Research On Schedule Optimization And Early Warning Technology For Highway Transportation Of Hazardous Chemicals

With the rapid development of global economy, the chemical industry chain keeps extending and the amount of hazardous chemicals transportation keeps increasing. The hazardous chemicals are mainly transported by railway, ship and motor. Because of its convenience and efficiency, motor transport becomes the most principal mode. But motor transport costs highly and brings much unpredictable risk, so managing and controlling highway transportation of hazardous chemicals efficiently becomes a problem focused by the whole society. To this end, the thesis designs and develops a scheduling and security supervisory system on the background of enterprises’ actual demands in order to decrease the cost and risk of hazardous chemicals transportation combining schedule optimization and early warning technology, as a result, to improve transport efficiency. The main research contents of the thesis are as follows:(1) The framework of scheduling and monitoring management system for transportation of hazardous chemical is given. The key technologies of the system are described in detail.(2) A multi-objective dynamic vehicle routing problem(MODVRP) model and two-phase solving strategy are studied. Based on the analysis of traditional model, a new model is built synthesizing dynamic demands, the effects on the road network, vehicle sharing, time windows and customer satisfaction. Multi-objective hybrid particle swarm optimization(MOHPSO) is adopted to get preliminary Pareto solutions combining adaptive grid technique in the first phase. In real-time optimization phase, the paths are adjusted according to dynamic demands. A more reasonable multi-objective dynamic vehicle scheduling is realized and experimentally proven.(3) A fatigue recognition algorithm based on deep learning is studied. After the introduction of related concepts of deep learning, FRADL is designed specific to the features of fatigue recognition to extract fatigue features automatically layer by layer, and then the algorithm recognizes state of fatigue from video images based on time window. The experimental result shows that it can recognize the state of fatigue quickly, accurately with high individual adaptability.(4) The management system for highway transportation of hazardous chemicals is implemented. The main technologies are applied to the system. The design and development of the system are finished, and the running instances are given.