Designing a Statistical Process Control Model Through a Fuzzy Inference System to Control Descriptive Characteristic in the Food Industry

Subject Areas :

Bahavar Azarmizad ¹ , Kamaleddin Rahmani Yoshanlui ² , Alireza Bafandeh Zendeh ³ , سیروس فخیمی آذر ⁴

1 - Ph.D Student, Department of Management, Faculty of Management, Economic and Accounting, Tabriz Branch, Islamic Azad University, Tabriz, Iran.
2 - Assistant professor, Department of Management, Faculty of Management, Economic and Accounting, Tabriz Branch, Islamic Azad University, Tabriz, Iran.
3 - Associate professor, Department of Management, Faculty of Management, Economic and Accounting, Tabriz Branch, Islamic Azad University, Tabriz, Iran.
4 - استادیار، گروه مدیریت، دانشکده مدیریت، اقتصاد و حسابداری، واحد تبریز، دانشگاه آزاد اسلامی، تبریز، ایران.

Received: 2023-01-05 Accepted : 2023-02-25 Published : 2025-03-09

Keywords: Fuzzy SPC, Fuzzy Mode, Middle Fuzzy, Fuzzy Inference System.,

Abstract :

The importance of quality in the industry to obtain and produce high quality products has been known for a long time. Quality in the production environment improves reliability, increase production and attracts customer satisfaction. Classical control diagrams, using precise and definite data, place production processes in two groups, «rejection» or «acceptance». Descriptive characteristics are in fuzzy conditions due to ambiguity in the number of defects in the product and decision making by the inspector, and fuzzy sets by defining continuous membership functions and using ambiguous and indefinite data by using triangular and trapezoidal fuzzy numbers in the form of control categories are classified and express the quality level of the product more realistically. This research is an applied and descriptive research, which was carried out with the aim of designing model of Statistical Process Control through a fuzzy inference system to control descriptive characteristics in the food industry. Sampling system has been used in the inspection station to collect information and according to sensory and physical characteristics, the quality level of the produced chocolates was determined. In the Classical Method, 28 cases were identified «under control» and only 2 cases were «out of control». But in the investigation with the fuzzy designed model, 28 samples were «under control», 1 sample was «relatively under control» and 1 sample was «out of control»; Based on the research result, practical suggestions were recommended to the relevant industry

 

