Static Voltage Stability Margin Calculation of AC/DC Power System Based on Data-Driven Convex Approximation

In an AC/DC power system with renewable energy source (RES) integration, the optimal power flow (OPF) model of calculating the static voltage stability (SVS) margin with the operation characteristics constraints of DC transmission lines is non-convex and difficult to solve. To address this problem, a data-driven convex approximation (DDCA) method is applied to learn the mapping between the injected bus power, DC variables and the bus voltage at the SVS critical point. In the OPF model, the operation characteristics constraints are rewritten as quadratic equations which can be trained as convex formulations. Combined with the adjacency matrix of the system, an improved relaxed least absolute shrinkage and selection operator (IR-LASSO) regression is proposed to approximate the extended power flow equations by training the data samples to obtain a convex OPF model of SVSM calculation, which can increase the computational efficiency and has good convergence. Case studies on the modified IEEE-39 bus AC/DC system show that the approximate convex OPF model can calculate the SVSM accurately. Moreover, the IR-LASSO regression can greatly increase the computational efficiency.