Invited Session 10

Modelling, Fault Diagnosis and Control of Complex System

Submission code: r2838

Organized by

Jun Huang (jhuang@suda.edu.cn)

Soochow University, China

 

Yuanhao Shi (yhshi@nuc.edu.cn)

North University of China, China

 

Zhezhuang Xu zzxu@fzu.edu.cn

Fuzhou University, China

 

Shanying Zhu (shyzhu@sjtu.edu.cn)

Shanghai Jiao Tong University, China

 

With the rapid development of control theory and the improvement of the complexity of system structure, new challenges are presented to the existing control theory and methods. Many interesting results have been achieved in modeling, fault diagnosis, prognostics, and optimization for complex industrial objects.          

 

In the issue of process control and optimization, the establishment of the industrial process model and mastering the law of process behavior are the basis and premise for designing a control system correctly and reasonably. Recently, for increasingly complex systems, several modeling solutions have been proposed, e.g., CPS based methods, active semi-supervised clustering algorithm, and coupling of integral methods. Besides, for an automatic control system, it is necessary to ensure its safety and reliability. Once the automatic control system breaks down, it may cause significant property loss and casualties. Therefore, to improve the safety, reliability, and economy over the life cycle, it is necessary to diagnose the fault in time, prognostics and take corresponding countermeasures. The methods of fault diagnosis and prognostics can be divided into model-based methods, data-based methods, knowledge-based methods, and signal processing-based methods. Among them, model-based methods have attracted the most extensive interests. As an effective tool in fault detection, interval observer has been widely concerned and some results have been achieved. Some scholars have applied the new interval estimation methods to fault estimation, which further enriches the theory of fault diagnosis. The residual useful life(RUL) prediction and optimization decisions are studied in many complex engineering fields, such as lithium battery RUL prediction and soot-blowing optimization. However, these new methods bring new challenges and opportunities to algorithm design, theoretical analysis, and system implementation.

 

The purpose of this invited session is to solicit recent achievements of structural design and control theory of complex engineering systems so as to further improve and develop the theoretical methods of complex industrial system modeling, fault diagnosis,prognostics, and optimization, and promote the development level of informatization and automation of industrial systems.