Photovoltaic (PV) power generation is pivotal to global carbon neutrality, yet dust accumulation severely impairs its efficiency and lifespan. Condensation-driven self-cleaning, a cost-effective and eco-friendly solution, is highly dependent on PV panel installation angles (θ). This study experimentally investigates condensation characteristics on clean/soiled PV panels and clarifies the influence mechanism of θ on condensation dynamics, establishing a quantitative correlation between θ and surface condensation behavior for self-cleaning optimization. Filmwise condensation enables gravity-driven cleaning, but residual films/dust compromise performance: clean panels exhibit higher condensate sliding (10.2 mg, 7 cycles) than soiled ones (7.4 mg, 6 cycles) in 4-h tests. Inclination modulates condensation via liquid film thickness and mobility; dust accumulation elevates roughness/thermal resistance, suppressing subcooling and condensate generation. A novel analytical framework decouples sliding (enhancing cleaning) and residual (impeding performance) condensate effects. Increasing θ from 15° to 90° boosts sliding condensate (2.95-fold), cleaning frequency (3.75-fold), and transmittance improvement (52.0%), while reducing residual condensate (80.0%) (200 mg dust). The exponential model TCM = Ae^(B·θ) (R²> 0.89) and Eqs. (3) and (6) characterize θ′s counteracting effects (B₁: positive sliding contribution; B₂: negative residual adhesion), providing a quantitative basis for optimizing PV self-cleaning and offering critical theoretical guidance.