Published: 8 September 2020 Abstract
This case study provides feasibility analysis of adapting Spiking Neural Networks (SNN) based Structural Health Monitoring (SHM) system to explore low-cost solution for inspection of structural health of damaged buildings which survived after natural disaster that is, earthquakes or similar activities.
Various techniques are used to detect the structural health status of a building for performance benchmarking, including different feature extraction methods and classification techniques (e.g., SNN, K-means and artificial neural network etc.).
The SNN is utilized to process the sensory data generated from full-scale seven-story reinforced concrete building to verify the classification performances. Results show that the proposed SNN hardware has high classification accuracy, reliability, longevity and low hardware area overhead.
1. Introduction
Earthquake is an oscillatory movement caused by the abrupt release of strain energy stored in the rocks within the crust of earth surface. Areas are always vulnerable to natural disasters, which can lead to extreme damages in nearby populations in terms of fatality, communication and infrastructure loss.
Flood, earthquake, cyclones and so forth, are among the most common occurring natural disasters across the world. The impact of these disasters differs in different geological and geographic locations.
These disasters come with no advance warning but an effective, well prepared and maintained infrastructure will decrease the potential impact of future disasters. The structural health of buildings and other infrastructure suffers degradation due to environmental catastrophes caused by ageing, hazards and natural disasters
[1]. In any area, public infrastructures, like schools, hospitals, fire stations, administrative buildings, bridges and treatment plants, are more prone to being highly affected by these disasters. Therefore, regular structural health monitoring is required to ensure the heath and endurance of these mega structures. In the event of a disaster, it is particularly important (i) to detect and quantify the severity of damage caused by environmental disasters at an early stage; (ii) to assess the current structural health and reliability of buildings to ensure their safe use; and (iii) to estimate repair costs for damage to minimize economic losses
[2]. Traditional monitoring methods rely on an inspection and assessment of the buildings and requires experienced inspectors. Many structures are not convenient for on-site monitoring due to the terrain obstacles, that is, the lack of access to such buildings, which sometimes make it too late due to the retrospective nature of inspections
[3]. An automated process such as installation of a Structural Health Monitoring (SHM) system for vulnerable structures, for example, buildings, bridges and even special launch vehicles, could periodically detect and notify of structural damages
[4]. An advance SHM system should include the current health profile of the structure, the functions of damage detection, structural life prediction and so forth
[5]. The lifespan of a typical structure lasts for decades whereas sensory instruments and microprocessors used by SHM systems come with a limited lifespan, for example, in an ideal operating environment the three-axis accelerometer of IIS3DHHC from the STMicroelectronics has a ten-year production life which shrinks in harsh outdoor environments. Therefore, after installation and regular use for several years, SHM systems may fatigue and fail. Due to technical and economic difficulties for secondary deployment, the longevity and reliability of SHM systems are key challenges that must be considered.
BRN Price at posting:
58.0¢ Sentiment: Buy Disclosure: Held
(20min delay)