Accurate and reliable traffic flow prediction is critical to the safe and stable deployment of intelligent transportation systems. However, it is very challenging since the complex spatial and temporal dependence of traffic flows. Most existing works require the information of the traffic network structure and human intervention to model the spatial-temporal association of traffic data, resulting in low generality of the model and unsatisfactory prediction performance. In this paper, we propose a general spatial-temporal graph attention based dynamic graph convolutional network (GAGCN) model to predict traffic flow. GAGCN uses the graph attention networks to extract the spatial associations among nodes hidden in the traffic feature data automatically which can be dynamically adjusted over time. And then the graph convolution network is adjusted based on the spatial associations to extract the spatial features of the road network. Notably, the information of rode network structure and human intervention are not required in GAGCN. The forecasting accuracy and the generality are evaluated with two real-world traffic datasets. Experimental results indicate that our GAGCN surpasses the state-of-the-art baselines on one of two datasets.
For a honeycomb structure used for absorbing crash energy and protecting the safety of human or instruments, the bigger the specific energy absorption (SEA) is, the more popular it would be when the peak crushing stress (σp) was retained small enough. In order to improve the energy absorption capacity, crashworthiness optimization for honeycomb structures with various cell specifications are studied in this paper. Detailed numerical models are established for those honeycomb structures by using an explicit finite element method code LS-DYNA. The numerical simulation results are then used as the design samples for constructing metamodels. The optimal Latin hypercube design (OLHD) method is employed for the selection of sampling design points in the design space, and the polynomial functions, radial basis functions (RBF), Kriging, multivariate adaptive regression splines (MARS), and support vector regression (SVR) are utilized to formulate the two optimal objectives SEA and σp. It is found that the polynomial function is the most efficient in constructing the crashworthiness metamodels of honeycombs among the above-mentioned methods. Then, the polynomial function models of SEA and σp are chosen as the surrogate models in the crashworthiness optimization. In order to further validate the polynomial function models, the polynomial function models of SEA and σp are compared with the analytical solutions based on Wierzbicki's theory and Kunimoto and Yamada's theory, respectively. An excellent correlation has been established. As such, the multi-objective particle swarm optimization algorithm (MOPSOA) is applied to obtain the Pareto front of SEA with σp of the honeycomb structures with various cell specifications, which has resulted in a range of optimal designs of honeycomb structures by the multi-objective optimization.
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.