The distributed maximum power point tracking (DMPPT) methods, which employ a DC optimizer (DCO) for each individual photovoltaic (PV) panel, are being increasingly introduced to alleviate solar energy waste caused by PV array mismatch issues. Nevertheless, the stability challenge of DMPPT-based distributed PV grid-connected systems incorporating numerous DCOs requires further exploration. Consequently, accurate modeling serves as a fundamental prerequisite for stability analysis. Typically, the model of a PV power plant comprises hundreds or even thousands of DCOs, leading to a significant computational burden during simulations. To address the modeling challenge, this chapter introduces a matrix-variable-based modeling method for distributed PV grid-connected systems. The central concept of the proposed method is to transform the intricate model, which includes numerous PV-DCO generation units, into an average model consisting of only two representative sub-modules by formulating variables using block matrix structures. This approach endows the model with enhanced scalability and improved simulation efficiency. Moreover, it enables the utilization of the vectorized simulation capability available in Matlab/Simulink. Additionally, linearization results can be obtained directly using the Linearization Toolbox in Simulink, circumventing the complexity of manual programming for linearization calculations when analyzing the stability of large-scale systems. With the average model and its corresponding linearization results, the primary influences on the small-signal stability of the system are identified through eigenvalue analysis and root locus methods. It is demonstrated that factors such as the total active power output from the PV array, the controller parameters of the grid-connected inverter, and the strength of the AC system play a pivotal role in affecting the small-signal stability of distributed PV grid-connected systems. Various simulation results validate the effectiveness of the proposed approach.