Detecting and mitigating cascading errors in a network to improve network resilience
In an embodiment, a computer implemented method is provided. The method may include quantifying a plurality of component level risks for at least a subset of components in the network. The method may further include simulating cascades of the component level risks, with each corresponding component designated as a risk seed of the subset of components, throughout the network. The method may additionally include quantifying the network level risk as a risk status in a resilience spectrum based on the simulated cascades.
1. A system comprising:
a non-transitory storage medium storing computer program instructions; and
one or more processors configured to execute the computer program instructions to cause the system to perform operations comprising:
quantifying a plurality of component level risks for at least a subset of components in a network;
simulating cascades of the component level risks, with each corresponding component designated as a risk seed of the subset of components, throughout the network;
quantifying a network level risk as a risk status in a resilience spectrum based on the simulated cascades by:
fitting a least-squares line to the component level risks and a fractal dimension based on corresponding exceedance probabilities derived from the simulated cascades to generate at least two risk parameters;
calculating, based on the generated at least two risk parameters, a tipping point within the resilience spectrum, wherein first portion of the spectrum at one side of the tipping point indicates that the network is resilient to cascading failures, and wherein a second portion of the spectrum at the other side of the tipping point indicates that the network is non-resilient to cascading failures; and
calculating a resilience reserve as a factor of safety between the risk status and the tipping point; and
displaying the risk status, the resilience reserve, and the tipping point within the resilience spectrum on a graphical user interface, such that the displayed information may be used to mitigate the network level risk.
2. The system of claim 1 , the subset of component comprising a plurality of network nodes and a plurality of network links.
3. The system of claim 1 , the network comprising at least one of water distribution network, pump driven stormwater and wastewater network, electrical power distribution network, electrical transmission network, supervisory control and data acquisition (SCADA) system, telecommunication network, transportation network, computer network, building structure, socio-technical network, supply chain network, internet of things (IoT) network, or human managerial network.
4. The system of claim 1 ,
the network simulating cascades of component level risks comprising simulating the cascades throughout the network, the network being interdependent; and
quantifying the network level risk comprising quantifying the interdependent network level risk.
5. The system of claim 1 , quantifying the network level risk further comprising:
determining a spectral radius of the network, wherein the spectral radius comprises the largest eigenvalue of a connection matrix representing the network.
6. The system of claim 1 , quantifying the network level risk further comprising:
calculating a network failure probability based on a component level risk.
7. The system of claim 1 , quantifying the network level risk further comprising:
calculating a corresponding exceedance probability indicating that a loss exceeds a predetermined value based on a component level risk.
8. The system of claim 1 , quantifying the network level risk comprising:
calculating a maximum probable loss based on a component level risk.
9. The system of claim 1 , the subset of components comprising a plurality of network nodes, and quantifying the network level risk comprising:
generating a list of blocking network nodes from the plurality of network nodes.
10. The system of claim 1 , the subset of components comprising a plurality of network links, and quantifying the network level risk comprising:
generating a list of blocking network links from the plurality of network links.
11. The system of claim 1 , quantifying the network level risk comprises:
calculating a cost for mitigating the network level risk.
12. The system of claim 1 , the operations further comprising:
determining a mitigation measure to reduce the network level risk.
13. The system of claim 12 , the subset of components comprising a plurality of network links, and the mitigation measure comprising rewiring of at least one of the plurality of network links.
14. The system of claim 12 , the subset of components comprising a plurality of network nodes, and the mitigation measure comprising at least one of:
splitting at least one network node of the plurality of network nodes; or
combining at least two network nodes of the plurality of network nodes.
15. The system of claim 12 , the mitigation measure comprising:
reducing vulnerability of at least one component of the subset of components.
16. The system of claim 12 , the mitigation measure comprising:
reducing a consequence of at least one component of the subset of components.
17. The system of claim 12 , the mitigation measure comprising:
improving at least one of a recovery cost, recovery time, or recovery order.
18. A non-transitory storage medium storing computer program instructions, which when executed cause operations comprising:
quantifying a plurality of component level risks for at least a subset of components in a network;
simulating cascades of the component level risks, with each corresponding component designated as a risk seed of the subset of components, throughout the network;
quantifying a network level risk as a risk status in a resilience spectrum based on the simulated cascades by:
fitting a least-squares line to the component level risks and a fractal dimension based on corresponding exceedance probabilities derived from the simulated cascades to generate at least two risk parameters;
calculating, based on the generated at least two risk parameters, a tipping point within the resilience spectrum, wherein first portion of the spectrum at one side of the tipping point indicates that the network is resilient to cascading failures, and wherein a second portion of the spectrum at the other side of the tipping point indicates that the network is non-resilient to cascading failures; and
calculating a resilience reserve as a factor of safety between the risk status and the tipping point; and
displaying the risk status, the resilience reserve, and the tipping point within the resilience spectrum on a graphical user interface, such that the displayed information may be used to mitigate the network level risk.
19. The non-transitory storage medium of claim 18 , the subset of component comprising a plurality of network nodes and a plurality of network links.
20. The non-transitory storage medium of claim 18 , the network comprising at least one of water distribution network, pump driven stormwater and wastewater network, electrical power distribution network, electrical transmission network, supervisory control and data acquisition (SCADA) system, telecommunication network, transportation network, computer network, building structure, socio-technical network, supply chain network, internet of things (IoT) network, or human managerial network.