Distributed Reinforcement Learning in Emergency Response Simulation

López, César A.; Martí, José; Sarkaria, Sarbjit

doi:10.1109/access.2018.2878894

Cited by 19 publications

(16 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Scenario 3 explores the changes in system performance when we increase the weight of CNs in the power system. We weigh the nodes 16,17,29,30, and 32 as the CNs (red marked nodes in Fig. 5) with ten times size larger than the normal ones.…”

Section: Results and Analysis Of Adaptation Ability Of The Casementioning

confidence: 99%

“…This section presents the modelling of the failure propagation based on i2SIM, whose layers and components are newly defined and developed. The i2SIM is a functional modelling approach that allows for capturing the interdependent interactions between different critical infrastructure systems [17]. Besides, based on the simulation process, some hidden interdependencies, i.e.…”

Section: Cps Resilience Evaluation Based On I2simmentioning

confidence: 99%

“…On the other hand, the weight of each path is defined opposite to how much changes (or improvements) can be gained, defined as ΔI RA , and its value is calculated through (16). Each status of the node availability is then calculated to obtain ΔSGC through updating the system from the current state to the next state (17).…”

Section: Decision Layer Designmentioning

confidence: 99%

“…Infrastructure interdependencies simulator (i2SIM) is a tool that provides the representation of interdependent infrastructures and modelling of their interdependencies. It has been widely used in many applications [13, 15, 16], and its suitability of the approach was previously tested in [17–19]. Specifically, Lopez et al [17] use i2SIM as the main simulation platform for the implementation of a coordinated decision‐making agent for emergency response scenarios.…”

Section: Introductionmentioning

confidence: 99%

“…It has been widely used in many applications [13, 15, 16], and its suitability of the approach was previously tested in [17–19]. Specifically, Lopez et al [17] use i2SIM as the main simulation platform for the implementation of a coordinated decision‐making agent for emergency response scenarios. In [19], learning agent using Monte Carlo estimation policy, through interactions with an environment of simulated disastrous scenarios using i2SIM is capable of making informed decisions for complex systems promptly.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Modelling cascading failure of a CPS for topological resilience enhancement

Yang

Chen

Martí³

2019

IET Smart Grid

View full text Add to dashboard Cite

This study focuses on the cyber-physical system (CPS) consisting of interdependent electrical distribution and communication networks, where the two networks are mutually dependent. A small disturbance in either of them can trigger a cascade of faults within the entire network. To investigate the failure mechanism, first, two features that affect topological resilience (TR) are defined in this study: adaptation and recovery abilities. Second, the authors model the process of cascading failures that occur in this coupled system by introducing and developing the infrastructure interdependencies simulator. The process of cascading failures is based on percolation theory, and they present a detailed analysis of cascading failure in a standard IEEE 33-bus system coupled with the 33-node communication system. This study proves that the adaptation ability of a coupled system is even lower than a single system. This is due to the interdependencies between systems, and the study of the failure mechanisms helps planers to make a better decision in the recovery process. Finally, the modified shortest path search is used to optimise the repair sequence. Their numerical results validate that the recovery ability of the coupled system is increased through the optimisation, which contributes to the TR enhancement.

show abstract

Section: Results and Analysis Of Adaptation Ability Of The Casementioning

confidence: 99%

Section: Cps Resilience Evaluation Based On I2simmentioning

confidence: 99%

Section: Decision Layer Designmentioning

confidence: 99%