摘要
磨机负荷参数是影响选矿流程产品质量和产量的难以检测关键过程变量.磨机研磨产生的多源机械信号频谱与磨机负荷参数间存在复杂的非线性映射关系.核潜结构映射(KPLS)算法适合构建基于频谱数据的磨机负荷参数预测(MLPF)模型.针对上述难点,本文提出一种面向MLPF的自适应多核潜结构映射选择性集成(SEN)模型.首先,基于经验模态分解(EEMD)和时频变换技术处理多源机械信号,得到基于不同时间尺度候选子信号的频谱数据;接着,采用KPLS和分支定界选择性集成(BBSEN)算法,构建基于多尺度频谱的候选子子模型和SEN子模型;最后,从候选子子模型和SEN子模型中优选获得不同时间尺度的候选子信号模型,并再次采用BBSEN算法优选集成子信号模型并加权组合,构建最终的MLPF模型.基于实验球磨机的实际运行数据仿真验证了所提方法的有效性.
Load parameters inside ball mill are difficulty-to-measure key process variables relative to production quality and quantity of the whole grinding process. There are complex nonlinear mapping relationships between mill load parameters(MLPs) and multi-source mechanical frequency spectral data. Kernel project to latent structure(KPLS) algorithm is suitable to build mill load parameter forecasting(MLPF) model based on such frequency spectral data. Aim to these problems, a new adaptive multi-kernel projection to latent structure selective ensemble(SEN) model for MLPF is proposed.At first, candidate sub-signals' frequency spectral data with different time scales are obtained by using ensemble empirical model decomposition(EEMD) and time/frequency transformation techniques from multi-source mechanical signals. Then,candidate sub-sub-models and SEN-sub-models are constructed based on different frequency spectral data by using KPLS and branch & bound SEN(BBSEN) algorithms. Finally, the candidate sub-signal models are optimal selected from these candidate sub-sub-models and SEN-sub-models; BBSEN is used again to select ensemble sub-signal models from these candidate ones and to weight them. Therefore, the final MLPF model is constructed. Simulation results of a laboratory-scale ball mill show effectiveness of the proposed approach.
引文
[1]HODOUIN D,JAMAS-JOUNELA S L,CARVALHO M T.State of the art and challenges in mineral processing control.Control Engineering Practice,2001,9(9):995-1005.
[2] ZHOU P, LU S W, MENG Y, et al. Survey on higher-level advanced control for grinding circuits operation. Powder Technology, 2016,288:324–338.
[3] CHAI Tianyou. Operational optimization and feedback control for complex industrial processes. Acta Automatica Sinica, 2013, 39(11):1744–1757.(柴天佑.复杂工业过程运行优化与反馈控制.控制理论与应用,自动化学报, 2013, 39(11):1744–1757.)
[4] TANG J, CHAI T Y, YU W, et al. Modeling load parameters of ball mill in grinding process based on selective ensemble multi-sensor information. IEEE Transactions on Automation Science and Engineering, 2013, 10(3):726–740.
[5] TANG Jian, CHAI Tianyou, ZHAO Lijie, et al. Soft sensing mill load in grinding process by time/frequency information fusion. Control Theory&Applications, 2012, 29(2):564–570.(汤健,柴天佑,赵立杰,等.融合时/频信息的磨矿过程磨机负荷软测量.控制理论与应用, 2012, 29(2):564–570.)
[6] KESHAV P, HAAS B D, CLERMONT B, et al. Optimisation of the secondary ball mill using an on-line ball and pulp load sensor-the sensomag. Minerals Engineering, 2011, 24(3/4):325–334.
[7] MULENGA F K, MKONDE A A, BWALYA M M. Effects of load filling, slurry concentration and feed flowrate on the attainable region path of an open milling circuit. Minerals Engineering, 2016, 89:30–41.
[8] LU S W, ZHOU P, CHAI T Y. Modeling and simulation of whole ball mill grinding plant for integrated control. IEEE Transactions on Automation Science&Engineering, 2014, 11(4):1004–1019.
[9] ZHAO D Y, CHAI T Y. Intelligent optimal control system for ball mill grinding process. Control Theory and Technology, 2013, 11(3):454–462.
[10] YU F U. Development and application on the grinding process modeling and simulation. Journal of Mechanical Engineering, 2015,51(7):197–207.
[11] POMERLEAU A, HODOUIN D, DESBIENS A. A survey of grinding circuit control methods:from decentralized PID controllers to multivariable predictive controllers. Powder Technology, 2000,108(108):103–115.
