Application of Fuzzy Logic in Artificial Intelligence

Abstract

In 1965, Lotfi Zadeh created fuzzy logic as a development of Boolean logic in light of the numerical hypothesis of fuzzy sets, which is a speculation of conventional set hypothesis. Fuzzy logic offers a great deal of adaptability for exploiting reasoning, adding an idea of degree when checking states and allowing states to be other than obvious or misleading. , taking into account possible errors and weaknesses. We begin by giving a brief overview of fuzzy logic before talking about its application to artificial and computational intelligence. Then, with a focus on fuzzy expert systems, we discuss cutting-edge applications of fuzzy logic, with one representative example being thoroughly examined. The article\'s conclusion makes recommendations for how fuzzy logic and artificial intelligence might complement one another. We focus our endeavors on few plainly characterized sub-spaces of artificial intelligence, staying away from scattering, and keeping a good arrangement between essential examination and applications.

Introduction

I. INTRODUCTION

The period of the intelligent machine, which began in 1948 with a paper by Alan Turing, generated many problems that are still open to debate today. The creation of the PC, a gadget that permitted individuals to computerize their points of view, essentially affected this time span. However, human thinking isn't exact. You would have a ton of trouble leaving your vehicle completely in one spot. The speculations of fuzzy puts and fuzzy logic were grown together to empower PCs to imitate human perspectives precisely. Because they mimic the fundamental limitations of the human brain to extract information and focus on data that can be applied to decision-making, they provide a formal system for representing vulnerabilities essential to real-world system administration that should be viewed as a numerical hypothesis.  As indicated by defenders of purported "solid" artificial intelligence, these PCs will sometime match our ongoing degree of intelligence. Decidedly, assuming that an issue is sufficiently interesting, innovation will ultimately complete each issue that is feasible by artificial techniques. Obviously, a few scholarly cycles, similar to "mindfulness," have been shown to be rising elements. Von Bertalanffy recently saw the chance of rising features of stunning frameworks during the 1920s while investigating complex organic frameworks. He saw that diverse plans of substances coordinated in unambiguous ways can show uncommon qualities not shared by the components alone. Assuming the entire is partitioned into parts or the parts are organized in an unexpected way, new characteristics are lost. Additionally, the investigation of individual parts is deficient to grasp emanant highlights. At the point when enough mass has gathered, a new element will unexpectedly show up, like the possibility of a minimum amount in material science. As opposed to reductionistic approaches, these methodologies for the most part surmise an all-encompassing perspective on the universe, for example something complex can be more than just the conglomeration or 'amount of its parts'. The moral ramifications should be painstakingly investigated, similarly as they were with the nuclear bomb, which was it could be said the primary artificial utilization of the previously mentioned actual impact. Each innovation can possibly be abused, whether for good or terrible. Since innovation has just at any point filled in as a device for people so far, the people use it who decide if it is helpful or evil. This may no longer be true in the case of intelligence since more sophisticated forms of intelligence may require new ethical standards, but these standards have not yet been attained by the new forms of intelligence in question.

II. FUZZY LOGIC

Fuzzy logic is a logical framework for thinking in uncertain situations that generalizes the traditional two value logic. It is a PC and guess thinking framework that is based on various thoughts and innovations that utilization fuzzy sets, which are classes of articles without clear lines whose not set in stone by degree. The fuzzy set hypothesis, which is an extension of the traditional set hypothesis, is connected to the numerical groundwork of fuzzy logic.

Lotfi A. Zadeh of the College of California at Berkeley posed the underlying viewpoint for fuzzy logic in 1965. At the point when Zadeh understood that customary PC logic couldn't be utilized to control information that reflected erratic or indistinct ideas, he created fuzzy logic, which empowers PCs to recognize information utilizing a scope of grayscales, similar as individuals' thought process.

He first proposed the idea of linguistic variables in 1973. So, by simulating the manner that this type of information is handled, electromechanical controller performance should be improved. The hypothesis took some time to develop at initially, but by the early 1970s, a small international group of scientists had become interested in it. There were a few Westerners, chiefly mathematicians, and a few Japanese engineers among them.

In 1975, process control for cement kilns constructed in Denmark was one of the first industrial applications of fuzzy theory. Yet, as the technology gained greater traction, fuzzy logic was applied to more practical situations.

