A novel grey–fuzzy–Markov and pattern recognition model for industrial accident forecasting

Subject Areas : Mathematical Optimization

Inyeneobong Ekoi Edem ¹ , Sunday Ayoola Oke ² , Kazeem Adekunle Adebiyi ³

1 - Department of Mechanical Engineering, Faculty of Engineering and Technology, Ladoke Akintola University of Technology, Ogbomoso, Oyo State, Nigeria|Department of Industrial & Production Engineering, Faculty of Technology, University of Ibadan, Ibadan, Nigeria
2 - Department of Mechanical Engineering, Faculty of Engineering, University of Lagos, Lagos, Nigeria|Industrial and Production Engineering Unit, Department of Mechanical Engineering, College of Engineering, Covenant University, Ota, Nigeria
3 - Department of Mechanical Engineering, Faculty of Engineering and Technology, Ladoke Akintola University of Technology, Ogbomoso, Oyo State, Nigeria

Received: 2020-10-06 Accepted : 2020-10-06 Published : 2018-09-01

Keywords: Forecasting . Manufacturing . Accidents . Fuzzy, grey, Markov . Pattern recognition,

Abstract :

Industrial forecasting is a top-echelon research domain, which has over the past several years experienced highly provocative research discussions. The scope of this research domain continues to expand due to the continuous knowledge ignition motivated by scholars in the area. So, more intelligent and intellectual contributions on current research issues in the accident domain will potentially spark more lively academic, value-added discussions that will be of practical significance to members of the safety community. In this communication, a new grey–fuzzy–Markov time series model, developed from nondifferential grey interval analytical framework has been presented for the first time. This instrument forecasts future accident occurrences under time-invariance assumption. The actual contribution made in the article is to recognise accident occurrence patterns and decompose them into grey state principal pattern components. The architectural framework of the developed grey–fuzzy–Markov pattern recognition (GFMAPR) model has four stages: fuzzification, smoothening, defuzzification and whitenisation. The results of application of the developed novel model signify that forecasting could be effectively carried out under uncertain conditions and hence, positions the model as a distinctly superior tool for accident forecasting investigations. The novelty of the work lies in the capability of the model in making highly accurate predictions and forecasts based on the availability of small or incomplete accident data.