Table 1 Some studies of PPV prediction using soft computing methods33,34,35,36,37,38,39.

From: PPV distribution of sidewalls induced by underground cavern blasting excavation

Technique

Input parameter

No. of dataset

R2

SVM

DI, C

80

R2 = 0.96

ICA-ANN

BS, ST, C, DI, Vp, E

95

R2 = 0.98

ANN

C, PF

232

R2ANN = 0.92

ANFIS

R2ANFIS = 0.98

CART

C, DI

86

R2 = 0.95

FS-ICA

W, D

50

R2 = 0.94

GMDH

SC, W, D

96

R2GMDH = 0.91

GS-GMDH

R2GS-GMDH = 0.94

ANFIS-GOA

PF, ST, RD, S, B

80

R2ANFIS-GOA = 0.97

ANFIS-CA

R2ANFIS-CA = 0.95

  1. SVM support vector machine, ICA imperialist competitive algorithm, ANN artificial neural network, ANFIS adaptive neuro-fuzzy inference system, CART classification and regression tree, FS fuzzy system, GMDH Group Method of Data Handling, GS generalized structure, GOA grasshopper optimization algorithm, CA cultural algorithm, DI distance from the blasting face, C charge per delay, BS burden to spacing, ST stemming, Vp P-wave velocity, E Young modulus, PF powder factor, W weight charge per delay, D distance from the blasting point, SC specific charge, RD rock density, S spacing, B burden, R2 coefficient of determination.
Back to article page