III. APPLICATIONS OF FUZZY LOGIC

Applications of fuzzy control include self-regulating systems for cameras, washing machines, automatic transmissions, subway controls, or in any case write-verify. These applications are usually very simple but very effective fuzzy rule-based systems. Applications of fuzzy innovation generally fall into two groups. These applications utilize fuzzy logic as a strong data portrayal strategy that conceals unimportant viewpoints and permits creative handling of sketchy data. The use of sensors and effectors is also essential to their effectiveness, but the interaction of many of these parts actually contributes to their success. The second category consists of much more advanced technologies, designed to work in addition to or in place of human experts. These applications include systems for clinical analysis, inventory and stock determination, traffic lights, fuzzy master systems, and fuzzy reservation systems. It is also important to deepen our understanding of how to monitor information systems with ambiguous and inaccurate information, and many issues remain to be resolved at this second meeting. 

Fuzzy logic is employed in a variety of industries, including those that govern the environment, residential appliances, and automobile systems. Many of the widespread uses include:

1. Used in aviation to control the altitude of shuttles and satellites.
2. It controls the speed and traffic of the car system.
3. It is used for dynamic and truly stable organizations and for evaluating individuals in large hierarchical organizations. 
4. We likewise oversee pH, drying and manufactured refining processes in our Substances business.
5. Fuzzy logic is utilized to handle standard dialects and different applications raised with man-made brainpower.
6. Frequently utilized in current control frameworks like expert frameworks.
7. Fuzzy logic imitates the manner in which individuals just decide, yet is a lot quicker. Consequently, it tends to be utilized in cerebrum association.

These were a portion of the typical utilizations of fuzzy logic. Presently we should take a gander at the benefits and obstructions of integrating fuzzy logic into man-made brainpower.

IV. REVIEW OF LITERATURE

To focus on the plan prerequisites by considering the level of their interconnectedness with the client needs, Buyukozkan et al., [2004] have depicted a nonexclusive type of the analytic hierarchy process (AHP) and all the more explicitly of the analytic network process (ANP). This technique utilizes three-sided fuzzy numbers to build the responsiveness to client needs and plan particulars.

Dr. Yoji Akao, a prime supporter of QFD, delivered a book in 2004 that subtleties QFD utilizing contextual analyses from huge Japanese organizations and ventures, covering all that from made and gathered products to development, substance processes, administrations, and software. The House of Quality (HOQ), a grid that gives a calculated guide to the plan process, fills in as the establishment for the QFD approaches.

Yeo et al[2004] .'s analysis of decision-making approaches for conceptual design desirability score. DSMs, or design structure matrices, are also employed as representation and analysis tools to control the design process from a variety of angles.

By proposing a fuzzy-logical derivation instrument that empowers joint effort among creators on the subjective detail of the interrelations among the plan issue's elements or errands, Saridakis et al. [2005] have tended to the DSM as a correspondence configuration device among various planners. To help consumer loyalty, quality function deployment (QFD) has likewise been used to change client needs into specialized plan prerequisites.

A fuzzy model-based strategy for generally assessing the carbon dioxide level of a wood ignition process was introduced by Ruusunen and Leivisk in 2004. The model results are used to appraise the estimated oxygen content as well as continuous burning quality estimations and lower caloric upsides of the wood fuel.

Osofisan and Obafaiye [2007] have talked about contextual analysis research done on the Fluidized Reactant Breaking Unit (FCCU) of the Kaduna Processing plant and Petrochemical Organization in Northern Nigeria utilizing the utilization of fuzzy logic control. This study has proven and confirmed the benefit of fuzzy logic's very simple control system design and implementation.

In order to deal with ambiguous and insufficient data, Gurcanli and Mungen [2009] presented a system for assessing the dangers that construction site employees are exposed to. This method combines historical accident data, expert opinions that are vulnerable to bias, and the present degree of safety at a construction site. Of the 40,000 unclassified work-related injuries across all industries, 5,239 work-related injuries were identified during the construction of this study. The procedure is then practiced at a tunnel construction site to calculate risk levels for all types of accidents. 

In noisy regression situations, Juang et al. [2010] described a type 2 fuzzy neural network using support vector regression. Several architectures can be used to build fuzzy neural systems. To improve the performance of your system, you need to solve three problems: Determine the ideal set of rules, identify a suitable membership function, and fine-tune both. 

Iranmanesh and Madadi (2008) used artificial intelligence (AI) techniques, specifically the Prolong programming language, to extract the project WBS from the mind map of the brainstorming project team. Expert system technology has been used by Biggs et al. [2008] to provide feedback on development to people from various ethnic minority groups.

