Research On Pedestrian Flow Simulation And Control Based On Directional Fuzzy Visual Field

Pedestrian flow is the distribution state of pedestrians in time and space,which transforms the location selection of many micro individuals into the macro aggregation phenomenon in a specific area.Pedestrian flow system is a dynamic and complex giant system that composed of pedestrians,building,pedestrian facilities,obstacles and other elements.Under the action of various static and dynamic factors,the system exhibits the characteristics of systematicness,dynamics,heterogeneity,and nonlinearity,and shows various complex macroscopic self-organization phenomena.In this thesis,the concept of Directional Fuzzy Visual Field(DFVF)is put forward from the perspective of pedestrian heterogeneity,which studies the difference of pedestrian’s action field in detail.Based on DFVF,the law of motion and the evolution characteristics of uni-directional,bi-directional and cross pedestrian flow are mainly studied.The main work of this thesis is reflected in the following aspects:(1)Considering the difference of pedestrian’s action field,the judgment model of DFVF is constructed in order to simulate the movement characteristics of pedestrian flow more accurately and refinedly.Different from the undifferentiated action field in the traditional floor field model,the DFVF fully reflects the heterogeneity of pedestrian’s action field selection,which provides an improved basis for the refined simulation of pedestrian flow.(2)Based on DFVF,a uni-directional pedestrian flow floor field model is established,which simulates the movement process of uniformly and asymmetrically distributed pedestrian flow in single-exit and multi-exit buildings.Simulation results show that the size of DFVF can significantly affect the moving efficiency of pedestrian flow system,and pedestrians are keeping a large DFVF can effectively reduce the evacuation time of pedestrian flow.These results are more obvious in the scenario of pedestrian flow with asymmetrical initial distribution.(3)A bi-directional pedestrian flow floor field model is established,the fuzzy distribution of pedestrian personal space and information processing space is defined by the concept of DFVF,which refined the action field of bi-directional pedestrian in detail.By introducing the k nearest neighbor rules in DFVF,The network topology of pedestrians’ interaction in personal space and information processing space is defined,which sets the interaction between pedestrians is not only dependent on physical distance,but also affected by the topological distance between groups.These rules make bi-directional pedestrian flow simulation more consistent with real pedestrians’ interaction.The movement characteristics and self-organizing phenomenon of bi-directional pedestrian flow with different distributions are simulated,and the influences of pedestrian density on the average velocity,system flow and evacuation time is analyzed.(4)Based on the foregoing results of the simulation analysis,an improved cross pedestrian flow floor field model is constructed.The stratification and bottleneck swing of pedestrian’s self-organizing phenomenon in the cross area is simulated,the variation process of pedestrian density,average velocity and average flow in the cross area is analyzed.The study found that the movement efficiency of cross pedestrian flow can be effectively improved by controlling the pedestrian conflict in the crossing area.(5)Considering the differences of pedestrian movement strategy choice,the evolutionary game model of pedestrians’ cooperation and competition strategy is established,which defines the pedestrians’ strategy evolution rules in DFVF based on the theory of the prisoner’s dilemma.The influence of different radius of DFVF,strategies adjusting income,pedestrian density,system scale and other factors on the strategy choice of pedestrian flow are analyzed,that provides theoretical support for the refinement improvement of the movement rules in floor field model.It is found that pedestrians have the larger the radius of DFVF;they are the more likely to choose competitive strategies.When pedestrians maintain a large visual field in the process of movement,the system should limit the advantages of competitive strategy,so as to achieve a dynamic balance between pedestrians’ cooperation strategy and competitive strategy,make pedestrian flow to keep the diversity of movement strategies,and effectively control the competitive pressure in the process of pedestrian flow movementIn this thesis,the difference of pedestrian action field selection is analyzed in depth,and the heterogeneity of pedestrian is described by the definition of DFVF.On the basis of DFVF,the floor field models of uni-directional pedestrian flow,bi-directional pedestrian flow and cross pedestrian flow are established respectively.This thesis makes an in-depth study on the movement law of pedestrian flow in different scenarios,and analyzes the static and dynamic influencing factors from multiple angles,which can provide certain theoretical support for the formulation of the control strategy of pedestrian flow.