00804675

Component

https://doi.org/10.3141/1739-10

http://scholar.google.com...ukry&publication_year=2000

A neural network modeling approach is used to identify concrete specimens that contain internal cracks. Different types of neural nets are used and their performance is evaluated. Correct classification of the signals received from a cracked specimen could be achieved with an accuracy of 75% for the test set and 95% for the training set. These recognition rates lead to the correct classification of all the individual test specimens. Although some neural net architectures may show high performance with a particular training data set, their results might be inconsistent. In situations in which the number of data sets is small, consistent performance of a neural network may be achieved by shuffling the training and testing data sets.

This paper appears in Transportation Research Record No. 1739, Evaluating Intelligent Transportation Systems, Advanced Traveler Information Systems, and Other Artificial Intelligence Applications.

English

p. 76-82

2000

Transportation Research Record

0309067421

Figures (7) ; References (12) ; Tables (4)

Accuracy; Concrete; Condition surveys; Neural networks; Performance evaluations; Specimens; Ultrasonic tests

Internal cracks; Signal classification; Ultrasonic signals

Highways; Materials; I32: Concrete

TRIS, TRB, ATRI

Jan 16 2001 12:00AM

• ARTIMIS TELEPHONE TRAVEL INFORMATION SERVICE: CURRENT USE PATTERNS AND USER SATISFACTION
• ASSESSMENT OF RURAL INTELLIGENT TRANSPORTATION SYSTEM WIRELESS COMMUNICATIONS SOLUTIONS
• BENEFIT-COST EVALUATION OF TRAVELER INFORMATION: SEATTLE'S WASHINGTON STATE DEPARTMENT OF TRANSPORTATION WEBSITE
• COMPARISON OF LOW-FIDELITY TRANSIMS AND HIGH-FIDELITY CORSIM HIGHWAY SIMULATION MODELS WITH INTELLIGENT TRANSPORTATION SYSTEM DATA
• DETERMINING THE OVERALL VALUE OF IMPLEMENTED NEW TECHNOLOGY IN TRANSPORTATION: INTEGRATED MULTIPLE OBJECTIVE-ATTRIBUTE METHODOLOGY
• DYNAMIC ESTIMATION OF ORIGIN-DESTINATION TRAVEL TIME AND FLOW ON A LONG FREEWAY CORRIDOR: NEURAL KALMAN FILTER
• ESTIMATING AN ORIGIN-DESTINATION TABLE UNDER REPEATED COUNTS OF IN-OUT VOLUMES AT HIGHWAY RAMPS: USE OF ARTIFICIAL NEURAL NETWORKS
• FAST EMERGENCY VEHICLE PREEMPTION SYSTEMS
• KNOWLEDGE ACQUISITION, VERIFICATION, AND VALIDATION IN AN EXPERT SYSTEM FOR IMPROVED TRAFFIC SAFETY
• LIKELY USERS OF ADVANCED TRAVELER INFORMATION SYSTEMS: EVIDENCE FROM THE SEATTLE REGION
• SEATTLE METROPOLITAN MODEL DEPLOYMENT INITIATIVE EVALUATION: RESULTS AND KEY FINDINGS FROM MODELING