Density-based evolutionary framework for crowd model calibration

Jinghui Zhong, Nan Hu, Wentong Cai, Michael Lees, Linbo Luo

Abstract: Highlights•We propose an evolutionary framework to calibrate crowd models, so that the simulated crowd behaviors can closely match the objective behaviors.•We introduce a density-based matching scheme to automatically evaluate the simulated crowd behaviors at a macroscopic level.•We propose a hybrid search mechanism based on differential evolution to reduce the computational time.•We design different simulation scenarios to test our framework, and the results demonstrate that our algorithm is effective and efficient for crowd model calibration.