Interdependent Network Design Problem (INDP)

Description

This analysis takes a decentralized approach to solve the Interdependent Network Design Problem (INDP), a family of centralized Mixed-Integer Programming (MIP) models, which find the optimal restoration strategy of disrupted networked systems subject to budget and operational constraints.

Contributors

• Implementation: Hesam Talebiyan, Chen Wang, and NCSA IN-CORE Dev Team

Related publications

• Talebiyan, Hesam. “Interdependent Restoration of Infrastructure Networks with Humans in the Loop: decentralized and strategic decision processes.” (2021) Diss., Rice University. https://hdl.handle.net/1911/111232

Input parameters

key name	type	name	description
network_type *	str	Network Type	Type of the network, which is set to from_csv for Seaside networks. e.g. from_csv, incore.
MAGS *	list	MAGS	The earthquake return period.
sample_range *	range	Sample Range	The range of sample scenarios to be analyzed.
dislocation_data_type	str	Dislocation Data Type	Dislocation Data Type.
return_model	str	Return Model	Type of the model for the return of the dislocated population. Options: step_function and linear.
testbed_name	str	Testbed Name	Sets the name of the testbed in analysis.
extra_commodity *	dict	Extra Commodity	Multi-commodity parameters dictionary.
RC *	list	Resource Caps	List of resource caps or the number of available resources in each step of the analysis. Each item of the list is a dictionary whose items show the type of resource and the available number of that type of resource. For example: * If network_type=from_csv, you have two options:* if, for example, R_c= [{“budget”: 3}, {“budget”: 6}], then the analysis is done for the cases when there are 3 and 6 resources available of type “budget” (total resource assignment).* if, for example, R_c= [{“budget”: {1:1, 2:1}}, {“budget”: {1:1, 2:2}}, {“budget”: {1:3, 2:3}}] and given there are 2 layers, then the analysis is done for the case where each layer gets 1 resource of type “budget”, AND the case where layer 1 gets 1 and layer 2 gets 2 resources of type “budget”, AND the case where each layer gets 3 resources of type “budget” (Prescribed resource for each layer).
layers *	list	Layers	List of layers in the analysis.
method *	str	Method	There are two choices of method: 1. INDP: runs Interdependent Network Restoration Problem (INDP). 2. TDINDP: runs time-dependent INDP (td-INDP). In both cases, if “time_resource” is True, then the repair time for each element is considered in devising the restoration plans.
t_steps	int	Time steps	Number of time steps of the analysis.
time_resource	bool	Time Resource	If time resource is True, then the repair time for each element is considered in devising the restoration plans.
save_model	bool	Save Model	If the optimization model should be saved to file. The default is False.
solver_engine	str	Solver Engine	Solver to use for optimization model. Default to use SCIP solver.
solver_path	str	Solver Engine Path	Executable Path to the Solver to use for optimization model. Default to SCIP solver.
solver_time_limit	int	Solve Time Limit	Solving time limit in seconds.

Input datasets

key name	type	name	description
wf_repair_cost *	incore:repairCost	Water Facility Repair Cost	Repair cost for each water facility.
wf_restoration_time *	incore:waterFacilityRepairTime	Water Facility Repair Time	Repair time at certain functionality recovery for each class and limit state.
epf_repair_cost *	incore:repairCost	Electric Power Facility Repair Cost	Repair cost for each electric power facility.
epf_restoration_time *	incore:epfRepairTime	Electric Power Facility Repair Time	Repair time at certain functionality recovery for each class and limit state.
pipeline_repair_cost *	incore:pipelineRepairCost	Water Pipeline Repair Cost	Repair cost for each water pipeline.
pipeline_restoration_time *	incore:pipelineRestorationVer1	Water Pipeline Resotarting Time	Pipeline restoration times.
power_network *	incore:epnNetwork	Electric Power Network Dataset	Electric power network dataset.
water_network *	incore:waterNetwork	Water Network Dataset	Water network dataset.
powerline_supply_demand_info *	incore:powerLineSupplyDemandInfo	Powerline Supply Demand Info	Supply and demand information for powerlines.
epf_supply_demand_info *	incore:epfSupplyDemandInfo	Electric Power Facility Supply Demand Info	Supply and demand information for electric power facilities.
wf_supply_demand_info *	incore:waterFacilitySupplyDemandInfo	Water Facility Supply Demand Info	Supply and demand information for water facilities.
pipeline_supply_demand_info *	incore:pipelineSupplyDemandInfo	Water Pipeline Supply Demand Info	Supply and demand information for water pipelines.
interdep *	incore:interdep	Interdependency	Interdepenency between water and electric power facilities.
wf_failure_state *	incore:sampleFailureState	Water Facility Failure State	MCS failure state of water facilities.
wf_damage_state *	incore:sampleDamageState	Water Facility Damage State	MCS damage state of water facilities.
pipeline_failure_state *	incore:sampleFailureState	Water Pipeline Failure State	Failure state of pipeline from pipeline functionality analysis.
epf_failure_state *	incore:sampleFailureState	Electric Power Facility Failure State	MCS failure state of electric power facilities.
epf_damage_state *	incore:sampleDamageState	Electric Power Facility Damage State	MCS damage state of electric power facilities
pop_dislocation *	incore:popDislocation	Population Dislocation	Population dislocation.
dt_params	incore:dTParams	Dislocation time parameters	Parameters for population dislocation time.
bldgs2elec	incore:bldgs2elec	Building To Electric Power Facility	Relation between building and electric power facility.
bldgs2wter	incore:bldgs2wter	Building To Water Facility	Relation between building and water facility.

