Adapting the Social Force Model for Low Density Crowds in Open Environments

TheSocialForce Model (SFM) successfully reproduces many collective phenomena in evacuations or dense crowds. However, pedestrians behaviour is context dependent and the SFM has some limitations when simulating crowds an open environment under normal conditions. Specifically, urban public square tend to expand their personal space try avoid areas reduce risk of collision. Based on SFM, proposed model splits perception into a large zone restricted frontal which they pay more attention. Through perceptions, agents estimate crowd density dynamically adapt space. Finally, original social force tuned reflect preference avoiding by turning rather than slowing down as long there enough Simulation results show that considered approach produces realistic behaviours SFM. The simulated less with same number collisions occur, better fits observations sparse place