Single loop detector data validation and imputation of missing data

The data derived from loop detectors are of great importance in terms of traffic monitoring and analysis. These data may contain many holes or incorrect values due to equipment malfunctions and communication faults that may produce unreliable results. These holes (missing samples) or incorrect values (bad samples) might be problematic for any algorithm that uses the data for analysis. In this paper, a method is described that detects bad data samples gathered by the loop detectors and imputes the best available samples in order to fill the holes caused by the bad declared samples. The diagnostics algorithm proposed in this paper is based on the statistical analysis. Unlike the previous approaches, this algorithm considers the time series of many samples, rather than basing decisions on single samples. The imputation algorithm proposed in this paper uses the “good” declared samples from the historical data of the investigated loop detector to fill the holes caused by the bad declared samples. This detection and imputation process allows the algorithms that use loop data to perform analysis without requiring them to compensate for missing or incorrect data samples.