The focal point of Lee et al[2001] .'s work is on proposing object models with relations expressed as fuzzy sets. Based on a summed-up object model that integrates the perspectives on semantic information displaying, artificial intelligence, and data set systems, Cross [2001] has investigated three sorts of connections intrinsic to protest models: (a) an occurrence of, (b) a-sort of, and (c) a class. Specialists have been utilizing fuzzy logic as a portrayal system in plan difficulties with innate vagueness during dynamic throughout the course of recent many years.

VI. FUZZY LOGIC IN AI: EXAMPLE

A fuzzy logic system is set up, starting with a set of registration functions for each data slice and a set of functions for each result.  The join capability then keeps a bunch of guidelines to give the right result benefit. To all the more likely comprehend Fuzzy Logic, you ought to allude to the Cycle Control Guide.[21]

1)  Step 1

The contribution in this case is temperature and the result is fan speed. For each piece of information, a set of membership grades should be encouraged. Simply put, a membership grades is a visual representation of an arrangement of fuzzy factors.  We'll utilize the fuzzy sets Chilly, Warm, and Hot for this model. Then, we'll foster an enrollment function for every one of the three temperature sets.:

The membership grades must abide by these requirements in order to generate the precise output value needed to power the system. Considering this, we converge the membership grades with an info worth of 52 degrees. As the convergence of the two grades occurs for this situation, we are applying two principles. To make a converging point, you can stretch out the convergence focuses to the result grades. The result functions(grades) can then be shortened at the level of the crossing focuses.

This description of the operation of fuzzy logic systems was incredibly straightforward. A truly working system would have a lot of inputs and may have a lot of outputs. A rather complex collection of functions and a lot more rules would arise from this.

VII. ADVANTAGES AND DISADVANTAGES OF FUZZY LOGIC IN ARTIFICIAL INTELLIGENCE

The benefits of using Fuzzy Logic systems are as follows:

1. It is a strong framework where no exact information sources are required.
2. These systems can oblige a few sorts of information sources including obscure, misshaped or loose information.
3. In the event that the criticism sensor quits working, we can reconstruct it as per the circumstance.
4. Fuzzy logic estimations can be encoded with less data, so they do not consume gigantic extra room.
5. As it looks like human thinking, these systems can tackle complex issues where questionable data sources are accessible and take choices as needs be.
6. These systems are adaptable and the guidelines can be adjusted.
7. The systems have a straightforward design and can be built without any problem.
8. Discreet sensors can be used through these systems, saving construction costs.
9. It is effectively reasonable.
10. It effectively takes care of mind-boggling issues by upgrading its capacity to achieve human-like navigation and thinking errands.
11. It manages vulnerabilities in designing.
12. Fuzzy logic is customizable, so you can easily customize FLS by simply adding or removing rules. 

The drawbacks of Fuzzy Logic systems are:

a. The precision of these systems is compromised as the framework generally deals with mistaken information and data sources.

b. There is no single orderly way to deal with tackle an issue utilizing Fuzzy Logic. Thus, numerous arrangements emerge for a specific issue, prompting disarray.

c. Because of error in results, they are not broadly acknowledged all of the time.

d. The main drawback of fuzzy logic control systems is their complete reliance on human information and control. 

e. You need to consistently refresh the standards of a Fuzzy Logic control framework.

f. These systems can't perceive AI or brain networks.

g. Systems require extensive testing for approval and verification.

Conclusion

Artificial intelligence systems that utilization fuzzy logic to reenact human thinking by and by need master support during advancement. This empowers you to depend on the information on professionals who are more acquainted with the framework. Fuzzy logic can likewise be used to further develop how calculations are done. Fuzzy logic and fuzzy semantics are utilized by IBM Watson. Fuzzy logic, which gives a scope of legitimate thinking that can happen as a subset of the human unique cycle, upholds understanding how individuals will work in a strong setting in a dark plan and directs machines. This report gives a speedy outline of the main discoveries in fuzzy logic and fuzzy simulated intelligence systems at the Artificial Intelligence Exploration Foundation from 1985 to the present. The start, the take-off, and the combination are the three-time spans in which the distribution is coordinated. As we see it, the IIIA has played and keeps on assuming a huge part in both the hypothetical turn of events and viable executions of fuzzy logic and fuzzy man-made intelligence systems.

