The Application Of The Genetic Algorithm In Coal Blending Optimization

China is now the middle stage of transforming from the industrial society to theinformation and service dominated society. The traditional industry is still in the leadingposition in China’s economy as a whole. The coking enterprise, as a member of theindustry, plays an important role in both national and local economic development. Coalblending is a critical process in the production of each coking enterprise, which refers tothe preparation for coking or carbonization of the coal. Specifically, it means theblending of various kinds of coal with an appropriate proportion to produce the coke ofhigh quality. Due to the properties, reserves, distribution, exploitation and transportationof the coal, choosing the best blending program with the most ideal feed is a problemworthy of attention. Coal blending optimization is of prominent significance in reducingthe cost, improving economic efficiency, optimizing energy consumption, and reducingenvironmental pressure.This paper takes the coal blending program and process of a coking enterprise asthe object, and establishes the mathematical model through genetic algorithm andgreedy algorithm to minimize the coal blending cost. Based on the characteristics ofgenetic algorithm and greedy algorithm, as well as its application in coking enterprises,the paper does the following research:To start with, the research is conducted on the process of coking coal blending, anda mathematical model is established. Secondly, it investigates the method and procedurefor coal optimization by using the traditional genetic algorithm, discussing theadvantages of traditional genetic algorithm and the place that in need of improvement.Thirdly, due to the deficiency of genetic algorithm and the areas in need ofimprovement, it combines greedy algorithm to make an update. Finally, based on the actual coal blending process of the coking enterprises, the author combines the greedyalgorithm and the MATLAB software to do a stimulation experiment, thus getting theresult of optimization algorithm.To compare the result of the simulation experiment with the blending ratio method,the conclusion can be drawn that the cost is greatly minimized by using the greedyalgorithm during the coal blending optimization. In the meantime, the deficiency of thetraditional genetic algorithm is avoided. Further more, the program has betteradaptability to the changing properties of raw coal, which greatly increases its practicalvalue.