RESEARCH OF ENERGY CONSUMPTION OF PNEUMATIC CONVEYING FLOUR SECTION BASED ON ERROR CONTROL BP (BACK PROPAGATION) NEURAL NETWORK
-
摘要: 小麦制粉中除研磨工段能量消耗外,气力输送工段也是能耗的一个主要方面,分析气力输送能耗模型是降低面粉气力输送工段能量损耗的基础。然而气力输送模型非线性强、设备匹配参数冗余度大,以及管路压损波动性大。综合考虑生产过程中粉料的状况与输送能耗之间呈现的非线性关系,选择具有较强自适应能力和自组织能力的BP神经网络调节设定处理数据的训练、测试比例和模型训练过程中的总误差和循环次数等参量,建立面粉气力输送工段能耗模型。选取河南某面粉厂500 t/d产能的制粉车间的气力输送工段,采集能耗相关的每小时加工吨麦、麦粒湿度、送风量实测数据为样本,对建立的能耗模型进行训练、预测、验证,结合该厂生产车间设备参数、管路铺设标注和单位生产量推测出吨麦电耗与计量值对比。结果表明:利用BP神经网络拟合算法,建立气力输送工段能量损耗模型可行,且该数学模型可以较好地反映面粉气力输送工段能耗情况,模拟值能够较为准确地预测面粉气力输送工段能耗情况,误差不超过7%。Abstract: In addition to energy consumption of grinding section in wheat pulverization,the pneumatic conveying section is also a major aspect of energy consumption.Analyilzing pneumatic conveying energy consumption can provide foundation for decreasing energy consumptionin pneumatic conveying section based on nonlinear mathematical model of pneumatic conveying,redundancy of equipment matching parameters,and pipe pressure loss by processing and flour texture fluctuation effect.So it was necessary to analyze the energy consumption model of pneumatic conveying in order to reduce the energy consumption in the flour pneumatic conveying section.The nonlinear relationship between the condition of the powder and the energy consumption wascomprehensively considered.The BP (Back Propagation) Neural Network with strong adaptive ability and self-organization ability was selected to establish the energy consumption model of pneumatic conveying flour section,and adjusting the training and testing ratio of the set processing data,the total error and the number of cycles in the model training process,etc.The pneumatic conveying section of the powder workshop of 500 t/da energy production in a flour mill in Henan,sample data of per hour processing ton,wheat grain humidity and air supply volume were collected as samples.The energy consumption model was trained,predicted and verified,and then the power consumption and the measurement value were compared with the equipment parameters,pipeline marking and unit production volume.The results showed that using the BP neural network fitting algorithm was feasible to build the energy loss model of the pneumatic conveying section.The mathematical model can better reflect the energy consumption in the pneumatic conveying section.Meanwhile the simulation results can predict accurately the conveying section of flour pneumatic energy consumption;the error was less than 7%.
-
-
[1] 李林轩.小麦制粉中的气力输送操作管理与故障分析[J].粮食加工,2014,39(5):13-15. [2] 张晋民.面粉厂气力输送的管理和操作[J].粮食与油脂,2002,15(11):41-42. [3] 李林轩.小麦制粉中的节能降耗措施[J].粮食加工,2010,35(2):70-71. [4] 武文斌,刘自然,张明明.面粉厂能量消耗的控制方法[J].粮食与食品工业,2009,16(1):1-2. [5] 刘宝华,李振亮,李亚,等.面粉多点供料气力输送系统能耗分析[J].粮食与油脂,2014,27(5):61-63. [6] 曹先珍.面粉厂电力消耗的费用构成分析与预测[J].粮食与饲料工业,2004(7):10-11. [7] 吴建章,黄学敏.面粉厂节能降耗的技术途径[J].粮食与食品工业,2011,18(4):22-25. [8] CHEN B,CENNA A A,WILLIAMS K C,et al.Investigation of energy consumption and wear in bypass pneumatic conveying of alumina[C]//LEE J,NI J,SARANGAPANI J,et al.Engineering Asset Management, 2011:221-230.
[9] 陈小素,吕先金.浅谈基本风压计算[J].浙江气象,2005,26(4):26-29. [10] KHOTANZAD A, AFKHAMI-ROHANI R,LU T L,et al.ANNSTLF-a neural-network-based electric load forecasting system[J].IEEE Transactions on Neural Networks,1997,8(4):835-846.
[11] 刘国欣.Kolmogorov三级数定理的推广[J].数学物理学报,1995,15(3):311-314. [12] 张甲,兴长策,岳旭东,等.神经网络模型在脱氮预报中的应用[J].辽宁工学院学报,2000(2):1-3. [13] 侯磊,许新裕,崔金山,等.基于BP神经网络的输油管道能耗预测方法[J].节能技术,2009,27(5):401-406. [14] 吴仕勇.基于数值计算方法的BP神经网络及遗传算法的优化研究[D].昆明:云南师范大学,2006. [15] 高山卜,钱成文,张沛,等.基于改进的BP神经网络管输能耗预测模型[J].油气储运,2014,33(8):869-872. -
期刊类型引用(0)
其他类型引用(2)
计量
- 文章访问数: 140
- HTML全文浏览量: 4
- PDF下载量: 30
- 被引次数: 2