Authors

Abstract

To investigate the mechanism of the movement law of sheep and theoretically support the study on the herd effect, a herd effect model based on the agent and cellular automata technology is built. The law of the herd effect is defined with the use of PyCharm for simulation, based on the characteristics of the flock tending to the top sheep. The flock activity area falls into several two-dimensional cell space structures, and the grid with the cell state of "sheep" in the cell space structure is considered a type of agent. The model assumed that there are five behaviors, including standing, walking slowly, looking for the leader, random and avoiding collisions in four scenarios. A herd effect model is built, the herd behavior is simulated, and the simulation results are compared with the actual herd behavior trajectory. The mean square error between the calculation model and the reality is 1.025. As revealed by the results, the model exhibits effective applicability, so it can better describe the trajectory of real herd behavior and provide theoretical guidance for the study on the herd effect.

Abstract in Chinese

为揭示羊群的运动规律机制，能够为羊群效应的研究奠定理论依据，构建了基于智能体和元胞自动机技术的羊群效应模型。利用PyCharm进行仿真，基于羊群趋向头羊的特点,定义了羊群效应规律，把羊群活动区域分割为若干个二维的元胞空间结构，把元胞空间结构中元胞状态为"羊"的网格看成是一种智能体。模型假设了四种情景下存在站立、慢走、寻找头羊、随机以及避免碰撞五种行为。构建羊群效应模型，对羊群的从众行为进行模拟仿真，并将仿真结果与现实羊群行为轨迹加以比较拟合，计算模型和现实的均方误差为1.025，结果显示该模型具备了有效的适用性，从而可以较好地描述现实羊群行为轨迹，为羊群效应研究提供理论指导。