Output datasets

key name	type	parent key	name	description
action *	incore:indpAction		Action	Restoration action plans.
cost *	incore:indpCost		Cost	Restoration cost plans
runtime *	incore:indpRuntime		Run Time	Run time duration (in second) to execute computations for each time step

(* required)

Execution

code snippet:

    indp_analysis = INDP(client)
    indp_analysis.set_parameter("network_type", "from_csv")
    indp_analysis.set_parameter("MAGS", [1000])
    indp_analysis.set_parameter("sample_range", sample_range)
    indp_analysis.set_parameter("dislocation_data_type", "incore")
    indp_analysis.set_parameter("return_model", "step_function")
    indp_analysis.set_parameter("testbed_name", "seaside")
    indp_analysis.set_parameter("extra_commodity", {1: ["PW"], 3: []})
    indp_analysis.set_parameter("RC", [{"budget": 240000, "time": 700}, {"budget": 300000, "time": 600}])
    indp_analysis.set_parameter("layers", [1, 3])
    
    indp_analysis.set_parameter("method", "INDP")
    # indp_analysis.set_parameter("method", "TDINDP")
    
    indp_analysis.set_parameter("t_steps", 10)
    indp_analysis.set_parameter("time_resource", True)
    
    indp_analysis.set_parameter("save_model", False)
    # indp_analysis.set_parameter("save_model", True)
    
     # scip
    indp_analysis.set_parameter("solver_engine", "scip")
    indp_analysis.set_parameter("solver_path", "/usr/local/bin/scip")

    # glpk
    # indp_analysis.set_parameter("solver_engine", "glpk")
    # indp_analysis.set_parameter("solver_path", "/usr/local/bin/glpsol")

    # cbc
    # indp_analysis.set_parameter("solver_engine", "cbc")
    # indp_analysis.set_parameter("solver_path", "/usr/local/bin/cbc")

    # gurobi
    # indp_analysis.set_parameter("solver_engine", "gurobi")

    indp_analysis.set_parameter("solver_time_limit", 3600)  # if not set default to never timeout

    indp_analysis.set_input_dataset("wf_restoration_time", wf_restoration_time)
    indp_analysis.set_input_dataset("wf_repair_cost", wf_repair_cost_result)
    indp_analysis.set_input_dataset("epf_restoration_time", epf_restoration_time)
    indp_analysis.set_input_dataset("epf_repair_cost", epf_repair_cost_result)
    indp_analysis.set_input_dataset("pipeline_restoration_time", pipeline_restoration_time)
    indp_analysis.set_input_dataset("pipeline_repair_cost", pipeline_repair_cost_result)
    indp_analysis.set_input_dataset("power_network", power_network_dataset)
    indp_analysis.set_input_dataset("water_network", water_network_dataset)  # with distribution noes
    indp_analysis.load_remote_input_dataset("powerline_supply_demand_info", powerline_supply_demand_info_id)
    indp_analysis.load_remote_input_dataset("epf_supply_demand_info", epf_supply_demand_info_id)
    indp_analysis.load_remote_input_dataset("wf_supply_demand_info", wf_supply_demand_info_id)
    indp_analysis.load_remote_input_dataset("pipeline_supply_demand_info", pipeline_supply_demand_info_id)
    indp_analysis.load_remote_input_dataset("interdep", interdep_id)
    indp_analysis.set_input_dataset("wf_failure_state", wterfclty_sample_failure_state)
    indp_analysis.set_input_dataset("wf_damage_state", wterfclty_sample_damage_states)
    indp_analysis.set_input_dataset("pipeline_failure_state", pipeline_sample_failure_state)
    indp_analysis.set_input_dataset("epf_failure_state", epf_sample_failure_state)
    indp_analysis.set_input_dataset("epf_damage_state", epf_sample_damage_states)
    indp_analysis.set_input_dataset("pop_dislocation", pop_dislocation_result)

    # # optional inputs
    # indp_analysis.load_remote_input_dataset("bldgs2elec", bldgs2elec_id)
    # indp_analysis.load_remote_input_dataset("bldgs2wter", bldgs2wter_id)

    # Run Analysis
    indp_analysis.run_analysis()

full analysis: indp.ipynb