ESTIMATION OF FUEL CONSUMPTION IN AGRICULTURAL MECHANIZED OPERATIONS USING ARTIFICIAL NEURAL NETWORKS

Borges, Pedro H. M.; Mendoza, Zaíra M. S. H.; Maia, João C. S.; Bianchini, Aloísio; Fernándes, Haroldo C.

doi:10.1590/1809-4430-Eng.Agric.v37n1p136-147/2017

Acessibilidade / Reportar erro

Brasil

Español English

sumário « anterior atual seguinte »

Sumário

Scientific Papers, Agricultural Machinery Management • Eng. agríc. (Online) 37 (01) • Jan-Feb 2017 • https://doi.org/10.1590/1809-4430-Eng.Agric.v37n1p136-147/2017 copy

ESTIMATION OF FUEL CONSUMPTION IN AGRICULTURAL MECHANIZED OPERATIONS USING ARTIFICIAL NEURAL NETWORKS

Authorship SCIMAGO INSTITUTIONS RANKINGS

ABSTRACT

This study aimed to develop artificial neural networks for the estimation of tractor fuel consumption during soil preparation, according to the adopted system. The multilayer perceptron network was chosen. As input data: the soil mechanical penetration resistance, the mobilized area by implements, the working gear and the tractor engine speed. The number of layers and neurons varied to form different architectures. The adjustment was verified based on various statistical criteria. The values estimated by the networks did not differ significantly from those obtained experimentally. The conclusion was that the networks showed adequate reliability and accuracy to predicting the fuel consumption in each tillage system, in function of the input data and this can be a useful tool for planning and management of agricultural operations.

KEYWORDS:
machine performance; artificial intelligence; agricultural planning

Statistic	Value	Unit
Average	1.64	MPa
Median	1.67	MPa
Standard deviation	0.23	MPa
Variance	0.05	MPa
Kurtosis Coefficient	-1.37	Dimensionless
Skewness Coefficient	-0.10	Dimensionless
Variation Coefficient	13.88	%

Statistic	Soil tillage system	Performance variables
Statistic	Soil tillage system	Mobilized area (m²).	Slipping (%)	Fuel consumption (L ha^-1)
Average	Conventional	0.55	12.60	19.60
	Reduced	0.52	11.70	7.89
	No tillage	0.31	3.10	7.12
Median	Conventional	0.55	12.55	19.65
	Reduced	0.52	11.98	7.81
	No tillage	0.31	3.00	7.47
Standard deviation	Conventional	0.04	2.12	1.81
	Reduced	0.02	2.25	1.05
	No tillage	0.04	1.11	1.16
Variance	Conventional	0.01	4.51	3.28
	Reduced	0.00	5.08	1.11
	No tillage	0.01	1.23	1.35
Kurtosis Coefficient (dimensionless)	Conventional	-1.22	0.17	-0.49
	Reduced	-1.55	-0.54	0.24
	No tillage	-1.22	0.76	-1.23
Skewness Coefficient (dimensionless)	Conventional	-0.13	0.31	-0.31
	Reduced	-0.02	0.17	0.61
	No tillage	-0.15	0.43	-0.35
Variation Coefficient (%)	Conventional	7.21	16.86	9.24
	Reduced	3.74	18.25	13.33
	No tillage	11.67	19.73	16.29

ANN	Architecture	Program parameters		Average error	R²	RMSE
ANN	Architecture	Selection	Test	Average error	R²	RMSE
1	MLP 12:12-12-1:1	0.2116	0.2245	-0.0022	0.9875	4.1091
2	MLP 12:12-18-1:1	0.1996	0.1587	0.0013	0.9905	3.5820
3	MLP 12:12-21-1:1	0.1754	0.1937	-0.0010	0.9910	3.5352
4	MLP 12:12-24-1:1	0.2170	0.2685	-0.0061	0.9854	4.4634
5	MLP 12:12-10-3-1:1	0.1743	0.1689	-0.0014	0.9906	3.5467
6	MLP 12:12-18-2-1:1	0.2255	0.1886	0.0064	0.9891	3.9061
7	MLP 12:12-18-3-1:1	0.1788	0.1767	-0.0025	0.9874	4.1942
8	MLP 12:12-21-2-1:1	0.2978	0.2045	-0.0048	0.9808	4.9883

ANN and Architecture	Normality test
ANN and Architecture	Kolmogorov-Smirnov (D statistical)	Lilliefors	Shapiro-Wilk (W statistical)
2) MLP12:12-18-1:1	D= 0.1752 p<0.01 ^ * and **** Significant at 5 y 1%.	p<0.01^ * and **** Significant at 5 y 1%.	W= 0.9045 p=0.0011^ * and **** Significant at 5 y 1%.
3) MLP12:12-21-1:1	D= 0.1174 p<0.10 ns	P<0.08 ns	W= 0.9598 p=0.0742 ns
5) MLP 12:12-10-3-1:1	D= 0.0959 p<0.05 ^* * and ** Significant at 5 y 1%.	p<0.01^ * and **** Significant at 5 y 1%.	W= 0.9324 p=0.0351^* * and ** Significant at 5 y 1%.
6) MLP 12:12-18-2-1:1	D= 0.1662 p<0.01 ^ * and **** Significant at 5 y 1%.	p<0.01^ * and **** Significant at 5 y 1%.	W= 0.9178 p=0.0005^ * and **** Significant at 5 y 1%.

ANN and Architecture	t-Student Values			Significance
	Calculated	Tabulated
	Calculated	α = 5%	α = 1%
2) MLP 12:12-18-1:1	0.0387	1.6520	2.3439	ns
3) MLP 12:12-21-1:1	0.0276			ns
5) MLP 12:12-10-3-1:1	0.1810			ns
6) MLP 12:12-18-2-1:1	0.1872			ns

ANN and Architecture	Statistical Criterion
ANN and Architecture	Correlation coefficient	Willmont Index	Performance Index
2) MLP 12:12-18-1:1	0.9905	0.9951	0.9857
3) MLP 12:12-21-1:1	0.9910	0.9954	0.9865
5) MLP 12:12-10-3-1:1	0.9891	0.9943	0.9835
6) MLP 12:12-18-2-1:1	0.9906	0.9953	0.9859

Associação Brasileira de Engenharia Agrícola SBEA - Associação Brasileira de Engenharia Agrícola, Departamento de Engenharia e Ciências Exatas FCAV/UNESP, Prof. Paulo Donato Castellane, km 5, 14884.900 | Jaboticabal - SP, Tel./Fax: +55 16 3209 7619 - Jaboticabal - SP - Brazil
E-mail: revistasbea@sbea.org.br

Acompanhe os números deste periódico no seu leitor de RSS

[1] * Corresponding author. Federal University of Mato Grosso/ Cuiabá - MT, Brasil. E-mail: pborges@ufmt.br