Traffic flow forecasting is the key to an intelligent transportation system (ITS). Currently, the short-term traffic flow forecasting methods based on deep learning need to be further improved in terms of accuracy and computational efficiency. Therefore, a short-term traffic flow forecasting model GA-TCN based on genetic algorithm (GA) optimized time convolutional neural network (TCN) is proposed in this paper. The prediction error was considered as the fitness value and the genetic algorithm was used to optimize the filters, kernel size, batch size, and dilations hyperparameters of the temporal convolutional neural network to determine the optimal fitness prediction model. Finally, the model was tested using the public dataset PEMS. The results showed that the average absolute error of the proposed GA-TCN decreased by 34.09%, 22.42%, and 26.33% compared with LSTM, GRU, and TCN in working days, while the average absolute error of the GA-TCN decreased by 24.42%, 2.33%, and 3.92% in weekend days, respectively. The results indicate that the model proposed in this paper has a better adaptability and higher prediction accuracy in short-term traffic flow forecasting compared with the existing models. The proposed model can provide important support for the formulation of a dynamic traffic control scheme.

To provide reliable traffic information and more convenient visual feedback to traffic managers and travelers, we proposed a prediction model that combines a neural network and a Macroscopic Fundamental Diagram (MFD) for predicting the traffic state of regional road networks over long periods. The method is broadly divided into the following steps. To obtain the current traffic state of the road network, the traffic state efficiency index formula proposed in this paper is used to derive it, and the MFD of the current state is drawn by using the classification of the design speed and free flow speed of the classified road. Then, based on the collected data from the monitoring stations and the weighting formula of the grade roads, the problem of insufficient measured data is solved. Meanwhile, the prediction performance of NARX, LSTM, and GRU is experimentally compared with traffic prediction, and it is found that NARX NN can predict long-term flow and the prediction performance is slightly better than both LSTM and GRU models. Afterward, the predicted data from the four stations were integrated based on the classified road weighting formula. Finally, according to the traffic state classification interval, the traffic state of the road network for the next day is obtained from the current MFD, the predicted traffic flow, and the corresponding speed. The results indicate that the combination of the NARX NN with the MFD is an effective attempt to predict and describe the long-term traffic state at the macroscopic level.

In this paper, we propose an improved cellular automaton model for the traffic operation characteristics of variable direction lanes in an Intelligent Vehicle Infrastructure Cooperation System (I-VICS). According to the proposed flow of variable oriented lane operation in the I-VICS environment, the idea for the improved model has been determined. According to an analysis of different signal states, an improved STCA model is proposed, in combination with the speed induction method of I-VICS and the variable direction lane switching strategy. In the assumed regular simulation environment, the STCA and STCA-V models are simulated under different vehicular densities. The results indicated that traffic parameters such as traffic flow and average speed of the variable direction lanes in the I-VICS environment are better than those in the conventional environment according to the operating rules of the proposed model. Moreover, lane utilization increased for the same density.

Traffic collisions are one of the leading causes of traffic congestion. In the case of urban intersections, traffic accidents can even result in widespread traffic paralysis. To solve this problem, we developed an autoencoder-based model for identifying intersection entrance accidents by analyzing the characteristics of traffic volume. The model uses the standard deviation of the intersection entrance lanes’ traffic volume as an input parameter and identifies intersection entrance accidents by comparing predicted data to actual measured data. In this paper, the detection rate and average detection time are chosen to evaluate the effectiveness of algorithms. The detection rate of the autoencoder model reaches 94.33%, 95.47%, and 81.64% during the morning peak, evening peak, and daylight off-peak periods, respectively. Compared to the support vector machine and the random forest, autoencoder has better performance. It is evident that the research presented in this paper can effectively enhance the detection effect and has a shorter detection time of intersection entrance accidents.