Optimisation of Sensor and Sensor Node Positions for Shape Sensing with a Wireless Sensor Network-A Case Study Using...

Area of Science:

• Structural Engineering
• Sensor Networks
• Machine Learning

Background:

• Operational condition data from structural components is crucial for product optimization through load adaptation.
• Optimal sensor placement is essential for high-quality shape and load sensing, especially in large-scale structures.
• Existing research often addresses sensor placement or node positioning separately, not holistically.

Purpose of the Study:

• To develop a unified methodology for optimal sensor placement and wireless sensor network (WSN) node positioning.
• To implement this methodology in a Python tool for practical application.
• To demonstrate the effectiveness of the optimized WSN for real-time shape sensing on a test component.

Main Methods:

• Application of the modal method for shape sensing.
• Utilization of a physics-informed neural network (PINN) for inverse problems in shape sensing (iPINN).
• Implementation of a WSN using strain gauges, HX711 A/D converters, and Arduino Nano 33 IoT microprocessors, with data processing on a Python Flask server.

Main Results:

• Successful realization of an optimized WSN on a demonstration part under test bench loading.
• Demonstration of high-accuracy shape sensing capabilities enabled by the integrated methodology.
• Validation of the Python tool's applicability for optimizing sensor placement and WSN configuration.

Conclusions:

• The presented methodology and Python tool effectively integrate optimal sensor placement and WSN node positioning.
• The developed system achieves high-accuracy shape sensing, crucial for structural health monitoring and product optimization.
• This approach provides a valuable framework for real-time monitoring and data processing of structural components.