Uncertainty Level of Voltage in Distribution Network: An Analysis Model With Elastic Net and Application in Storage Configuration

Abstract

Quantitatively assessing the impact of uncertainties on node voltage in a distribution network is an important issue. This paper proposes a new model to analyze the uncertainty level of voltage from the perspective of the linear quantitative description with elastic net regression. To obtain the model parameters defined as the voltage influence coefficients, the procedures are as follows: first, estimate the expected values of node voltage by point-estimate method, and conduct the analysis sample collection; then determine the dominant factors and calculate their voltage influence coefficients for the above sample by elastic net regression. To further illustrate the practical application of the proposed model, a coordinated optimization configuration method for energy storage mounted on distributed generations is presented in detail. The verification results based on the modified IEEE 33-bus system and a modified practical 113-bus system demonstrate that the analysis model with elastic net has an excellent performance and is of high accuracy in assessing the influence of uncertainties on node voltage. Furthermore, the optimization results of multi-ESS for the test system illustrates that the proposed configuration method can produce a global and practical configuration scheme for the whole distribution system.