Pedestrian Traffic Flow Research Based On Regional Potential-Field And Cost Potential-Field

After initial cellular automata model developed, a lot more mature model came about, of which a floor-field cellular automata model being effective one that simulated the pedestrian traffic flow. In the dissertation, pedestrian traffic flow cellular automata model is built combined with regional potential-field and cost potential-field based on floor-field cellular automata model. The basic rules of the model is established on common floor-field cellular automata model. The focus of dissertation is on calculating the probability of movement of cellular, thus choosing the optimized pedestrian path to get to the destination in shortest time.First of all, the dissertation briefly summarizes the floor field cellular automata model, then calculates the probability of movement and making update rules of cellular through establishing floor-field, thereby to guide the pedestrians. On the basis, the dissertation establishes difference scheme of Eikonal equation by using minimum time and path algorithm and leapfrog scheme, solve the equation, and then establishes the pedestrian regional potential-function-field. Secondly, according to change of the pedestrian density, pedestrian cost potential-field is established.The key part of the dissertation lies in calculating probability of movement combined with regional potential-function-field and pedestrian potential-function-field. Quantifies the probability of pedestrian movement in different condition, in consideration of pedestrians walking preference and traffic congestion, and determines the movement probability formula. Movement probability is the key of the update rules, one establish can update rules once confirming the probability. Resolving conflicts is the most to be considered, and the dissertation solves the problems in the movement by introducing variable. The modeling process mainly considers time and comfort of pedestrians, because pedestrians always choose the most comfortable walking way when meeting familiar paths, which can be measured by the change of the potential function. The potential function of a cellular automata model in the dissertation is to better solve the issue in the process.After establishing the potential-field cell automata model, the dissertation applies it into cases like single exit area, multi-exits area and the pedestrian area with obstacles. In each case, draws the corresponding formula of movement probability by changing formulas and parameters of model, changes the updated rules, and simulates the pedestrian traffic flow. Meanwhile, makes compares the model in the dissertation with the traditional floor-field cellular automata model. It concluded that in the same situation, the pedestrian path is more reasonable.Innovation points of the research are:1. Establishing difference scheme of Eikonal equation by using leapfrog scheme, solving the equation, improving the accuracy of truncation error, and making the model more accurate.2. Establishing the pedestrian regional potential-field model, quantifying the probability of pedestrian movement, and better programming the walking path.