Introduction
Quality is a decision criterion for the customer and not a decision tool for engineers, marketers, or management. It is based on the actual customer experience of the product or service, determined against the desired requirements, and the main purpose of quality measurement is to determine and evaluate the degree and level of the product or service used. Statistical process control (SPC) is a powerful set of problem-solving tools that provide consistency in manufacturing processes and increase the ability to produce quality products (2). Statistical process control during production is the main tool needed to achieve this goal; it is also a sampling technique that measures the quality of the items produced. Control charts and process efficiency are two important tools for statistical quality control (3). In traditional control charts, data consists of explicit values.But the measurement system, which mainly includes the operator, the machinery and the environmental conditions, can contain “uncertainties” or “ambiguities” in the explicit data. These uncertainties are based on the process and the measurement system, which can lead to some difficulties in obtaining explicit and definite values ​​from the process (11). In these situations, control charts based on fuzzy set theory are more useful for process evaluation. For this purpose, fuzzy set theory can be adopted for control charts. In these situations, fuzzy set theory supports the development of concepts and techniques to deal with sources of uncertainty or imprecision. A significant contribution of fuzzy set theory is its ability to present and model linguistic and approximate data for the quality control process (10).In many real systems where accurate and definitive information is not always available and information exists in a vague and fuzzy form, fuzzy methods can provide a more accurate examination of the state of the production process by using appropriate linguistic expressions and fuzzy numbers. In this research, fuzzy control charts are developed using fuzzy rules and then the actual process efficiency index is examined to evaluate the accuracy, correctness and performance of the production process in a fuzzy state. The theory of fuzzy sets and random variables has always tried to examine the issue of uncertainty separately. However, in real systems, we always encounter cases where processes are affected by two types of uncertainty. Uncertainty is due to the lack of complete and accurate information and uncertainty is due to the inherent random nature of the processes (4).Although fuzzy set theory has many advantages for modeling uncertainty, sometimes only considering fuzzy sets and considering an integrated system cannot model uncertainty as a result of clear definitions for membership functions (13). Therefore, a fuzzy inference system that includes fuzzy membership functions has been proposed to improve the quality of uncertainty modeling. This system will be able to model uncertainties directly based on existing knowledge and inference rule bases (13). Various studies have focused most of their attention on fuzzifying control limits and have refrained from providing an evaluation and calculation system based on the knowledge base.In many of these studies, such as the study by Esmaeilpour et al. (2009), the logic of designing control charts has been the same as that of deterministic control charts, in which only the numbers are fuzzified and no significant change has been made in simplifying and increasing the efficiency of control charts for companies. Accordingly, it is necessary to present an applied model based on a fuzzy inference system using fuzzy mode and median in a research. Regarding the importance of the subject in terms of applicability, the researcher's studies and investigations show that not much work has been done so far in terms of presenting applied models in the field of control charts based on fuzzy mode and median in companies in East Azerbaijan province in terms of presenting local patterns.Therefore, considering the issues raised, the issue that led the researcher to conduct this research was primarily to pay attention to the fact that in the confectionery and chocolate industry, the evaluation of descriptive characteristics of products such as color, aroma and flavor, texture, chewiness, sugar bloom and foreign substances is carried out through sensory evaluation tests to convert descriptive data into quantitative data and analyze them using methods such as 5-point, 7-point and 9-point tests. However, the use of this method has limitations and errors, because the sensitivity of human senses is subject to various individual, racial, physiological and other factors, and its results can vary with changing conditions if repeated. Accordingly, in this research, fuzzy theory was used to empower statistical control charts, and fuzzy control charts were presented for the desired quality characteristic.For this purpose, a fuzzy inference system will first be designed and modeled so that, based on what Kaya (2021) stated, uncertainty can be modeled based on membership functions and a decision support system can be provided so that through this decision support system, the product evaluation ratio can be evaluated with high accuracy and speed at different times and places. This system will be able to directly model uncertainties based on existing knowledge and the inference rule base. This research aims to design an applied statistical process control model using the fuzzy mode and median method and compare its results with the classical method. In line with this research objective, the main research question is formulated as follows: what kind of model is the applied statistical process control model using the fuzzy mode and median method and how will its results differ from the classical method

 

 Materials and Methods

This research is applied in terms of purpose and descriptive in terms of method, and it is also descriptive-survey in terms of operation that is implemented in the field. Its method is as follows: by conducting library studies, a fuzzy inference system for fuzzy control charts was first designed. In designing this fuzzy inference system, verbal variables and trapezoidal fuzzy numbers were used. Then, based on the designed inference system, the control limits of the control charts were determined based on the mode and fuzzy median. This system was used to control the number of defects. The desired model was designed in the MATLAB R2018A software environment and based on the fuzzy inference system toolbox. The designed system had seven input variables and one output variable.The present study was conducted at Dadash Baradar Industrial Company, which produces various types of chocolates, biscuits, wafers, snacks, cakes, chewing gum, toffee, and caramels, for the descriptive (qualitative) characteristics of milk chocolate. The reason for choosing qualitative characteristics is that quantitative characteristics cannot be expressed in a fuzzy form, but qualitative and descriptive characteristics can be expressed in a fuzzy form. The document study method was used to collect data. In this study, documents are forms that company inspectors use to control defects in a definite way. The definite defect control is used to validate and compare it with the fuzzy mode and median method; also, forms for controlling defects are designed in the form of fuzzy observations, and a trapezoidal fuzzy spectrum is used to enter the fuzzy observations of the number of defects (C). This spectrum is part of the data collection tool.Trapezoidal fuzzy spectrums were designed using research data. To run the model after collecting information, a sampling system was used at the final inspection station. The data was collected in the form of 30 samples of 50, which were related to 30 working days of chocolate product production. Considering the nature of the seven defects, a separate membership function was defined for each of them. The membership functions created trapezoidal fuzzy numbers from the numbers obtained from the production process of Dadash Baradar Industrial Company, and finally, in the data analysis stage, Minitab 21.1.0 and MATLAB R2018A software were used

