition is employed to enhance the accuracy of internal resistance measurement and state-of-health assessment. Furthermore, to address issues like error accumulation in the ampere-hour integration method and filter dispersion in extended Kalman filtering, the two approaches are fused by using a weighting factor, and the prediction model is dynamically updated to implement temperature compensation and aging correction. Experiments show that this approach can rapidly estimate the initial capacity of the battery and achieve precise estimation of the available residual capacity under varying conditions, enhancing the effectiveness of lifecycle capacity prediction.</p>