基于自适应调节的人类动力学模型

Human Dynamic Model Based on Adaptive Adjustment

近期人类动力学的研究揭示:大量人类行为的时间间隔分布具有标度性。尽管最近建立的人类动力学的排队论模型在解释这种非泊松特性的成因方面取得了很大成功,但还有大量存在标度性的人类行为不能用排队论模型描述。基于人类对自身行为的兴趣或活性的自适应调节现象,提出了一个非排队论的人类动力学模型,并对其进行了数值模拟和解析研究。该模型可以获得指数为-1的幂律形式的时间间隔分布,与一些实证结果相吻合。这一结果表明,自适应调节的兴趣或活性也是人类行为的非泊松特性的成因之一。

Recently, many scaling laws of interevent time distribution of human behaviors are observed, Although the human dynamics models based on the queuing process have got great successes in the explanation of such non-Poisson property of human behaviors, there are many human behaviors with non-Poisson property cannot be explained by the queuing models. In this paper, based on the adaptive adjustments of the interests or activities of human behaviors, a non-queuing process model for human dynamics is proposed and studied by numerical simulations and analytical methods. This model can obtain the power-law interevent distribution with exponent - 1, in agreement with many empirical results. These results imply that the adaptive adjustment of interests or activites is another source of the non-Poisson property of human behaviors.