 

References:

1. اسماعیل‌پور ر، رمضانیان م، کاظم اف ف. ارائه مدل کنترل فرآیند آماري فازي با روش مد فازي براي کنترل تعداد نقص هاي محصول. نشریه مدیریت صنعتی. 1388؛1(2): 18-3.
2. جعفرنژاد، ا.، قاسمیان صاحبی، ا.، معصومی، ب.، و رفعتی، آ. 1398. تحلیل روابط میان عوامل حیاتی موفقیت کنترل فرآیند آماری. چهارمین کنفرانس بین المللی مدیریت صنعتی، یزد.
3. علینژاد خ، دباغ ر، شیرزاد ا. کنترل کیفیت آماری بر‌اساس شاخص کارایی فرآیندها و نمودارهای کنترل با رویکرد فازی (مطالعه موردی: شرکت‌ آب و فاضلاب شهری استان آذربایجان غربی). نشریه مهندسی عمران امیرکبیر. 1398؛ 51(4): 712-699.
4. فراز، ع. 1390. بناسازی نمودار کنترل آماری– فازی فرآیند. تحقیق در عملیات در کاربردهای آن (ریاضیات کاربردی)، 54-45.
5. مرادی طادی، ا. و آوخ دارستانی، س. 1393. توسعه‎ی نمودار کنترل IX-MR با استفاده از روش مد فازی. کنفرانس بین المللی مهندسی، هنر و محیط زیست.
6. Dombi J, Hussain A. A new approach to fuzzy control using the distending function. Journal of Process Control. 2020; 86: 16-29.
7. Ertugrul İ, Aytac E. Construction of quality control charts by using probability and fuzzy approaches and an application in a textile company. Journal of intelligent manufacturing. 2009; 20(2): 139-149.
8. Gulbay M, Kahraman C. Development of fuzzy process control charts and fuzzy unnatural pattern analyses. Computational statistics & data analysis. 2006; 51(1): 434-451.
9. Gulbay M, Kahraman C, Ruan D. α‐Cut fuzzy control charts for linguistic data. International journal of intelligent systems. 2004; 19(12): 1173-1195.
10. Kaya I. Construction of u Control Chart by Type-2 Fuzzy Sets, Proceedings Of the 4th International E-Conference on Advances in Engineering, Technology and Management –ICETM. 2021: 138-143.
11. Kaya, I., Devrim, E., Baracli, H. 2022. Design of Attributes Control Charts for Defects Based on Type-2 Fuzzy Sets with Real Case Studies from Automotive Industry.1173-1195.
12. Kaya I, Erdogan M, Yıldız C. Analysis and control of variability by using fuzzy individual control charts. Applied Soft Computing. 2017; 51: 370-381.
13. Kaya I, Turgut A. Design of variable control charts based on type-2 fuzzy sets with a real case study. Soft Computing. 2021; 25(1): 613-633.

14. Kim E, Park M, Ji S, Park M. A new approach to fuzzy modeling. IEEE Transactions on fuzzy systems. 1997; 5(3): 328-337.
15. Sugeno M, Yasukawa T. A fuzzy-logic-based approach to qualitative modeling. in IEEE Transactions on Fuzzy Systems. 1993;1(1): 7.


16. Wang X, Park J. H, She K, Zhong S, Shi L. Stabilization of chaotic systems with T–S fuzzy model and nonuniform sampling: A switched fuzzy control approach. IEEE Transactions on Fuzzy Systems. 2018; 27(6): 1263-1271.