References

[1] Akao.Y. [2004]: Quality Function Deployment: Integrating Customer Requirements into Product Design. Productivity Press, ISBN 1563273136 [2] Buyukozkan, G., Ertay, T., Kahraman, C. and D. Ruan [2004]: Determining the importance weights for the design requirements in the house of quality using the fuzzy analytic network approach. International Journal of Intelligent Systems, Vol. 19, pp. 443–461. [3] D. Heckerman and A. Mamdani, editors. Uncertainty in Artificial Intelligence (Proceedings of the 9th Conference). Morgan Kaufmann Publishers, 1993. [4] Didier Dubois, Henri Prade, and Ronald R. Yager, editors. Readings in Fuzzy Sets for Intelligent Systems. Morgan Kaufmann, 1993. [5] Dubois D., Godo L., Prade H., Zapico A. (1999). On the Possibilistic Decision Model: from Decision under Uncertainty to Case-Based Decision. International Journal of Uncertainty, Fuzziness and Knowledge-Based Systems 7(6): 631-670. [6] Erich P. Klement. Operations on fuzzy sets and fuzzy numbers related to triangular norms. In Proceedings of the Eleventh International Symposium on Multiple-Valued Logic, pages 218{225, Norman, 1981. IEEE, New York. [7] Flaminio, T.; Godo, L.; and Marchioni, E. (2013). Logics for belief functions on MValgebras. International Journal of Approximate Reasoning 54(4): 491–512. [8] Gurcanli, G.E., and U.Mungen [2009]: An occupational safety risk analysis method at construction sites using fuzzy sets. International Journal of Industrial Ergonomics, Vol. 39, pp. 371–387. [9] Iranmanesh, H., and M.Madadi [2008]: An Intelligent System Framework for Generating Activity List of a Project Using WBS Mind map and Semantic Network. Proceedings of World Academy of Science, Engineering and Technology, Vol. 30, pp. 338-345. [10] Juang, C., Lin, Y., and C. Tu [2010]: Recurrent self-evolving fuzzy neural network with local feedbacks and its application to dynamic system processing. Fuzzy Sets and Systems, Vol. 161, No. 19, pp. 2552-2562. [11] Lee, J., Kuo, J. Y., and N. L. Xuel [2001]: A note on current approaches to extending fuzzy logic to object-oriented modeling. International Journal of Intelligent Systems, Vol.16, pp. 807–820 [12] Maria Zemankova-Leech and Abraham Kandel. Fuzzy relational data bases { a key to expert systems. Verlag TUV Rheinland, 1984. [13] Osofisan, P.B., and O.J.Obafaiye [2007]: Fuzzy Logic Modeling of the Fluidized Catalytic Cracking Unit of a Petrochemical Refinery. The Pacific Journal of Science and Technology, Vol. 8, No. 1, pp. 59-67. [14] Ruusunen, M. and K. Leivisk [2004]: Fuzzy modelling of carbon dioxide in a burning process. Control Engineering Practice, Vol. 12, pp. 607–614. [15] Saridakis, K. M. and A. J. Dentsoras [2005]: A fuzzy rulebased approach for the collaborative formation of design structure matrices. 25th SGAI International Conference on Innovative Techniques and Applications of Artificial Intelligence, Cambridge. [16] Torres A., 2005 Torres A., N. J. J. (2005). Fuzzy logic in medicine and bioinformatics. Journal of Biomedicine and Biotechnology, 2005:1–7. [17] Wolfgang Slany. Scheduling as a fuzzy multiple criteria optimization problem. CD-Technical Report 94/62, Christian Doppler Laboratory for Expert Systems, Technical University of Vienna, 1994. Submitted to Fuzzy Sets and Systems. [18] Yeo, S.H., Mak, M.W. and S.A.P. Balon [2004]: Analysis of decision-making methodologies for desirability score of conceptual design. Journal of Engineering Design, Vol. 15, No.2, pp. 195–208. [19] Zapico A., Godo L. (2000). Representation of preference relations induced by lattice-valued, generalised possibilistic utility functions. International Journal of Uncertainty, Fuzziness and Knowledge-Based Systems 6(8): 719-734. [20] Zhaohao Sun and Sun., 2005 Zhaohao Sun, G. F. and Sun., J. (2005). Four new fuzzy inference rules for experience based reasoning. IFSA World Congress. [21] Sayantini[2023], Fuzzy Logic in AI, Edureka, https://www.edureka.co/blog/fuzzy-logic-ai/#example

Copyright

Copyright © 2023 Fayaz Rasool Ganaie. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.