Real time health monitoring of industrial machine using multiclass support vector machine

Abid, Anam; Khan, Muhammad Tahir; Ullah, Azmat; Alam, Mansoor; Sohail, Muhammad

doi:10.1109/iccre.2017.7935046

Cited by 11 publications

(3 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The first investigations were carried out at the beginning of the 70s, based on the design of observers and the use of filters. Later the principle of analytical redundancy was developed [ 2 , 10 , 11 ]. Subsequently, for the first time, a specialized monitoring system was implemented for the detection and diagnosis of failures in a nuclear plant [ 5 , 12 , 13 ].…”

Section: Discussion and Analysis Of The Existing Literaturementioning

confidence: 99%

“…One of the advantages of the proposed method is the use of only the input signal (control signal) and the output of the process to obtain the models. This is in addition to being a technique that does not require too much mathematical complexity, as only available signals and the use of parametric identification techniques are used; that is, it is therefore oriented towards a practical sense that can be implemented in any type of process [ 10 , 11 , 13 , 19 ]. In other words, it is possible to demonstrate that any abnormality in the process can be represented by a mathematical model through identification systems, based, of course, on the history of the process that is being studied, which helps to fulfill the initial objective of the AI detection [ 13 , 50 , 51 , 52 ].…”

Section: Discussion and Analysis Of The Existing Literaturementioning

confidence: 99%

See 1 more Smart Citation

Recent Advances in Intelligent Algorithms for Fault Detection and Diagnosis

Mercorelli

2024

Sensors

View full text Add to dashboard Cite

Fault-finding diagnostics is a model-driven approach that identifies a system’s malfunctioning portion. It uses residual generators to identify faults, and various methods like isolation techniques and structural analysis are used. However, diagnostic equipment doesn’t measure the remaining signal-to-noise ratio. Residual selection identifies fault-detecting generators. Fault detective diagnostic (FDD) approaches have been investigated and implemented for various industrial processes. However, industrial operations make it difficult to implement FDD techniques. To bridge the gap between theoretical methodologies and implementations, hybrid approaches and intelligent procedures are needed. Future research should focus on improving fault prognosis, allowing for accurate prediction of process failures and avoiding safety hazards. Real-time and comprehensive FDD strategies should be implemented in the age of big data.

show abstract

Section: Discussion and Analysis Of The Existing Literaturementioning

confidence: 99%

Section: Discussion and Analysis Of The Existing Literaturementioning

confidence: 99%

Recent Advances in Intelligent Algorithms for Fault Detection and Diagnosis

Mercorelli

2024

Sensors

View full text Add to dashboard Cite

show abstract

“…Riset ini dilakukan oleh (Chavhan, Chavhan and Chavan, 2013) [3] . Penggunaan Machine Learning untuk memonitor mesin diusulkan oleh (Abid et al, 2017) [4] . Metode machine learning yang digunakan yaitu menggunakan Multi Class SVM.…”

Section: A Penelitian Yang Terkaitunclassified

Sistem Monitoring Mesin Menggunakan Multiprosesor Seri Avr Dengan Metoda Pooling

Darsanto

2019

teknokom

View full text Add to dashboard Cite

Pemilihan teknologi untuk mendeteksi mesin industri dengan nilai ekonomis rendah, mudah perawatantapi dengan presisi yang baik umumnya menjadi pilihan di dalam suatu industri agar biaya operasionaltidak terlalu tinggi. Teknologi yang bisa digunakan adalah mikrokontroller. Mikrokontroller adalahsuatu prosesor yang penggunaan bersifat khusus bukan untuk multi user seperti halnya mikroprosesordi suatu perangkat komputer. Mikrokontroler dengan kelebihannya yaitu memiliki nilai ekonomis tapitetap memiliki presisi yang baik memiliki keterbatasan di kecepatan dan kapasitas memory. Untuk itudibutuhkan suatu algoritma yang cukup efisien agar dapat bekerja sesuai dengan yang diinginkan.Tantangan dalam riset ini adalah menggunakan mikrokontroler secara real time (waktu nyata) untukmemonitor mesin yang berjumlah puluhan. Dalam penelitian ini dikembangkan suatu multiprosesordalam arti menggunakan mikrokontroller lebih dari satu unit dan algoritma tertentu agar dapat secarabersamaan memonitor mesin dan saat itu juga menampilkan kondisi mesin industri yang berjumlahpuluhan. Dalam penelitian ini jumlah mesin yang dimonitor yaitu 80 buah dan bisa dirubah sesuaisetting di memori mikro.

show abstract

Software-basierte Funktionen

Wolf

2023

Software Im Automobil

View full text Add to dashboard Cite

Real time health monitoring of industrial machine using multiclass support vector machine

Cited by 11 publications

References 9 publications

Recent Advances in Intelligent Algorithms for Fault Detection and Diagnosis

Recent Advances in Intelligent Algorithms for Fault Detection and Diagnosis

Sistem Monitoring Mesin Menggunakan Multiprosesor Seri Avr Dengan Metoda Pooling

Software-basierte Funktionen

Contact Info

Product

Resources

About