Application Research Of AdaBoost Polynomial Algorithm In Electric Control System Of Mineral Processing

In the mineral processing plant of nonferrous metal, grinding and classification control process plays an important role, and the size of slurry particle is a key factor in evaluating the quality of ore grinding classification process. Hence,it is necessary to take real-time detection of the grinding pulp size. As the cyclone is widely used in large and medium-size ore dressing plant for classification of slurry particle size, cyclone overflow granularity is the most important factor in assessing pulp product quality. However, due to the complexity of the flow field and grinding process variability, it is difficult to find a complete, accurate but simple formula to estimate the size of the dressing overflow.In order to reduce manufacturing and maintenance cost, this thesis has made a development on existing cyclone overflow particle size measurement methods. Specifically, the thesis presents a polynomial-based AdaBoost cyclone overflow granularity predictive model, which is based on electrical system for controlling slurry flow and pressure of beneficiation process.This thesis describes an innovative predictive model based on the polynomial model and AdaBoost algorithm. In the proposed model, the mean square error is minimized to establish the objective function of polynomial model, and the training sample of the poor performance is strengthened to improve the generalization and the precision using AdaBoost. The proposed model is verified to accurately predict the cyclone overflow granularity to ensure the water conservancy cyclone overflow granularity pass rate, and improve the efficiency of mineral processing, which is envisioned to have practical applications.