•  
  •  
 

Keywords

Stability in voltages, Biting contingencies, Reactive power planning, Hybrid metaheuristic, Mixed-integer optimization, Power system resiliency

Document Type

Article

Abstract

In this paper, a two-stage optimization scheme of proactive resilience of the transmission system to severe multi-bus contingencies is developed. A hybrid PSO-GSA-Quantum metaheuristic is used to find the 4-bus reactive fault clusters in the IEEE 118-bus system in stage 1. The cluster identified with the performance index (PI) of 1.8875 × 103 has the 91.6th percentile of 10000 random Monte Carlo samples and 1.78 times worse than the worst N-1 contingency. The results of the Monte Carlo analysis indicate that the PI distribution is essentially bimodal (skewness = 0.5188, kurtosis = 3.2394) and the distribution is concentrated around 1.63 × 103 and 1.73 × 103, which prove that without any intervention, multi-bus faults tend to cluster into moderate and severe families. In Stage 2, the ideal positioning and size of the voltage support devices are formulated as a nonlinear integer program that is a mixed programming and is solved using a hybrid PSO-Bat algorithm. Exceptional consistency is shown by statistical analysis at 30 independent runs (coefficient of variation = 1.39%, 95% interval [ 1.2375 × 103,1.2642 × 103 ]) and Device 4 is an extremely consistent voltage support location (standard deviation = 6.24 bus indices). It is shown through Benchmarking that the conventional PSO has the lowest mean PI (1.2497 × 103), which is better than the hybrid by 13.93% (p < 0.01), whereas Genetic Algorithm is far quicker (23.49 s/run). The sensitivity analysis of weighting factors makes certain of the robustness and the change in PI of ± 54.8 between ± 50% changes in weight. The framework defines a statistically confirmed way of moving reactive N-1 security to proactive, intelligence-based resilience planning against multi-bus attacks coordinated.

DOI

10.30684/2412-0758.1555

First Page

1

Last Page

29

Download

Share

COinS