Abstract

The article discusses the information technology of a robust intelligent control system design based on quantum fuzzy inference. The application of the developed design methodology is based on the quantum self-organization of fuzzy controller’s imperfect knowledge bases and leads to an increase in the robustness of intelligent control systems in unpredicted situations. The results of mathematical modeling and physical experiment are compared using the example of an autonomous robot in the form of a “cart – pole” system. Experimental confirmation of the synergetic effect existence in the robust self-organized fuzzy controller formation from a finite number of non-robust fuzzy controllers in on-line has been demonstrated. The resulting effect is based on the existence of hidden quantum information extracted from the classical states of the controller’s time-varying gain coefficients processes schedule. The derived law of quantum information thermodynamics establishes the possibility to forming a thermodynamic control force due to the extracted amount of hidden quantum information and performing additional useful work, that guarantees the achievement of the control goal based on increasing the robustness of a self-organized quantum controller. At the same time, the amount of useful work performed by the control object (at the macro level) exceeds the amount of work spent (at the micro level) by a quantum self-organized controller to extract the quantum information hidden in the responses of imperfect knowledge bases without violating the second thermodynamics information law for open quantum systems with information exchange of entangled super correlated states. A concrete example of an autonomous robot is given, demonstrating the existence of a quantum self-organization synergetic effect to imperfect knowledge bases.

中文翻译：

---

自主机器人智能鲁棒控制：不完善知识库的量子自组织——实验

摘要

文章讨论了基于量子模糊推理的鲁棒智能控制系统设计的信息技术。所开发的设计方法的应用基于模糊控制器不完善知识库的量子自组织，并导致智能控制系统在不可预测情况下的鲁棒性提高。以“车-杆”系统形式的自主机器人为例，对数学建模和物理实验的结果进行了比较。实验证实了在线有限数量的非鲁棒模糊控制器形成鲁棒自组织模糊控制器时存在协同效应。所产生的效果基于从控制器时变增益系数过程调度的经典状态中提取的隐藏量子信息的存在。量子信息热力学的推导定律建立了由于提取大量隐藏量子信息并进行额外的有用功而形成热力学控制力的可能性，从而在提高自组织量子鲁棒性的基础上保证控制目标的实现控制器。同时，控制对象（宏观层面）所做的有用功超过了量子自组织控制器提取隐藏在响应中的量子信息所花费的功（微观层面）。不完善的知识库，而不违反具有纠缠超相关态信息交换的开放量子系统的第二热力学信息定律。给出了自主机器人的具体例子，证明了量子自组织协同效应对不完善知识库的存在。