This paper focuses on analyzing the following four aspects:First, light passing through galaxy clusters "lacks" a portion of cosmological redshift. The universe is expanding, but galaxy clusters, bound by gravity, do not expand. Consequently, light traversing a galaxy cluster does not undergo cosmological redshift during its passage through the cluster.Second, the expansion of galaxy clusters affects the calculation of the average scale factor of the universe in the past.Third, galaxy clusters accumulate a large amount of matter, leading to a reduction in the matter density of void regions. Although the matter within galaxy clusters also participates in constraining cosmic expansion, its constraining effect on the expansion of void regions is relatively small. This results in a slower time evolution of the expansion rate in the void regions of an inhomogeneous universe, implying that the age of our universe is greater than that of a homogeneous universe. Even the expansion rate of void regions (at the same time calculated backward from the present) is lower than the value predicted by the homogeneous model. By combining the time evolution curves of the Hubble parameter in both models, this paper demonstrates that even after supplementing the "missing" redshift caused by light passing through galaxy clusters, the distance to supernovae calculated using the homogeneous model formula remains smaller than that derived from the inhomogeneous model.Fourth, the standard cosmological model, which默认 uses the volume-weighted average Hubble parameter in the path integral formula to calculate supernova distances, has significant systematic biases. It employs the FLRW metric, effectively treating the large-scale uniformity of the universe as absolute uniformity. This introduces an implicit assumption that the volume-averaged Hubble parameter is equal to the path-averaged Hubble parameter, leading to substantial systematic errors in distance measurements. For example, at a certain redshift, in the model established herein, based on the derived formulas for the path-averaged and volume-averaged Hubble parameters combined with astronomical observation data, the supernova brightness calculated using the path-averaged Hubble parameter is 75% of the brightness predicted by the homogeneous model, while that calculated using the volume-averaged Hubble parameter is 99.1% of the predicted brightness. These results suggest that the formation of stable structures in galaxy clusters may be a crucial reason why the observed brightness of distant supernovae is dimmer than predicted.This paper is an exploratory study and has not yet undergone cross-validation. Any errors are welcome to be criticized and corrected.