Variable activation function extreme learning machine based on residual prediction compensation

Wang, Gaitang; Li, Ping; Cao, Jiangtao

doi:10.1007/s00500-012-0817-5

Cited by 8 publications

(1 citation statement)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…At present, industrial production mainly uses the single soft sensor modelling method. Including soft sensors based on support vector machines (SVMs; Kaneko and Funatsu, 2013;Liu et al, 2010), soft sensor modelling is based on a leastsquares support vector machine (LSSVM; Feng and Gui, 2012;Li C et al, 2008;Peng et al, 2008) or on artificial neural networks (Hu et al, 2010;Wang et al, 2012;Xu et al, 2011), etc. From the current research results, a single soft sensor model cannot fully describe the global characteristics of industrial systems. Due to the complexity of the process industry, often resulting in a single soft sensor model having a complex structure, long training time and because of the performance of the model constraints, the soft sensing model having satisfactory performance cannot be established; it is difficult to meet the real-time requirements of the dominant variable estimation.…”

Section: Introductionmentioning

confidence: 99%

A multi-model fusion soft sensor modelling method and its application in rotary kiln calcination zone temperature prediction

Tian

Wang

et al. 2015

Transactions of the Institute of Measurement and Control

View full text Add to dashboard Cite

A soft sensor is necessary for industrial process control and analysis, and the core problem is how to construct an appropriate model having a fast convergence speed and a good generalization performance. A multi-model fusion soft sensor modelling method is proposed. Firstly, kernel principal component analysis is applied to choose the non-linear principal component of the model input data space, and then the least-squares support vector machine applied to regression modelling, which could not only reduce the complexity of the calculation but improve the generalization ability. For a least-squares support vector machine, prediction model parameters are difficult to determine; an improved particle swarm optimization algorithm is proposed to optimize the least-squares support vector machine prediction model parameters, and prediction accuracy is improved. The calcination zone temperature of a rotary kiln is chosen as a simulation object because it is difficult to measure directly. Simulation results indicate the proposed soft sensor modelling method has a high learning speed, good approximation and generalization ability, and proved an efficient modelling method.

show abstract