Dec2py: Decentralized Decision-making tools

Decision support tools for the interdependent network restoration to enhance resilience

Note

This webpage contains code documentation and sample jupyter notebooks for an ongoing project. The documentation is constantly getting updated and revised, and the codes will be added to the website soon.

This website offers code documentation and sample jupyter notebooks for a set of analytic tools that model the restoration processes in synthetic and real-world interdependent networks subject to external shock and disruption. The codes are developed using Python language. Various restoration decision-making models are considered here:

• Centralized methods

  These methods solve one optimization problem for the whole interdependent network, which leads to optimal restoration plans. The implication of such models is that the decision-maker is one entity who has complete information and authority to restore all layers of the interdependent network. This kind of methods includes Interdependent Network Design Problem (INDP) [GDOSSM16] and time-dependent INDP (td-INDP) [GDOMSS16].

• Decentralized methods

  These methods model a multi-agent decision-making environment where each agent has the authority to restore a single layer, has complete information about her respective layer, and minimal or no information about other layers. It is assumed that agents communicate poorly and not on time. The decentralized methods include Judgment Call (JC) method [TDO20][TDO19] with and without auction-based resource allocations [TDO21b] and Normal-form Interdependent Network Restoration Games (N-INRG) and Bayesian Games (B-INRG) [TDO21a].

Below, you can find different parts of the website (topics), starting with demonstrating tools for an interdependent toy network followed by a notebook containing the complete analysis dashboard. The third section presents functions and classes under the hood:

Topics

• Interdepndent Network Restoration Decision-making (toy example)
• Interdepndent Network Restoration Decision-making (Complete Analysis Dashboard)
  • Run a toy example using different methods
  • Run different methods for different networks, post-process the outputs, and plot the result
• Backend Classes and Modules
  • High-level analysis modules
  • Infrastructure
  • INDP
  • Judgment Call
  • N-INRG and B-INRG

Indices and tables

• Index

• Module Index

• Search Page

To-do works

Todo

Games: refine how to reduce importance of action not consistent with the player’s type.

(The original entry is located in C:\Users\ht20\Documents\GitHub\Dec2py\codes\gameclasses.py:docstring of gameclasses.BayesianGame.set_states, line 5.)

Todo

Games: Expand the code to imitate td-INDP

(The original entry is located in C:\Users\ht20\Documents\GitHub\Dec2py\codes\gameclasses.py:docstring of gameclasses.InfrastructureGame.set_time_steps, line 4.)

Todo

Games: refine the way the game solution is chosen in each time step

(The original entry is located in C:\Users\ht20\Documents\GitHub\Dec2py\codes\gameclasses.py:docstring of gameclasses.NormalGame, line 97.)

Todo

Games: Add the geographical interdependency to find actions, which means to consider the arc repaires as indepndent actions rather than aggregate them in the ‘OA’ action.

(The original entry is located in C:\Users\ht20\Documents\GitHub\Dec2py\codes\gameclasses.py:docstring of gameclasses.NormalGame.find_actions, line 3.)

Todo

Games: Add the geographical interdependency to computation of payoff values

(The original entry is located in C:\Users\ht20\Documents\GitHub\Dec2py\codes\gameclasses.py:docstring of gameclasses.NormalGame.flow_problem, line 29.)

Todo

The calculated repair time and costs are written to node and arc info input files. It makes it impossible to run the analyses in parallel because there might be a conflict between two processes. Consider correcting this.

(The original entry is located in C:\Users\ht20\Documents\GitHub\Dec2py\codes\indp.py:docstring of indp.time_resource_usage_curves, line 4.)

Bibliography

GDOSSM16

Andres D. Gonzalez, Leonardo Duenas-Osorio, Mauricio Sanchez-Silva, and Andres L. Medaglia. The Interdependent Network Design Problem for Optimal Infrastructure System Restoration. Computer-Aided Civil and Infrastructure Engineering, 31(5):334–350, may 2016. URL: http://doi.wiley.com/10.1111/mice.12171, doi:10.1111/mice.12171.

GDOMSS16

Andres David Gonzalez, Leonardo Duenas-Osorio, Andres L. Medaglia, and Mauricio Sanchez-Silva. The time-dependent interdependent network design problem (td-INDP) and the evaluation of multi-system recovery strategies in polynomial time. In The 6th Asian-Pacific Symposium on Structural Reliability and its Applications, 544–550. Shanghai, China, 2016.

TDO19

Hesam Talebiyan and Leonardo Duenas-Osorio. Probabilistic Assessment of Decentralized Decision-making for Interdependent Network Restoration. In Junho Song, editor, 13th International Conference on Applications of Statistics and Probability in Civil Engineering, ICASP13. Seoul, South Korea, 2019. doi:10.22725/ICASP13.215.

TDO20

Hesam Talebiyan and Leonardo Duenas-Osorio. Decentralized Decision Making for the Restoration of Interdependent Networks. ASCE-ASME Journal of Risk and Uncertainty in Engineering Systems, Part A: Civil Engineering, 6(2):04020012, jun 2020. URL: http://ascelibrary.org/doi/10.1061/AJRUA6.0001035, doi:10.1061/AJRUA6.0001035.

TDO21a

Hesam Talebiyan and Leonardo Duenas-Osorio. Interdependent Network Restoration Games: Simultaneous and Bayesian. working paper, 2021.

TDO21b

Hesam Talebiyan and Leonardo Duenas-Osorio. Resource Allocation for Restoration of Interdependent Networks Using Auction Theory. Reliability Engineering and System Safety (under review), 2021.

Wu17

Jason Wu. End-to-end seismic risk analysis framework for the identification of infrastructure network retrofits. Ph.D. Dissertation, Stanford University, Stanford, California, 2017.