Mathematical Model Based On Neural Network Concentrator And Applied Research

The modeling of the mineral processing plant has been the key point of the research of all the persons in this industry. In recent years the application of ANN (Artificial neural network) greatly improved the accuracy and reliability and enriched the theory and approach of the modeling of mineral processing. But, What the mineral processing actually needed is a more complex and sounded non-linear model. So you can't expect the use of ANN could solve all the questions. The essay use ANN, Exploratory Data Analysis, Principal Components Analysis and SPSS (Statistical Package for the Social Science) etc. to explore a new way of the modeling of mineral processing plant.1. In the process of mineral processing the relationships of the different data may be linear, approximate linear or simple non-linear, but mainly non-linear. Principal Components Analysis can solve the interrelationships of the input and output of BP model(a kind of model of ANN) and improve the quality of the model.2. To explore the law of the mineral processing data by using Exploratory Data Analysis technique, which is to find the most efficient way of mineral separation and the best combination of the grade of fine mineral and recovery and to improve the efficiency and benefits of the mineral processing plant.3. To analyse the result of the Exploratory Data Analysis technique by using Principal Component Analysis & BP technique, i.e. the result of separation being input factor and the separation condition being output factor, to find the laws and use the laws to guide the reagent usage and improve the efficiency of mineral separation.What the author has done can improve the quality of mathematic model of mineral processing plant, enrich the non-linear modeling theory > methods& practice. At the other hand the model of mineral separation can not only predict the efficiency & benefits of the mineral processing plant, but also can be used to predict the conditions of the separations, such as floatations gravity separation & magnetic separation etc..