Prediction of Specific Fuel Consumption of a Tractor during the Tillage Process Using an Artificial Neural Network Method

Agronomy, Journal Year: 2024, Volume and Issue: 14(3), P. 492 - 492

Published: Feb. 28, 2024

In mechanized agricultural activities, fuel is particularly important for tillage operations. this study, the impact of seven distinct parameters on usage per unit draft power was examined. The are tractor power, soil texture index, plowing speed, depth, width implement, and both initial moisture content bulk density. This study investigated construction an artificial neural network (ANN) model tractor-specific consumption predictions two implements: chisel moldboard plows. ANN created based collection related data from previous research studies, validation performed using actual field experiments in clay a plow. developed (9-22-1) confirmed by graphical assessment; additionally, root-mean-square error (RMSE) computed. Based RMSE, results demonstrated good agreement specific between observed predicted values, with corresponding RMSE values 0.08 L/kWh 0.075 training testing datasets, respectively. novelty work presented paper that, first time, farm machinery manager can optimize carefully controlling certain parameters, such as content, implement width, depth plowing. show that input make significant contribution to output over used different percentages. Accordingly, analysis showed had high plows at 30.13%; contributed 4.19% 4.25% predicting power. concluded practical useful advice production be achieved through optimizing rate selecting proper levels affecting reduce costs. Moreover, could develop future fuel-planning schemes

Language: Английский

Identification of Armyworm-Infected Leaves in Corn by Image Processing and Deep Learning

Acta Technologica Agriculturae, Journal Year: 2024, Volume and Issue: 27(2), P. 92 - 100

Published: June 1, 2024

Abstract Corn is rich in fibre, vitamins, and minerals, it a nutritious source of carbohydrates. The area under corn cultivation very large because, addition to providing food for humans animals, also used raw materials industrial products. exposed the damage various pests such as armyworm. A regional monitoring intended actively track population this pest specific geography; one ways using image processing technology. Therefore, aim research was identify healthy armyworm-infected leaves deep neural network form 4 structures named AlexNet, DenseNet, EfficientNet, GoogleNet. total 4500 images, including infected leaves, were collected. Next, models trained by train data. Then, test data evaluated evaluation criteria accuracy, precision, F score. Results indicated all classifiers obtained precision above 98%, but EfficientNet-based classifier more successful classification with 100%, accuracy 99.70%, -score 99.68%.

Language: Английский

Optimisation and Modelling of Soil Pulverisation Index Using Response Surface Methodology for Disk Harrow Under Different Operational Conditions

Acta Technologica Agriculturae, Journal Year: 2024, Volume and Issue: 27(2), P. 76 - 83

Published: June 1, 2024

Abstract The study aimed to determine the optimal pulverisation index of soil for disk harrow by modelling. A mathematical model was developed using a Design-Expert software and response surface methodology. Experiments were carried out in silty loamy with three different levels moisture content 9.25%, 17.56%, 22.32%, operating depths 10 cm, 15 20 speeds 3.17, 4.85, 5.47 km·h -1 . quadratic proposed statistically significant ( P <0.01), strong correlation relationship R 2 = 0.989) between actual predicted values. adequacy precision achieved at 41.84 showed models‘ ability navigate design space. However, statistical analysis, t -test -value, values have no differences soil. (8.61 mm) desirability 1.00, 14.43%, an depth 11.64 forward speed 5.30 Model validation confirmed acceptability 0.974) 99% accuracy predicting index.

Language: Английский