Large-scale access of distributed photovoltaic (PV) systems to low-voltage (LV) distribution networks causes voltage overruns and back-feeding problems, which seriously affect the safe and stable operation of power grids. Aiming at the voltage overrun and back-feeding overload caused by the large-scale access of low-voltage distributed photovoltaic (PV) systems, this paper proposes a user regulation and station autonomy strategy based on particle swarm optimization algorithm. A mathematical model is constructed with total power loss, voltage deviation and PV consumption ratio as the optimization objectives, and the particle swarm algorithm is used to solve the optimal access location and capacity configuration of PV system. This includes the establishment of a two-stage topology model for distributed PV systems, the design of PV MPPT control and inverter double-loop control strategies, and the use of PSO algorithm for iterative optimization of decision variables. Pilot application is carried out in a village #2 station area, and the results show that: the total voltage deviation after single-point PV optimization configuration is reduced from 1.311kV to 0.0885kV, with a voltage deviation reduction of 94.36%; the total voltage deviation of multi-point configuration is further reduced to 0.0349kV, with a reduction of 98.47%. After the implementation of the autonomous control strategy, the midday peak voltage is stabilized from over 250V to below 240V, effectively suppressing the backward flow. The optimized PV rated capacity is 970kW, which maximizes the PV consumption under satisfying the voltage constraints. It is proved that the proposed method can effectively improve the voltage quality of distribution network and guarantee the safe and efficient grid connection of distributed PV.