Nur Rezky Safitriani, Erlyne Nadhilah Widyaningrum, Rizka Amalia Putri, Husna Afanyn Khoirunnisa, Morina A. Fathan

Aplikasi Sistem Monitoring Produksi Menggunakan Diagram Kontrol Multivariat Fuzzy T² Hotelling berbasis Alpha-cut dan Transformasi Median

Introduction

Aplikasi sistem monitoring produksi menggunakan diagram kontrol multivariat fuzzy t² hotelling berbasis alpha-cut dan transformasi median. Deteksi penyimpangan proses produksi akurat dengan sistem monitoring adaptif. Menggunakan Diagram Kontrol Multivariat Fuzzy T² Hotelling, alpha-cut, dan transformasi median.

Abstract

Pengendalian kualitas produksi yang adaptif menjadi kebutuhan mendesak dalam menghadapi data multivariat dengan ketidakpastian, disertai tuntutan untuk meningkatkan kualitas produk. Hal ini dapat diatasi menggunakan teori himpunan fuzzy melalui alat Statistical Process Control berupa diagram kontrol. Penelitian ini mengembangkan aplikasi sistem monitoring produksi menggunakan diagram kontrol multivariat fuzzy T2 Hotelling berbasis alpha-cut dan transformasi median. Aplikasinya dilakukan pada industri material bangunan di UD Tiga Beton sebagai penghasil batako press. Monitoring dilakukan pada dua karakteristik kualitas yang saling berkorelasi, yaitu kondisi fisik dan bidang permukaan, yang direpresentasikan dalam bentuk linguistik. Data pengamatan dikonversi ke dalam bilangan fuzzy menggunakan Triangular Fuzzy Number dan proses defuzzifikasi melalui transformasi median serta tambahan alpha-cut sebesar 0,6 agar dapat monitoring pergeseran mean yang kecil. Hasil penerapannya menunjukkan bahwa empat pengamatan terdeteksi berada di luar batas sehingga mengindikasikan proses produksi berada dalam keadaan out of control. Dengan demikian, aplikasi sistem ini terbukti mampu mendeteksi penyimpangan proses secara lebih akurat dan praktis. Diagram kontrol fuzzy multivariat berbasis alpha-cut dan transformasi median menjadi alternatif yang adaptif dalam pengendalian kualitas pada berbagai produksi.

Review

The paper, "Aplikasi Sistem Monitoring Produksi Menggunakan Diagram Kontrol Multivariat Fuzzy T² Hotelling berbasis Alpha-cut dan Transformasi Median," addresses a pertinent challenge in modern manufacturing: adaptive quality control amidst multivariate data and inherent uncertainty. The authors propose a sophisticated application of a fuzzy T² Hotelling multivariate control chart, enhanced by the integration of alpha-cut and median transformation techniques. This approach is positioned as a timely and practical solution for improving the accuracy and efficiency of production monitoring, directly contributing to enhanced product quality in industrial settings where data can be imprecise or linguistically represented. The methodology is robust and thoughtfully detailed, providing a clear pathway for handling qualitative and uncertain data. The conversion of linguistic quality characteristics, such as physical condition and surface area, into Triangular Fuzzy Numbers effectively bridges the gap between subjective observations and quantitative analysis. A significant strength lies in the defuzzification process, which utilizes median transformation, further refined by a strategic alpha-cut of 0.6. This specific alpha-cut value is introduced with the explicit aim of detecting subtle shifts in the process mean, a critical capability for proactive quality management. The real-world application at UD Tiga Beton, monitoring two correlated characteristics of batako press production, provides a compelling demonstration of the system's practical utility. The empirical findings underscore the effectiveness of the proposed monitoring system, with four observations successfully identified as out-of-control, signaling process deviations. The authors effectively argue that this application offers a more accurate and practical mechanism for detecting anomalies compared to traditional quality control methods. The paper concludes by advocating for this fuzzy multivariate control chart, augmented by alpha-cut and median transformation, as a highly adaptive and robust alternative for quality control across a diverse range of production environments. The claims of improved detection and adaptability are well-supported by the described methodology and its successful application.

Full Text

Comments

You need to be logged in to post a comment.