Introduction
Very recently artificial neural network (ANN) models have been gaining immense popularity for predicting the properties of pelletized iron ore. Neural networks are a type of machine learning model which imitate the processing of data in a human brain and are used to predict the desired output when the rules are known for a given input. In the present time, Neural network models have been used in a variety of applications such as banking, medical diagnosis, optical character recognition, artificial intelligence and process control.
The ability of neural networks to make successful predictions of pelletized iron ore properties has been studied by several researchers. In this paper we present an overview of a neural network model that has been developed to predict the performance of pelletized iron ore.
Description of the Neural Network Model
The neural network model discussed here is based on a 3-layered feed-forward network with backpropagation. The network architecture consists of one input layer with three neurons and one output layer with one neuron. Each neuron in the input layer is connected to one neuron in the output layer. The weights of the connections between the neurons are adjusted using a supervised learning algorithm such as the gradient descent or the backpropagation algorithm.
The input to the neural network is the performance indices of pelletized iron ore. These indices include Acid Solubility Index (ASI), Oxidation-Reduction Potential (ORP), Iron-Unsoluble in Sulphuric Acid (IUS), Total Iron Concentration (TC) and Water Solubility Index (WSI). The output from the neural network is the performance prediction of the pelletized iron ore.
Training of the Neural Network Model
The neural network model is trained using data from a database containing the performance indices and the respective performance predictions of pelletized iron ore. The training data is divided into two sets. The first set is used to generate the initial weight matrix using a trial and error approach. The second set is used to refine the weights through a supervised learning algorithm.
The supervised learning algorithm used by the neural network model is the backpropagation algorithm. During backpropagation, the error between the predicted output and the desired output is calculated and used to adjust the weights of the neural network. The weights are adjusted so that the predicted output of the neural network is as close as possible to the desired output.
Results of the Neural Network Model
The results of the neural network model show that it is capable of making accurate predictions of the performance of pelletized iron ore. The results also show that the model can generalize well, i.e. it is able to make predictions even for novel data points.
Conclusion
In conclusion, the research presented here on the use of neural networks for predicting the performance of pelletized iron ore has produced promising results. The neural network model developed is capable of making accurate predictions and can generalize well to novel data points. As a result, this model can be used for process control and decision making in the mining industry, thus improving efficiency and reducing the cost of production.
