ISERME - 2018

Permanent URI for this collectionhttp://192.248.9.226/handle/123/14734

Browse

Browsing ISERME - 2018 by Subject "Artificial neural network"

Filter results by typing the first few letters
Now showing 1 - 1 of 1
  • Thumbnail Image
    item: Conference-Full-text
    A review of prediction of blast performance using computational techniques
    (Department of Earth Resources Engineering, 2018-08) Bhatawdekari, RM; Danial, JA; Edy, TM; Abeysinghe, AMKB; Samaradivakara, GVI
    In hard rock excavation, drilling and blasting is commonly used for loosening rock. Optimum rock fragmentation due to blasting is desirable for downstream operation productuivity. Environmental impacts due to blasting consist of flyrock, ground vibration, air over pressure (AOp). Blast performance depends upon mainly 3 factors consisting of rock mass properties, blast design and explosives system utilised. Mean fragment size is commonly used for rock fragmentation analysis. During 1960-80, blast performance was evaluated using empirical methods. With advancement of computing power during the last two decades, various computentional techniques have been developed for predicting fly rock distance, peak particle velocity, air over pressure with various input paramters based on set of blasts. Technique involves training and testing blast data and comparing results with different computentional algorithm. Various computetntional techniques consisting of Artifical bee algorithem (ABC), Artifical Neural Network (ANN) , Fuzzy Interface System (FIS), GA Genetique algorithm (GA), Imperialist Competitive Alorithm (ICA), Particle Swarm Optimization (PSO), Supoort Vector Machine( SVM) for predicting blast performance are reviewed. Presently, various computentional techniques are ustilsed by researchers. This paper further discusses h ow these techniques can be implemented at operating mines by mining engineers, blasting team for predicting blast performance.