On the basis of considering the space and time correlation of photovoltaic (PV), a multi-objective distributed PV grid-access planning model including PV total cost, power grid operation cost and active management fee is established by using the chance constrained programming under the active power distribution network. Firstly, the spatial and temporal probability models of different types of PV are established and the random variables are introduced. When the PV is the spatial probability model, the equal probability conversion principle and the Cholesky decomposition technology are used to deal with the spatial correlation of the random variables; when the PV is the time probability model, the method of selection sampling is used to deal with the temporal correlation and form the sample matrix. Finally, a hybrid intelligent algorithm combining random simulation techniques and Cuckoo Algorithm based on Particle Swarm Optimization (CAPSO) is used to solve the above model. The IEEE33 system is selected for example analysis, the simulation results show that the proposed method improves the distributed PV network penetration and reduces the comprehensive economic cost in the process of active distribution network planning. This work is supported by National Natural Science Foundation of China (No. 61473265).