重叠世代模型的预测与控制:人工智能混沌镇定
Forecasting and control in overlapping generations model: chaos stabilization via artificial intelligence
摘要:
不规则的,尤其是混乱的,行为往往是不可取的经济过程,因为它提出了预测其动态的挑战。在这种情况下,控制这样一个过程的数学模型可以用来抑制混沌行为,并从不规则的系统过渡到规则的动态。本文构造了一个带控制函数的世代重叠模型。通过应用进化算法,我们表明,在没有控制的情况下,在这个模型中可以观察到规则和不规则的行为(周期和混沌)。然后,我们使用的综合控制的金字塔控制方法与两个控制参数,以解决控制问题的不规则行为的模型。我们解决了一些优化问题,应用进化算法来选择控制参数,以确保稳定的周期轨道。我们比较了应用控制前后模型动态的定性和定量特征,并使用仿真验证了所获得的结果。因此,我们表明,人工智能技术(特别是进化算法)结合金字塔控制方法非常适合于在本文所考虑的模型中的不规则动态的深入分析和稳定。
英文摘要:
Irregular, especially chaotic, behavior is often undesirable for economic processes because it presents challenges for predicting their dynamics. In this situation, control of such a process by its mathematical model can be used to suppress chaotic behavior and to transit the system from irregular to regular dynamics. In this paper, we have constructed an overlapping generations model with a control function. By applying evolutionary algorithms we showed that in the absence of control, both regular and irregular behavior (periodic and chaotic) could be observed in this model. We then used the synthesis of control by the Pyragas control method with two control parameters to solve the problem of controlling the irregular behavior of the model. We solved a number of optimization problems applying evolutionary algorithms to select control parameters in order to ensure stability of periodic orbits. We compared qualitative and quantitative characteristics of the model's dynamics before and after applying control and verified the results obtained using simulation. We thus demonstrated that artificial intelligence technologies (in particular, evolutionary algorithms) combined with the Pyragas control method are well suited for in-depth analysis and stabilization of irregular dynamics in the model considered in this paper.
机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。