Abstract:Traffic monitoring agencies collect traffic data samples to estimate annual average daily traffic (AADT) at short duration count sites. The steps to estimate AADT from sample data introduce error that manifests as uncertainty in the AADT statistic and its applications. Past research suggests that the assignment of a short duration count site to a traffic pattern group (TPG), characterized by known traffic periodicities, represents a significant but poorly quantified source of error. This paper presents an appr… Show more
“…It was validated that this method's effectiveness was superior to that of traditional methods [17]. To reduce the error of estimating annual average daily traffic from sample data, experts such as G. Grande proposed a method to quantify the error range, and used a new data-driven allocation method to lessen the error, which could reduce the average absolute error by 2.46% [18]. Z. Liu and other researchers proposed a pattern recognition method based on image processing to reduce the incidence of highway traffic accidents.…”
Section: Related Workmentioning
confidence: 99%
“…The main function of the CNN model is to collect features from data, and it mainly includes convolutional, activated, pooling and fully connected layers [40][41]. Common activation function of CNN models include Sigmoid [42], ReLU [43] and tanh functions [44], and their calculations are shown in equation (18). 18), ϕ is the input variable.…”
Section: B Design Of Stop Point Recognition and Construction Of Dtprp...mentioning
To reduce traffic congestion, it is particularly important to use advanced technology to predict urban traffic flow. Therefore, a dynamic traffic pattern prediction model is proposed, which includes convolutional neural network, long and short term memory network and attention mechanism. The validity of the prediction model is verified by the loss function and the average absolute percentage error. In addition, the study also constructs a model for user travel pattern and parking point recognition based on deep learning and mobile signaling data. The performance of the recognition model is verified by the accuracy and other indicators. The research outcomes demonstrated that the max average absolute percentage error of the dynamic traffic mode prediction model was 7.8%, and the mini value was 2.9%. The average accuracy of the user travel pattern recognition model was 83.34%, and that of the parking point recognition model was 88.56%.The dynamic traffic model recognition and prediction model designed by the research institute has better results, and has practical guiding significance in smart city traffic management.
“…It was validated that this method's effectiveness was superior to that of traditional methods [17]. To reduce the error of estimating annual average daily traffic from sample data, experts such as G. Grande proposed a method to quantify the error range, and used a new data-driven allocation method to lessen the error, which could reduce the average absolute error by 2.46% [18]. Z. Liu and other researchers proposed a pattern recognition method based on image processing to reduce the incidence of highway traffic accidents.…”
Section: Related Workmentioning
confidence: 99%
“…The main function of the CNN model is to collect features from data, and it mainly includes convolutional, activated, pooling and fully connected layers [40][41]. Common activation function of CNN models include Sigmoid [42], ReLU [43] and tanh functions [44], and their calculations are shown in equation (18). 18), ϕ is the input variable.…”
Section: B Design Of Stop Point Recognition and Construction Of Dtprp...mentioning
To reduce traffic congestion, it is particularly important to use advanced technology to predict urban traffic flow. Therefore, a dynamic traffic pattern prediction model is proposed, which includes convolutional neural network, long and short term memory network and attention mechanism. The validity of the prediction model is verified by the loss function and the average absolute percentage error. In addition, the study also constructs a model for user travel pattern and parking point recognition based on deep learning and mobile signaling data. The performance of the recognition model is verified by the accuracy and other indicators. The research outcomes demonstrated that the max average absolute percentage error of the dynamic traffic mode prediction model was 7.8%, and the mini value was 2.9%. The average accuracy of the user travel pattern recognition model was 83.34%, and that of the parking point recognition model was 88.56%.The dynamic traffic model recognition and prediction model designed by the research institute has better results, and has practical guiding significance in smart city traffic management.
The widespread nature of cell phones and connected vehicle navigation systems has led to the development of commercially available probe-based traffic data products. This study assesses the accuracy of annual average daily total traffic, truck traffic, medium-duty truck traffic, and heavy-duty truck traffic volumes obtained using probe-based traffic activity indices from a North American company called StreetLight Data (StL). The probe-based estimates were compared with 2019, 2020, and 2021 volumes at eleven continuous count sites and 2019 volumes at twenty-nine short-duration count (SDC) sites in the Winnipeg Metropolitan Region. The results showed reasonable agreement between the ground truth and probe-based total traffic estimates with mean absolute percent errors (MAPEs) ranging from 8.8% to 22.1% across the study years. The medium-duty truck estimates had larger errors than total traffic with MAPEs of 29.9% to 37.5%. Despite having higher volumes than medium-duty trucks, heavy-duty trucks had the smallest probe data sample and largest errors with MAPEs of 56.6% to 96.4%. Benefiting from its larger sample size, the StL medium-duty truck index was found to be a better predictor of heavy-duty truck traffic than the heavy-duty truck index. Further, the total truck volumes estimated using only the medium-duty index were more accurate than those taken as the sum of the medium and heavy-duty truck volumes obtained using their respective indices. Finally, the percent differences for the 2019 annual average daily total traffic and truck traffic estimates at the SDC sites were comparable when only the medium-duty index was used for truck volume estimation.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.