Wyniki wyszukiwania - Biblioteka Nauki

Ten serwis zostanie wyłączony 2025-02-11.

Nowa wersja platformy, zawierająca wyłącznie zasoby pełnotekstowe, jest już dostępna.
Przejdź na https://bibliotekanauki.pl

Ograniczanie wyników

Znaleziono wyników: 2

Liczba wyników na stronie

Wyniki wyszukiwania

Wyszukiwano:
w słowach kluczowych: ANN model

Sortuj według:

Ogranicz wyniki do:

Prediction of the Discharge Coefficient of a Labyrinth Weir Type D by an Artificial Neural Network Method

100%

Belaabed F. , Arabet L.

2024

tom Vol. 71, nr 1

59--72

This study presents the use, and its advantages, of artificial intelligence methods to predict the discharge coefficient (Cw), considering the approach conditions of the labyrinth weir type D. The study suggests modifying the training and validation rates in AI tools, which are often fixed without proper justification in previous studies. Unlike most studies that use geometric dimensions as inputs, this work focuses on the approach conditions (the emplacement of the labyrinth weir and filling the alveoli upstream and downstream) of the labyrinth weir type D. The results, based on laboratory experiments, show that these modified inputs significantly impact the efficiency and cost of constructing the weir. Moreover, the Cw predictions based on these inputs are highly satisfactory compared to laboratory test results. In terms of training and validation ratios, the study confirms that the optimal ratio is 70/30 for accurate and highly satisfactory predictions.

Estimation of grade and recovery in the concentration of barite tailings by the flotation using the MLR and ANN analyses

86%

Deniz V. , Umucu Y. , Deniz O. T.

Physicochemical Problems of Mineral Processing

2022

tom Vol. 58, iss. 5

art. no. 150646

This study aimed to find optimal models in a comparative framework to estimate the recovery and grade of barite concentrate obtained from the rougher flotation of the barite tailings. Therefore, firstly, the effect of four operating parameters (flotation time, pH, collector dosage, and depressant dosage) on the rougher flotation of the barite tailings containing 37.23% BaSO4 was experimentally investigated. Secondly, two models called the multivariable linear regression (MLR) and the artificial neural network (ANN) were used for the estimation of the recovery and grade of the barite concentrate for the rougher flotation optimization. The R2 values found from the MLR and ANN models were 0.828 and 0.995 for the concentrate recovery, and 0.977 and 0.960 for the barite concentrate grade, respectively. In the comparison of the models determined, it was found that the ANN model expressed quite well than the MLR models, especially for the recovery of the rougher concentrate.