[12] SHUKLA S, MISHRA S, SINGH B. Power quality event classification under noisy conditions using EMD-based de-noising techniques.IEEE Transaction on Industrial Informatics, 2014, 10(2):1044–1054.
[13] TANG J, CHAI T Y, YU W, et al. Feature extraction and selection based on vibration spectrum with application to estimate the load parameters of ball mill in grinding process. Control Engineering Practice, 2012, 20(10):991–1004.
[14] LEI Y G, HE Z J, ZI Y Y. Application of the EEMD method to rotor fault diagnosis of rotating machinery. Mechanical Systems and Signal Processing, 2009, 23(4):1327–1338.
[15] CUSIDO J, ROMERAL L, ORTEGA J A, et al. Fault detection in induction machines using power spectral density in wavelet decomposition. IEEE Transaction Industial Electronic, 2008, 55(2):633–643.
[16] RIERA-GUASP M, ANTONINO-DAVIU J A, PINEDA-SANCHEZ M, et al. A general approach for the transient detection of slipdependent fault components based on the discrete wavelet transform.IEEE Transaction Industial Electronic, 2008, 55(12):4167–4180.
[17] SESHADRINATH J, SINGH B, PANIGRAHI B K. Vibration analysis based interturn fault diagnosis in induction machines. IEEE Transaction on Industrial Informatics, 2013, 10(1):340–350.
[18] KANKAR P K, SHARMA S C S, HARSHA S P. Rolling element bearing fault diagnosis using auto correlation and continuous wavelet transform. Neurocomputing, 2011, 74(10):1638–1645.
[19] WU Z, HUANG N E. Ensemble empirical mode decomposition:a noise-assisted data analysis method. Advances in Adaptive Data Analysis, 2009, 1:1–41.
[20] SHUKLA S, MISHRA S, SINGH B. Power quality event classification under noisy conditions using EMD-based de-noising techniques.IEEE Transaction on Industrial Informatics, 2014, 10(2):1044–1054
[21] LI RY, HE D. Rotational machine health monitoring and fault detection using EMD-based acoustic emission feature quantification. IEEE Transaction on Instrumentation and Measurement, 2015, 61(4):990–1001.
[22] SINGH D S, ZHAO Q. Pseudo-fault signal assisted EMD for fault detection and isolation in rotating machines. Mechanical Systems&Signal Processing, 2016, 81(5):202–218.
[23] FELDMAN M. Time-varying vibration decomposition and analysis based on Hilbert transform. Journal of Sound and Vibration, 2006,295(3/5):518–530.
[24] TANG Jian, CHAI Tianyou, CONG Qiumei, et al. Modeling mill load parameters based on selective fusion of multi-scale shell vibration frequency spectrum. Control Theory&Applications, 2015, 32(12):1582–1591.(汤健,柴天佑,丛秋梅,等.选择性融合多尺度筒体振动频谱的磨机负荷参数建模.控制理论与应用, 2015, 32(12):1582–1591.)
[25] XU L, ZHANG J Q, YAN Y. A wavelet-based multisensor data fusion algorithm. IEEE Tranctions on Instrumentation and Measurement, 2004, 53(6):1539–1545.
[26] TANG Jian, CHAI Tianyou, ZHAO Lijie, et al. Ensemble modeling for parameters of ball-mill load in grinding process based on frequency spectrum of shell vibration. Control Theory&Applications, 2012,29(2):183–191.(汤健,柴天佑,赵立杰,等.基于振动频谱的磨矿过程球磨机负荷参数集成建模方法.控制理论与应用, 2012, 29(2):183–191.)
[27] HANSEN L K. Neural network ensembles. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1990, 12(10):993–1001.
[28] ZHOU Z H, WU J, TANG W. Ensembling neural networks:many could be better than all. Artificial Intelligence, 2002, 137(1/2):239–263.
[29] TANG Jian, CHAI Tianyou, CONG Qiumei, et al. Soft sensor approach for modeling mill load parameters based on EMD and selective ensemble learning algorithm. Acta Automatica Sinica, 2014,40(9):1853–1866.(汤健,柴天佑,丛秋梅,等.基于EMD和选择性集成学习算法的磨机负荷参数软测量. 2014, 40(9):1853–1866.)
[30] TANG J, YU W, CHAI TY, et al. Selective ensemble modeling load parameters of ball mill based on multi-scale frequency spectral features and sphere criterion. Mechanical Systems&Signal Processing,2016, 66/67:485–504.