Neural network based on adaptive chaotic gradient descending optimization algorithm and its application in matte converting process

Abstract

An adaptive chaotic gradient descending optimization algorithm for single objective optimization was presented. A local minimum judged by two rules was obtained by an improved mutative-step gradient descending method. A new optimal minimum was obtained to replace the local minimum by mutative-scale chaotic search algorithm whose scales are magnified gradually from a small scale in order to escape local minima. The global optimal value was attained by repeatedly iterating. At last, a BP (back-propagation) neural network model for forecasting slag output in matte converting was established. The algorithm was used to train the weights of the BP neural network model. The simulation results with a training data set of 400 samples show that the training process can be finished within 300 steps to obtain the global optimal value, and escape local minima effectively. An optimization system for operation parameters, which includes the forecasting model, is achieved, in which the output of converter increases by 6.0%, and the amount of the treated cool materials rises by 7.8% in the matte converting process.

Keywords

matte converting / chaotic search / gradient descending / neural network

References

[1] JingGuang-jun, LiangXue-mei, FanXun-li. Design and application of prediction model based on genetic neural network [J]. Computer Engineering and Application, 2001, 30(9): 1-3(in Chinese)

[2] ZhuHui-jian, ZhangTai-shan. Learning algorithm of neural network weight based on simulated annealing algorithm [J]. Journal of Central South University of Technology, 2000, 31: 830-832(in Chinese)

[3] ChoiC, LeeJ. Chaotic local search algorithm [J]. Artificial Life and Robotics, 1998, 2(1): 41-47

[4] ZhouC, ChenT. Chaotic annealing for optimization [J]. Physical Review E, 1997, 55(3): 2 580-2 587

[5] De JongK AAn Analysis of the Behavior of a Class Genetic Adaptive system [D], 1995, Michigan, Michigan University

[6] WangL, SmithK. On chaotic simulated annealing [J]. IEEE Transaction on Neural Networks, 1998, 9(4): 716-718

[7] ZhangTong, WangHong-wei, WangZi-cai. Mutative scale chaos optimization algorithm and its application [J]. Control and Decision, 1999, 14(3): 285-288(In Chinese)

[8] ZhengLi-hui, GuoYa-jun, PanDe-hui. A new global optimization algorithm based upon gradient search [J]. Control Theory and Application, 1997, 14(3): 343-47(in Chinese)

[9] WangLing, ZhengDa-zhong, LinQing-sheng. Survey on chaotic optimization methods [J]. Computing Technology and Automation, 2001, 20(1): 1-5(in Chinese)

[10] KwokT, SmithK. An unified framework for chaotic neural-network approaches to combinatorial optimization [J]. IEEE Transactions on Neural networks, 1999, 10(4): 978-981

[11] LiBing, JiangWei-sun. Chaos optimization method and its application [J]. Control Theory and Application, 1997, 14(4): 163-165(in Chinese)

[12] Goto S. Equilibrium calculations between matte, slag and gaseous phases in copper smelting [A]. Copper Metallurgy-practice and Theory [C]. London, 1974.

[13] KirkpatrickS, GelattC D, VecchiM P. Optimization by simulated annealing [J]. Science, 1983, 220: 671-680

[14] TangWei, ZhangXue-yi, LiDian-pu. Chaotic optimization method for neural networks weights [J]. Journal of Harbin Engineering University, 2000, 21(3): 12-14(in Chinese)

[15] WuMin, TangZhao-hui, GuiWei-hua. Expert control strategy using neural networks for electrolytic zinc process [J]. Trans Nonferrous Met Soc China, 2000, 10(4): 555-560

[16] HuZhi-kun, MeiChi, PengXiao-qi. Intelligent decision and endpoint prediction system for PS furnace matter converting process [J]. Nonferrous Metals, 2002, 54(1): 40-42(in Chinese)