作者: Stephen G Ritchie , Mohamed Kaseko , Behnam Bavarian
DOI:
关键词: Data acquisition 、 Learning rule 、 Pavement management 、 Process (engineering) 、 Artificial intelligence 、 Systems engineering 、 Artificial neural network 、 Expert system 、 Engineering 、 Automation 、 Backpropagation
摘要: A potential automated pavement evaluation system to address multisensor applications; integrate different types of sensors, techniques, and information; offer more sophisticated intelligent processing capabilities for improved management is described. The separate components this either now exist in prototype form or are under development. Such a could automate real time much the data acquisition, interpretation, process, capture experience judgment expert engineers performing condition assessments identification appropriate rehabilitation maintenance strategies. This research directed toward an innovative, noncontact, nondestructive (INDE) system, using novel artificial intelligence (AI)-based approach that would three AI technologies: computer vision, neural networks, knowledge-based systems, addition conventional algorithmic modeling techniques. focus current, initial development advanced sensor capability network technology determine type, severity, extent distresses from digitized video image representations surface acquired time. properties networks provide solutions inherently difficult nature integration output interpretation evaluation. background conceptual INDE evaluation, results demonstrate feasibility case study application multilayer perception backpropagation learning rule,