A Personnel Evacuation Model Considering Competitive Turning Preference

In recent years,the increase in crowd gathering scenes due to economic development and societal progress has resulted in a rise in incidents related to crowd safety.During emergency evacuations,the competitive behavior of individuals in selecting local movement directions directly impacts the micro-level evacuation process.Simultaneously,the selection of exits and path planning behaviors regulate the macroscopic movement state of evacuating crowds.This study focuses on a model research concerning speed-direction selection and exit selection in individual’s evacuation.Firstly,to address the limitations of constrained level of flexibility and blindly following model deadlock behaviors observed in the local direction selection during evacuations,this study introduces a speed model that incorporates competitive direction preference.The model considers pedestrians’ preferences for both forward distance and direction during movement.By applying the principle of conflict avoidance,the model calculates the forward distance between individuals and other pedestrians or obstacles.The Gaussian distribution is employed to capture individuals’ specific direction selection willingness,while a greedy strategy algorithm determines the direction decision-making.The optimal speed function is utilized to determine the appropriate speed magnitude.Additionally,the model introduces an elastic collision interval and deviation correction mechanism to regulate collisions and adjust speeds during competitive scenarios.This model addresses the issue of blindly following prevalent in collision-free models,enhancing the flexibility of direction selection.Simulated scenarios involving high competition bottlenecks demonstrate the model’s superior performance,achieving a flow rate error of 3.19%and individual distribution error prior to exit of 6.46%.These results showcase the model’s superiority compared to collisionfree models and social force models.Furthermore,to overcome the limited accuracy in exit selection resulting from inadequate evaluation of individual distribution characteristics in exit selection strategy,an experiment on exit selection behavior is designed focusing on exit selection behavior,considering factors such as personnel distribution.Additionally,an exit selection strategy model is developed,which incorporates personnel distribution characteristics.The model employs a compound field model to quantify the personnel distribution characteristics of the scene.Algorithms such as equivalent distance analysis and expected queueing time are integrated to predict the simulation time for exits and facilitate exit decision-making.Comparative simulations using experimental data under various working conditions demonstrate that the proposed exit selection model improves exit selection accuracy by an average of 19%compared to Liu’s exit queueing prediction model.In conclusion,a model research on speed-direction selection and exit selection is conducted in this study,with the following research significances identified:Firstly,algorithms for speed-direction selection and exit selection are proposed,enabling effective simulation and prediction of evacuations.Secondly,the reference value of the proposed elastic collision interval,deviation correction mechanism,and discrete field model for the construction and optimization of other models is emphasized.Lastly,corresponding experiments and simulations are included in the research,providing valuable data support for related studies on evacuation strategy and crowd behavior.