Real-world utility of non-singleton fuzzy logic systems: A case of environmental management
Real-world utility of non-singleton fuzzy logic systems: A case of environmental management
复制标题
DOI：
10.1109/fuzz-ieee.2015.7338007
复制
发表时间：
2015-08
期刊：
影响因子：
--
通讯作者：
Amir Pourabdollah;Christian Wagner;Michael Smith;K. Wallace
中科院分区：
文献类型：
--
作者：
Amir Pourabdollah;Christian Wagner;Michael Smith;K. Wallace
非单点模糊逻辑系统在处理不确定性方面的潜力是众所周知的。然而，他们的效用和可能的挑战，在现实世界中的应用，特别是超越模糊控制，仍然没有得到广泛的研究。本文提出了一些以用户为中心的设计方法，使NSFLSs可用于现实世界的环境管理问题。在以前的工作中，一个单一的FLS开发的基础上建立的环境管理框架。在进一步调查用户的要求之后，认识到有效地获取、表示和显示系统的投入和产出至关重要，特别是在数据收集和决策过程中存在不确定性的情况下。为了满足新的需求，该系统已扩展到NSFLS，因此它可以在处理不确定性（例如，噪声）环境数据。受以用户为中心的设计这个特定的系统扩展，本文的贡献是开发一些实用的方法来捕捉/表示输入/输出的不确定性NSFS。根据进一步的用户评估，所解释的方法有可能被用于许多类似的现实世界的应用，从而扩展了NSFL的适用性，以更广泛的背景下比现在。
The potentials of non-singleton fuzzy logic systems (NSFLSs) in dealing with uncertainties are widely known. However, their utilities and possible challenges in real-world applications, particularly beyond fuzzy controls, are still not widely examined. This paper presents some user-centric design approaches in making NSFLSs usable in a real-world problem of environmental management. In previous work, a singleton FLS was developed based on an established environmental management framework. After further investigation of the users' requirements, it was realized that the effective capture, representation and visualization of the system's inputs and outputs are critical, particularly when there are uncertainties involved in data collection and decision-making processes. For addressing the new requirements, the system has been extended to a NSFLS, so it can make use of non-singleton fuzzification in handling uncertain (e.g., noisy) environmental data. Inspired by the user-centric design of this particular system extension, the contribution of this paper is the development of some practical methods to capture/represent input/output uncertainties in NSFLSs. Subject to further users evaluation, the explained methods have potential to be employed in many similar real-world applications, thus extending the NSFLSs applicability to a wider context than the present.