Font Size: a A A

Research On Speed Management Scheme Of Eight-Lane Freeway

Posted on:2022-05-12Degree:MasterType:Thesis
Country:ChinaCandidate:H L YuFull Text:PDF
GTID:2492306740984099Subject:Traffic and Transportation Engineering
Abstract/Summary:PDF Full Text Request
Eight-lane freeway has considerable lanes,complex vehicle composition and good driving environment.Drivers can choose their driving speed more freely.These objective conditions will lead to greater speed dispersion between vehicles in some sections of the freeway and frequent vehicle lane changes,etc.circumstances,resulting in rear-end collisions and other traffic safety accidents,which not only poses a threat to the lives and property of road users,but also seriously reduces the efficiency of road operation.Therefore,from the two perspectives of comprehensive traffic efficiency and driving safety,the adoption of a scientific and reasonable speed management scheme for an eight-lane freeway can effectively reduce the possibility of traffic accidents caused by improper speed selection by drivers,which may enable the expressway to achieve both economic and social aspects.Firstly,starting from the basic concepts of different speeds,the relationship between speed management and traffic safety is analyzed,the necessity of speed limit measures on eight-lane expressways is clarified,and the advantages and disadvantages of different speed limit schemes are discussed.Based on the idea of first determining the mainline speed limit scheme,and then determining the per-lane speed limit scheme,the formulation process of the overall speed limit plan for the eight-lane freeway was determined.Analyzed the influencing factors of the main-line speed limit value,combined with the relevant speed management specifications and engineering technology demonstration method,put forward the formulation process of the main-line speed limit scheme and related implementation suggestions.Secondly,through outdoor aerial photography experiments,a macro analysis of the characteristics of the traffic volume and vehicle composition of the eight-lane freeway was carried out;using indoor driving simulation equipment,three lane-dividing speed limit schemes were constructed for experimentation,starting and stopping lanes,and the average duration from five perspectives: average longitudinal speed,average lateral acceleration,and maximum lateral acceleration,a microscopic analysis of the lane-changing behavior of eight-lane freeway drivers is carried out.The results show that the differentiated per-lane speed limit management will have a significant impact on the driver’s lane-changing decision(lane selection)and lane-changing execution(lane change duration,longitudinal average speed and lateral acceleration).Again,the deep belief network(DBN)is selected to construct the lane-changing decision model,and the combination of input layer,continuous restricted Boltzmann machine(CRBM),and BP neural network output layer is defined as the basic structure of the model.The 15lane-changing decision variables in a single set of input variables are defined,and data collection is realized through Houdini 3D software.The square root mean square error(RMSE)and the mean absolute error(MAE)are selected as the error evaluation indicators,and the number of hidden layers and the number of neurons in each hidden layer are determined by the layer-by-layer test method.The model was trained through 1400 sets of training data.After training,the remaining 592 sets of test data were used to test the prediction accuracy of the built DBN model.The prediction accuracy of the model is 93.92%,which can effectively predict the passenger car lane-changing behavior in eight-lane freeway.Then,a cellular automaton(CA)model for the traffic flow of an eight-lane freeway was constructed,and the improved comfortable driving model was used as the basic evolutionary rule of the CA model;the DBN lane-changing decision model was used to replace the lane-changing rules of the passenger car in the traditional CA model;The relationship between lane-changing motivation excitation and safety distance condition is used as the lane-changing rule for medium and large passenger cars and trucks,and the validity of the model is verified.The results show that the model can simulate the traffic flow of eight-lane freeway.Finally,the traffic flow scenarios with different truck proportion and traffic volume are constructed,and the CA model is used to carry out simulation experiments.The quantitative evaluation ideas of different differentiated per-lane speed limit scheme are formulated.Four evaluation indexes are selected as the output values of the model from the two aspects of traffic efficiency and traffic safety.The objective weighting method,critic method is used to determine the weight coefficient of each index,and select the linear comprehensive evaluation model to get the final evaluation value.The differentiated per-lane speed limit schemes under different traffic flow scenarios are compared and analyzed,moreover,relevant suggestions on the formulation of differentiated per-lane speed limit schemes are put forward.This paper puts forward a complete set of demonstration method of speed limit scheme for main line of eight-lane freeway and quantitative comparison and selection method of speed limit scheme by lane.The research results can provide reference for speed limit scheme formulation of newly built or expanded eight-lane freeway and above in the future,and provide theoretical basis for speed management measures of traffic management department,It can also be used as a supplement to the related theory and technology in the field of highway speed limit.
Keywords/Search Tags:eight-lane freeway, speed limit management, lane-changing decision, deep belief network, cellular automata
PDF Full Text Request
Related items