The penetration of distributed generation (DG) is becoming higher and higher in the distribution network, and so it is difficult to completely eliminate the voltage ‘out of limit’ of the network. This paper constructs a bi-level optimization model considering equipment planning and the operation of an active distribution network. The upper model considers the capacity allocation optimization of DG, static var compensator (SVC) and capacitor bank (CB) in the network to reduce the investment and operational cost. In the lower model, the coordinated control of demand response (DR), on-load tap changer (OLTC), energy storage (ES) and DG are considered to improve voltage stability. From the characteristics of the model, a hybrid optimization algorithm based on the improved sparrow algorithm (CLSSA) and second-order cone (SOC) programming are used to solve the problem. The CLSSA algorithm is used to determine multi-dimensional variables externally to improve the solution speed, and the active management and control strategy of the network is obtained internally based on SOC programming. Finally, the modified IEEE33-bus system is used for simulation verification. The results show that the bi-level optimization model can effectively enhance the operation economy of the distribution network, improve the network voltage distribution and stabilize peak-valley difference. This work is supported by the National Natural Science Foundation of China (No. 52067020).