Information-theoretic sensor placement for large sewer networks.

Area of Science:

• Environmental Engineering
• Network Analysis
• Data Acquisition

Background:

• Sewer network management requires proactive monitoring to prevent issues.
• Current methods for sensor placement are often suboptimal for large-scale networks.

Purpose of the Study:

• To develop a framework for optimized sensor placement in sewer networks.
• To maximize the information obtained about the network's state through sensor data.

Main Methods:

• Utilized mutual information to quantify sensor evidence.
• Developed a one-step modified greedy algorithm to address combinatorial placement challenges.
• Applied the algorithm to two real-world sewer networks in England and Denmark.

Main Results:

• The modified greedy algorithm demonstrated superior performance in estimating unmonitored locations compared to other heuristic methods.
• Sensor placements from the algorithm showed a tendency to cluster, enabling efficient network monitoring.
• The proposed framework provides a novel approach for designing data acquisition infrastructure for large sewer networks.

Conclusions:

• The developed framework and algorithm offer a computationally tractable solution for optimizing sensor placement in complex sewer systems.
• This method empowers utility operators to enhance sewer network monitoring and infrastructure design effectively.