Prediction of Road Condition and Smoothness for Flexible and Rigid Pavements in Louisiana Using Neural Networks: Research Project Capsule [21–1P]

Details

• Creators:
  Wu, Zhong
• Corporate Creators:
  Louisiana Transportation Research Center
• Contributors:
  Rupnow, Tyson ; Zhang, Zhongjie
• Corporate Contributors:
  Louisiana. Department of Transportation and Development ; Louisiana State University (Baton Rouge, La.)
• Subject/TRT Terms:
  Flexible Pavements Mathematical Prediction Neural Networks Pavement Management Systems Rigid Pavements Smoothness
• Publication/ Report Number:
  21-1P
• Resource Type:
  Brief
• Geographical Coverage:
  Louisiana;United States
• Corporate Publisher:
  Louisiana Transportation Research Center
• Abstract:
  To achieve the objectives of this study, a neural network based system will be developed, using the data collected by DOTD on regular basis and local weather information. The proposed methodology can be summarized as follows. A review and documentation of the current state of practice by DOTD will evaluate short-term and long-term road condition and maintenance strategies. PMS data for the Louisiana pavements will be obtained from DOTD databases as well as meteorological condition data for the project sites from local agencies. These data will be categorized as per research scope and assessed through statistical measures. The team will also conduct statistical analyses to identify the significant variables from the historic pavement distress and condition data, pavement structure, traffic data, climatic data to predict the pavement distresses, and smoothness. Also, the set of variables that dictates the prediction of pavement condition will be used as an input to the ANN models. These variables can be different for different condition indices in the prediction models. ANN models will also be developed for each performance indicator index using software applications. The models will be validated using an existing dataset. The proposed ANN models will be presented in an implementation ready format through the development of a single windows-based interface that can accommodate all the ANN models. All findings and analyses of this study will then be documented in a detailed report.
• Format:
  PDF
• Funding:
  80/20 SPR
• Collection(s):
  US Transportation Collection
• Main Document Checksum:
  urn:sha-512:5f0c39a5867520716bd89b54574e4b49b662f0c53b2036cd173355f1e85034807ea79c1ab14f320fd240533f499b226a8f9d00f759946ce4bf90b12dfa23791e
• Download URL:
  https://rosap.ntl.bts.gov/view/dot/65607/dot_65607_DS1.pdf
• File Type:
  [PDF - 246.92 KB ]