A hybrid intelligent optimization approach to improving quality for serial multistage and multi-response coal preparation production systems. (April 2018)

• Record Type: Journal Article
• Authors: Yin, Xianhui; He, Zhen; Niu, Zhanwen; Li, Zhaojun (Steven)
• Is Part Of: Journal of manufacturing systems. Volume 47(2018)
• Page Start: 199
• Page End: 216
• Publication Date: 2018-04
• Subjects: Coal preparation system -- Quality improving -- Multistage and multiple response optimization (MMRO) -- Support vector regression (SVR) -- Genetic algorithm (GA); Manufacturing processes -- Periodicals; Production engineering -- Data processing -- Periodicals; Robots, Industrial -- Periodicals; Production, Technique de la -- Informatique -- Périodiques; Robots industriels -- Périodiques; Electronic journals; 670.42
• DOI: 10.1016/j.jmsy.2018.05.006
• Languages: English
• ISSNs: 0278-6125

Abstract

Graphical abstract: Conformance and stability of quality on coal preparation systems is critical for customers. However, conventional quality control methodologies for coal preparation process could become problematic in context of multistage and multiple response. Besides, the existing researches on multistage and multi-response optimization problems have some limitations under the highly nonlinear, high dimensional and limited sample data conditions. Thus, there is little evidence showing that the previous approaches have strong generalization ability. However, the advancement of cyber-physical system and data mining techniques, especially machine learning and intelligent optimization algorithm provides the potential to solve these problems. Therefore, this paper attempts to develop a new and intelligent approach to improve and optimize quality for serial multistage and multi-response coal preparation system. The proposed approach includes two modules. In the first module, using the support vector regression method, we construct the mechanism-data hybrid driven model to reflect and describe the change and propagation mechanism of response characteristic along the stage. In the second module, combining with the constructed response models and modified desirability function, the genetic algorithm -based intelligent optimization search strategy is presented to inversely determine the globally best processing setting conditions. Highlights: The methodology of multistage and multiple responses optimization is presented to cope with the quality improvement problem in sophisticated coal preparation process. The proposed solutions give the support to assure the quality conformance and stability for serial multistage and multiple responses coal preparation system (MMCPS). The forward iterative modeling method based on support vector regression for MMCPS is presented, which not only could reflect the complex correlation between inputs and outputs, but also could consider the interdependency between neighboring stages. A goal-oriented and backward iterative optimization approach based on genetic algorithm is proposed to determine the globally optimal operating conditions of coal preparation system. Abstract: Modeling and optimization of multiple quality response characteristics (i.e., ash content and calorific value) has been a very important issue that can assure the quality conformance and stability for coal preparation system with multiple operating stages. Although extensive research works have been reported on multistage and multiple responses optimization problem, the traditional physical based model and statistical cause-effect model reaches its limits due to the fast increasing complexity, nonlinearity and high-dimensionality of modern coal preparation system. In addition, these conventional modeling and optimization methods have poor self-adaption and self-learning ability to different operation conditions. Internet-of-things and cyber manufacturing techniques make it convenient to collect larger volumes of sensor data that can provide powerful support for efficient data analytics. The combination of massive industrial data, advanced machine learning models and intelligent optimization algorithms bring new opportunities to deal with these problems. Therefore, this paper attempts to propose a hybrid intelligent technique based multistage and multi-response optimization approach for serial coal preparation system. Firstly, using the support vector regression theory, we construct a mechanism-data hybrid driven and forward-iteration model to reflect the change and propagation mechanism of quality response characteristics along the stage. Secondly, combining with the constructed response models and modified desirability function, the genetic algorithm based backward-iteration optimization search method is presented to determine the globally best processing setting conditions. A real life case study is used to verify the usefulness of proposed approach. Additionally, the performance of surrogate models based on different propagation modeling mode, kernel parameter optimization methods and modeling techniques are comprised and discussed. The results show the effectiveness and superiority of proposed approach.