Mixed-Integer Optimization for Bio-Inspired Robust Power Network Design

Huang, Hao; Panyam, Varuneswara; Narimani, Mohammad Rasoul; Layton, Astrid; Davis, Katherine

doi:10.1109/naps50074.2021.9449755

Cited by 6 publications

(14 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Unlike the optimization problem in [20,21] that searches all possible network structures, the MINLP approach defines the possible connections to improve the efficiency while sacrificing a globally-optimal solution. Comparing the results in [24] and [21], the MINLP solution improves the network reliability under N − 1 contingency and solves significantly faster. However, the optimization formulation in [24] fails to capture the feasible space, since Equations 1-3 involve the logarithm function (whose inputs Figure 2: Reliability improvement for bio-inspired networks and original networks under N-1, N-2 and N-3 contingencies.…”

Section: Related Workmentioning

confidence: 96%

Section: Related Workmentioning

confidence: 99%

“…A promising approach preliminarily investigated in [24] was to use PowerModels.jl [25] to combine the bio-inspired power network problem with its network expansion problem. The proposed MINLP in [24] integrates the power flow constraints with ecological robustness as the objective function. Unlike the optimization problem in [20,21] that searches all possible network structures, the MINLP approach defines the possible connections to improve the efficiency while sacrificing a globally-optimal solution.…”

Section: Related Workmentioning

confidence: 99%

“…Ecological robustness has further been improved in grid studies while considering power system constraints by modeling the problem as a mixed-integer nonlinear programming (MINLP) problem that solves for a bioinspired robust power network design [24]. A binary decision variable was introduced for candidate branches, so that the ecological robustness was dependent on binary decision variables as well as the power flow distribution.…”

Section: Introductionmentioning

confidence: 99%

“…However, the formulation of ecological robustness associates with logarithm functions, which requires the entries' domain to be positive. With the integration of power flow models, the MINLP problem in [24] is a non-convex, nonlinear problem. Thus, the previous MINLP was unable to guarantee feasible solutions for large power system cases.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

An Ecological Robustness-Oriented Approach for Power System Network Expansion

Huang

Mao²,

Panyam³

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

Electric power grids are critical infrastructure that support modern society by supplying electric energy to critical infrastructure systems. Incidents are increasing that range from natural disasters to cyber attacks. These incidents threaten the reliability of power systems and create disturbances that affect the whole society. While existing standards and technologies are being applied to proactively improve power system reliability and resilience, there are still widespread electricity outages that cause billions of dollars in economic loss annually and threaten societal function and safety. Improving resilience in preparation for such events warrants strategic network design to harden the system. This paper presents an approach to strengthen power system security and reliability against disturbances by expanding the network structure from an ecosystems perspective.Ecosystems have survived a wide range of disturbances over a long time period, and an ecosystem's robust structure has been identified as the key element for its survivability. In this paper, we first present a study of the correlation of ecological robustness and power system structures. Then, we present a mixed-integer nonlinear programming problem (MINLP) that expands the transmission network structure to maximize ecological robustness with power system constraints for an improved ability to absorb disturbances. We solve the MINLP problem for the IEEE 24 Bus Reliability Test System and three synthetic power grids with 200-, 500-and 2000-buses, respectively. Our evaluation results show the optimized power systems have increased the network's robustness, more equally distributed power flows, and less violations under different levels of contingencies.

show abstract