He Yaqun

Research on identification of coal and waste rock based on GLCM and BP neural network

When exploring the parameters for the coal and waste rock, eight characteristic parameters were selected as the whole characters of an image according to their significant differences in gray scale and texture features. On the basis, the error back-propagation algorithm of neural network is applied for the nonlinear identification of samples. The identification network was trained successfully through learning samples. Then, the validity of eight characteristic parameters was verified through the tests of experimental images. Meanwhile, the goal of intelligent identification of coal and waste rock is achieved successfully.

Notice of Retraction The experimental and practical teaching system of the discipline of mineral processing engineering in China University of Mining and Technology

Taking the technology of the processing and utilization of coal resource and clean coal technology as its characteristic, the discipline of mineral processing engineering of China University of Mining and Technology (CUMT) is a Chinese national key discipline, and it has formed a passel of world advanced research achievements and became an important base for the education and scientific research of Chinese mineral processing engineering, and it also has high position and effect in the scientific field of the world. This discipline has a strong back ground of engineering, therefore, the reform of discipline experiments and practical teaching is very important for the cultivation of the spirit of innovation and practical engineering ability of the students. The base and preceptor condition of this discipline were introduced in the paper, and the experimental and practical engineering cultivation system of this discipline in CUMT was discussed from three layers, and the achievement of this system was introduced, too.

A Profit-Oriented Expert System for Coal Washery Optimization

When designing a coal preparation plant, a process flowsheet should be determined to maximize profits. For a given constraint such as clean coal ash or yield, and by applying twelve basic flowsheets for coal preparation recently used in China, the optimum ash and yield composition of the coal products--coal product structure--can be obtained with the maximum profitability. In this article we studied the method of optimization in a coal plant designing expert system to acquire a required coal product structure. By comparing three fitted mathematical models of the raw coal washability curve and seven models of the partition curve, the optimum models are obtained. Based on the models, applying the profits of coal product as an objective function and utilizing the method of generalized Lagrange operators to conditionally constrain the yield and ash content of the coal product, the optimum coal preparation flowsheet was achieved using the Zangwill method. It provides an efficient theoretical basis for defining a technical plan in coal preparation plant design. This paper also studies the computer programming development and implementation of coal product structure optimization by applying an object oriented programming method in a coal plant design expert system. The overall scheme of the coal plant designing expert system, an explanation of the reasoning method, and a discussion of the self-learning mechanism are presented in this paper.

Notice of Retraction Research on identification of coal and waste rock based on PCA and GA-ANN

When exploring identification of coal and waste rock, 17 characteristic parameters of gray-scale histogram and gray level co-occurrence matrix (GLCM) were chosen according to their differences in gray scale and texture. Then, the principal component analysis (PCA) algorithm was used to get principal components from all the parameters chosen above. The principal components were defined as the inputs of GA-ANN to identify the experimental samples of coal and waste rock. The identification rate reaches up to 100% through simulating experiments which proves the feasibility of the characteristic parameters and principal component analysis. Finally, by comparing the simulation results with BP neural network, the identification not only demonstrates the validity in extracting the principal components from the 17 parameters, but also shows that the GA-ANN algorithm is superior to the traditional BP neural network in pattern recognition. The application of PCA combined with GA-ANN provides a new method in intelligent identification for coal and waste rock.

Applying computer multimedia to all teaching/learning links in the national excellent curriculum—Mineral processing plant design

Mineral processing plant design curriculum is innovated ceaselessly so as to improve the education quality and to bring students up the engineering diathesis and design competence through the courses, teaching material, experiment, practice in plants, and scientific research. The computer multimedia technic plays a important role in the curriculum innovating. In practice, a serial research, named APPLYING COMPUTER MULTIMEDIA TO A SET OF TEACHING LINKS is carried out. The computer multimedia is applied to compiling the teaching material, classroom teaching, classroom-work, practice after classroom, graduation project and second classroom, which have made great progress.