Industrias

Computers &Chemical Engineering
ELSEVIER Computers and Chemical Engin~dng 24 (2000) 785-791 www.elsevier.com/locate/compehemeng

Challenges in the industrial applications of fault diagnostic systems
Sourabh Dash, Venkat Venkatasubramanian *
Laboratory for Intelligent Process Systems, School of Chemical Engineering, Purdue University, W. Lafayette, IN 47907, USA

Abstract Process faultdiagnosis (PFD) involves interpreting the current status of the plant given sensor readings and process knowledge. Early diagnosis of process faults while the plant is still operating in a controllable region can help avoid event progression and reduce the amount of productivity loss during an abnormal event. PFD forms the first step in abnormal situation management (ASM), which aims at timelydetection, diagnosis and correction of abnormal conditions. However the problem of PFD is made considerably difficult by the scale and complexity of modem plants. We briefly outline the various challenges in the area of PFD and review the existing methods to tackle them. We argue that a hybrid blackboard-based framework utilizing collective problem solving is the most promising approach. The efforts ofthe ASM consortium in pursuing the implementation of the state-of-the-art technologies at plant sites are also described. © 2000 Elsevier Science Ltd. All rights reserved.

Keywords: Fault diagnosis; Qualitative trend analysis; Diagnostic methods; Abnormal situation management
ever, considerable progress has been made in all areas of this field. The paper is structured as follows: we firstprovide a brief list of desirable attributes of a diagnostic system achieving all of which in a single technique is a difficult task. We then describe the different diagnostic philosophies in a broad sense, concentrating on a few specific ones to illustrate. Next the state-of-the-art technology being pursued by the ASM consortium at plant sites will be discussed. We end with conclusions and futuredirections. We begin with a list of desirable attributes in the next section.

1. Introduction Abnormal situations occur when processes deviate significantly from their normal regime during on-line operation. Abnormal situation management (ASM) deals with timely detection and diagnosis, assessment of the abnormal situation and countermeasure planning. Process fault diagnosis (PFD) is the first stepin ASM dealing with detection and isolation of abnormal events, i.e. analysis of root causes that result in abnormal behavior. The area of fault detection and diagnosis is an important aspect of process engineering. Not only is it important from a safety viewpoint, but also for the maintenance of yield and quality in a process. This area has received considerable attention from industry andacademia alike because of the economic and safety impact involved (Nimmo, 1995). Intelligent, real-time operator support systems are seen as a way to address ASM. There are however a number of practical challenges in designing such systems due to several factors such as complexity of process dynamics, lack of adequate models, incomplete uncertain data, diverse sources of knowledge, amount of effort andexpertise required to develop and maintain the systems etc. How* Corresponding author. Tel.: + 1-765-4940734; fax: + 1-7654940805. E-mail address: venkat@ecn.purdue.edu(V. Venkatasubramanian)

2. Desirable attributes of a diagnostic system In this section we list some of the desirable characteristics that a diagnostic system should ideally possess to be effective. These are useful to benchmarkvarious methodologies and can also aid in designing better diagnostic methods that meet most requirements.

2.1. Early detection and diagnosis
Early and accurate diagnosis is an important and highly desirable attribute. However, the challenge in realizing it lies in the fact that quick response to failure diagnosis and tolerable performance during normal operation are two conflicting goals....