| With the rapid development of our country's economy and society and the increasing proportion of infrastructure investment,the freeway construction of our country and the development of related undertakings have achieved world-renowned achievements.With the increase in the number of motor vehicles and the continuous increase of residents' travel and transportation demands,the frequent occurrence of freeway traffic accidents has seriously threatened the economic development and people's life.However,for a long time,there are still many practical problems in the emergency support and rescue freeway in our country.The emergency mode is relatively simple,ignoring the scenario dependence of emergency plans,lack of cognition of the potential factors of emergencies,and failure to explore the heterogeneity of different accident scenarios,emergency decision-making lacks innovation in theory and technology.Therefore,the emergency rescue is not timely,the human and material support is insufficient,and the response and rescue organization and management are chaotic,which greatly affects the development of emergency rescue work.It is urgent to improve the emergency prevention mechanism and response level of freeway.This paper is based on solving the problem of low efficiency of emergency rescue on freeway,aiming at the traffic accidents that occur on freeway,on the basis of in-depth analysis of the potential factors of frequent accidents and the characteristics of different accident scenarios,from the emergency preparation before the accident and the accident after the accident.Starting from the perspective of emergency response,the key issues in the emergency decision-making system,such as the static allocation of emergency resources,the dynamic dispatch of emergency vehicles,and the optimization of dispatch paths,are studied respectively,aiming to improve the efficiency of emergency rescue and reduce emergency response in all aspects of emergency rescue.time.The main work and research results of this paper focus on the following four aspects:(1)Clustering and analyzing of freeway accident scenarios based on TSC-LC modelIn this paper,combined with the analysis method of situational emergency and traffic safety,the two step clustering(TSC)algorithm is employed to preprocess and automatically cluster the freeway accident data.The input reference of latent variable(LC)regression logistic model,and proposed four LC regression logistic models with their own advantages,namely LCMNL,LCMNP,LCOL and LCOP models.Based on the characteristics of LC models that are good at capturing heterogeneous utilities,we aim at analyzing the impact of explanatory variables on the severity of traffic accidents in each scenario category,exploring the heterogeneity between homogeneous categories,and the impact of unobservable factors on different categories,so as to provide a solid foundation for the subsequent allocation and scheduling of emergency resources.(2)Research on two-stage semi-static emergency resource allocation method of freeway based on SP-RRM2010 before the accidentTaking the emergency time as the axis,the accident as the node,and the situation clustering as the background,this paper adopts the relevant knowledge of regret theory from the psychological point of view that emergency decision-making is closer to the "avoidance" in "seeking advantages and avoiding disadvantages".On the basis of the accident black spots identified by the improved accident frequency method,a two-stage semi-static emergency resource allocation model of freeway based on SP-RRM2010 is proposed.This method not only considers the accident probability,but also considers the time attribute,cost attribute and the importance of attributes and demand attributes in emergency decision-making.In order to better estimate the proportion of these three attributes in the model,we actually investigated 119 related practitioners or researchers,and actively listened to the research and practical experience of front-line personnel to obtain a more realistic and reliable reference for modeling parameters,which provides strong material guarantee and risk management for possible future accidents.(3)Research on dynamic dispatch and reconfiguration method of emergency vehicles based on MDP after the accidentFrom the perspective of efficiency and fairness,this paper proposes a method for dynamic dispatch and reconfiguration of emergency vehicles for command and dispatch system decision makers.This method is a flexible optimization framework based on Markov decision process(MDP)model to reduce total response time and delay risk of the system,with the goal of improving the fairness of redistribution.In terms of efficiency goals,the improved BPR model is used to evaluate the reliability index of the current traffic conditions,and the travel time of emergency vehicles is reasonably corrected through the reliability index.According to the M/G/c queuing system theory,the vehicle waiting and scheduling time is approximated.In addition,due to the errors in the classification of call responses,the delay of response time has different effects on emergency needs of different severity,so a delay penalty for calls with different priorities is designed.For the fairness index,this paper divides fairness into vertical fairness and horizontal fairness.Vertical fairness emphasizes on-demand distribution,and horizontal fairness emphasizes that everyone has a share.The Gini coefficient is used to measure the fairness of reconfiguration.In this paper,the approximate dynamic programming(ADP)algorithm is used to solve the MDP model,and the upper bound based on the LAPI algorithm and the lower bound based on the linear programming are designed to deal with the infinite and unbounded state space in the MDP model,which can not only solve each decision process,but also ensure the quality of the optimal solution,which is especially suitable for realtime dynamic scheduling strategies.Finally,this paper uses the data of the Dutch Emergency Medical Service(EMS)system to dynamically dispatch and reconfigure ambulances,and compares the ADP strategy with the other four benchmark strategies.The results show that the emergency strategy proposed in this paper performs best in the aspects of all the four indicators,including response time,delay ratio,average risk level and Gini coefficient,which provides scientific support for theories and methods in the field of emergency rescue and response.(4)Research on optimization method of emergency vehicle dispatching route based on MIPSSTW modelAiming at accidents on freeway,a mixed integer linear programming(MIPSSTW)model with semi-soft time windows considering node consumption is proposed to optimize the route of emergency vehicle scheduling.The methods of estimating travel time for urban intersections and freeway in-ramp and on-ramp are introduced and solved using the Mutated Cuckoo(MCS)algorithm,enhancing the selection of the optimal solution through the mutation strategy used in each iteration.Finally,three cities in Flevoland,the Netherlands are selected as the research areas to verify the model.Not only the path selection method considering node consumption is compared with the method that not considering node consumption,but also the MCS algorithm is compared with CS algorithm,GA algorithm,PSO algorithm and ACO algorithm.The feasibility and effectiveness of the path optimization method and solution algorithm proposed in this chapter is proved to be better.The strategy provides a reasonable emergency scheduling path for decision executors.In this paper,aiming at the emergency decision-making system of freeway,using scientific,practical and balanced research ideas,from the perspectives of cluster analysis of freeway accident scenarios,emergency resource allocation,emergency vehicle scheduling and path planning,the progressive and innovative methods aim to enrich the theoretical support of the freeway emergency decision support systems are proposed,which improve the efficiency of emergency rescue response,ensure the safety of people's lives and property,and provide important academic value and practical significance for China's transportation industry and emergency field. |
PDF Full Text Request