Flow Direction Level Traffic Flow Prediction Based on a GCN-LSTM Combined Model

Traffic flow prediction plays an important role in intelligent transportation systems and is of great significance in the applications of traffic control and urban planning.Due to the complexity of road traffic flow data,traffic flow prediction has been one of the challenging tasks to fully exploit the spatiotemporal characteristics of roads to improve prediction accuracy.In this study,a combined flow direction level traffic flow prediction graph convolutional network(GCN)and long short-term memory(LSTM)model based on spatiotemporal characteristics is proposed.First,a GCN model is employed to capture the topological structure of the data graph and extract the spatial features of road networks.Additionally,due to the capability to handle long-term dependencies,the longterm memory is used to predict the time series of traffic flow and extract the time features.The proposed model is evaluated using real-world data,which are obtained from the intersection of Liuquan Road and Zhongrun Avenue in the Zibo High-Tech Zone of China.The results show that the developed combined GCNLSTM flow direction level traffic flow prediction model can perform better than the single models of the LSTM model and GCN model,and the combined ARIMA-LSTM model in traffic flow has a strong spatiotemporal correlation.

Modeling a Local Geoid: Application in Thies, Senegal

Many applications in geodesy, hydrography and engineering require determining heights linked to the geoid. Direct leveling, which is the traditional method of obtaining these elevations, is slow, time consuming and expensive. The contribution of space techniques can make it possible to overcome these constraints provided that we have a precision geoid model compatible with that obtained by the GNSS method. There are today relatively precise regional geoid models, at least outside of mountain ranges, in all developed countries, which is not yet the case in developing countries like Senegal. An alternative is to use local models restricted to a small area. Thus, this study aims to produce a geoid model by combining multi-source data for the city of Thies intended mainly to support leveling operations by GNSS. To achieve this objective, direct precision leveling and GNSS leveling (static mode) were carried out covering the study area. The reference points used are, among others, those of the RRS04 (Reference Network of Senegal 2004) and the NGAO53 (General Leveling of West Africa 1953). Additionally, gravimetric measurements were conducted using the Sensor Play-Data Recorder application. The calculation of the model was carried out by the SRBF (Spherical Radial Basis Function) method using the PAGravf4.5 software. The SRBF method uses EGM08 to first calculate height and gravity anomalies. These are then compared with the raw data in order to determine the residuals which will allow the model to be refined. In order to validate our model, control points (GNSS/leveled) were chosen based on a homogeneous geographical distribution in the area in order to evaluate their altitude. An accuracy of less than 2 cm was obtained. Comparing our model with the existing local model GGSV12v1 shows that our model is more accurate.