燕山大学学报
燕山大學學報
연산대학학보
JOURNAL OF YANSHAN UNIVERSITY
2014年
3期
221-225,251
,共6页
王葛%李珊%张瑞忠%安领军%李强
王葛%李珊%張瑞忠%安領軍%李彊
왕갈%리산%장서충%안령군%리강
漏钢预报%神经网络%模式识别%粒子群优化算法%PSO-BP网络模型
漏鋼預報%神經網絡%模式識彆%粒子群優化算法%PSO-BP網絡模型
루강예보%신경망락%모식식별%입자군우화산법%PSO-BP망락모형
breakout prediction%neural network%pattern recognition%particle swarm optimization%PSO-BP network model
针对BP神经网络在训练过程中收敛速度慢以及用于模式识别泛化能力差的问题，将粒子群优化算法PSO引入到BP神经网络的训练过程，建立了PSO-BP神经网络模型，并将其应用到连铸漏钢预报系统中。结合某钢厂连铸现场历史数据对该连铸漏钢预报系统进行了测试，测试结果以98.03%的预报率及100%的报出率，验证了基于粒子群优化算法的BP神经网络连铸漏钢预报系统模型的可行性和有效性。
針對BP神經網絡在訓練過程中收斂速度慢以及用于模式識彆汎化能力差的問題，將粒子群優化算法PSO引入到BP神經網絡的訓練過程，建立瞭PSO-BP神經網絡模型，併將其應用到連鑄漏鋼預報繫統中。結閤某鋼廠連鑄現場歷史數據對該連鑄漏鋼預報繫統進行瞭測試，測試結果以98.03%的預報率及100%的報齣率，驗證瞭基于粒子群優化算法的BP神經網絡連鑄漏鋼預報繫統模型的可行性和有效性。
침대BP신경망락재훈련과정중수렴속도만이급용우모식식별범화능력차적문제，장입자군우화산법PSO인입도BP신경망락적훈련과정，건립료PSO-BP신경망락모형，병장기응용도련주루강예보계통중。결합모강엄련주현장역사수거대해련주루강예보계통진행료측시，측시결과이98.03%적예보솔급100%적보출솔，험증료기우입자군우화산법적BP신경망락련주루강예보계통모형적가행성화유효성。
An approach of Particle Swarm Optimization (PSO) is introduced to the training process of BP neural network and a PSO-BP neural network model is designed in this paper in order to overcome the defects of slow speed and the shortcomings of gen-eralization ability for pattern recognition of the BP neural network. The designed model is applied to the steel leakage prediction of continuous casting process. The PSO-BP neural network model is tested with the historical data acquired from a steel mill. The feasibility and the validity of the model are verified by the results of the accuracy rate of 98.03%and the prediction rate of 100%.
