Dyn-WNTR

Dyn-WNTR is an extension of the WNTR framework(here), designed to enable real-time and dynamic simulations of water distribution networks (WDNs). This version allows for the continuous updating of water network models based on real-time data, providing a flexible and interactive platform for optimization, predictive modeling, and system management.

Key Features:

• Real-time Simulation: Unlike traditional simulators, Dyn-WNTR supports real-time updates, enabling continuous changes to valves, pumps, demands, and network components during the simulation.
• Integration with IoT and Digital Twins: Dyn-WNTR can integrate real-time IoT sensor data through technologies like LoRaWAN, improving the accuracy and adaptability of water network simulations. This makes it a suitable framework for use in digital twins of WDNs, allowing operators to optimize and predict network behavior in real-time.
• Dynamic Control and Analysis: Users can interact with the simulation, adjusting network parameters on the fly. This capability enables dynamic testing of various scenarios, improving decision-making and system resilience.
• Machine Learning Integration: Dyn-WNTR can work alongside machine learning models, using real-time data to improve predictions and system optimization.

Benefits:

• Optimized Resource Management: By enabling dynamic control and feedback, Dyn-WNTR helps in managing water resources more efficiently, adjusting parameters like pressure, flow, and demand in real-time.
• Improved System Resilience: Real-time updates and adaptability ensure that the system can respond to operational changes and potential issues as they arise.

Installation

To use Dyn-WNTR, you’ll need to clone this repo (anything useful is in the mwntr folder).

Example: Real-time Simulation with Dynamic Control

import mwntr

# Initialize the model and simulation
wn = wntr.network.WaterNetworkModel()
wn.add_reservoir('R1', base_head=100.0)
wn.add_junction('J1', base_demand=10.0)
wn.add_pipe('P1', start_node_name='R1', end_node_name='J1', length=100, diameter=0.3)

# Create a simulation object
sim = MWNTRInteractiveSimulator(wn)
sim.init_simulation()

# Modify the system dynamically
sim.start_leak('J1')  # Introduce a leak at J1
sim.step_sim()  # Run the simulation for one timestep
sim.stop_leak('J1')  # Stop the leak after one timestep

Contribution

We welcome contributions to improve and expand the functionality of Dyn-WNTR. Feel free to fork the repository, submit issues, or open pull requests with new features or bug fixes. Contributions can help us make the platform more scalable, user-friendly, and adaptable to real-world applications.

Future Work

1. Integration with Machine Learning:
   We plan to integrate reinforcement learning (RL) models for optimizing system management decisions, such as valve control and demand adjustment.

2. Expansion to Larger Networks:
   Additional work will focus on improving the scalability of Dyn-WNTR, ensuring it can efficiently simulate large-scale water distribution networks with real-time updates.

3. Digital Twin Integration:
   The future direction includes expanding Dyn-WNTR's ability to interface with real-world sensor data, enhancing the effectiveness of digital twins and predictive maintenance.

4. Merge into WNTR official repository The plan is to merge this fork into the original project to enhance the capabilities of WNTR itself.

License

Dyn-WNTR is open-source and distributed under the MIT License.