Accelerating Integer Neural Networks On Low Cost Dsps

During the recent years, numerous utilities have experienced catastrophic rupture of critical Prestressing Concrete Cylinder Pipe (PCCP) lines throughout the world. Because PCCP is usually used throughout many water and wastewater utilities as critical mains with high flow rates, sudden failures can have significant negative consequences. Much attention has been focused on reliably assessing the condition of PCCP mains. Some techniques, such as visual inspections, electromagnetic inspections, acoustic monitoring and fiber-optic monitoring, are used to assess the condition of a pipeline now. Each of these techniques has capabilities and limitations that are important to understand when assessing the condition of a main. This can lead to the adoption of inadequate or over-conservative mitigation strategies. In this paper the combination of continuous acoustic monitoring and comprehensive dynamic risk management modeling is proposed. It provides an assessment of remaining time to failure for each pipe segment. This strategy can provide the opportunity to identify problematic pipe sections and repair the pipe prior to failure, and also can assess the presence and extent of deterioration in these large-diameter water and wastewater pipelines.

show abstract

A regeneratable dynamic differential evolution algorithm for neural networks with integer weights

Bao

Chen

2010

J. Zhejiang Univ. - Sci. C

View full text Add to dashboard Cite

Accelerating Integer Neural Networks On Low Cost Dsps

Cited by 3 publications

References 7 publications

A Framework of Web-GIS for PCCP Risk Management System

A Framework of Web-GIS for PCCP Risk Management System

The reseach on PCCP risk management based on wireless sensor network

A regeneratable dynamic differential evolution algorithm for neural networks with integer weights

Contact Info

Product

Resources

About