A synthesis method for chemical plant operating procedures

https://doi.org/10.1016/0098-1354(87)85018-4Get rights and content

Abstract

A strategy is proposed for the systematic synthesis of operating procedures for chemical processing facilities. This strategy is based on artificial intelligence techniques, most importantly planning with constraints. Introduced in this paper is the technique of symbolic functional modelling, which has particular application in planning operating procedures for processes with recycle streams. Also introduced is the concept of stationary states, which are used as targets during procedure planning and real time operations, along with a decomposition strategy for the discovery of stationary states. Some of this methodology is implemented in a computer program called POPS (Prototype Operating Procedure Synthesis program). The application of POPS to a hydrocarbon chlorination process startup is illustrated.

References (20)

  • P.P. Chakrabarti et al.

    Heuristic search through islands

    Artif. Intell.

    (1986)
  • R.E. Fikes et al.

    STRIPS: A new approach to the application of theorem proving to problem solving

    Artif. Intell.

    (1971)
  • F.P. Lees

    Research on the process operator

  • T.L. Teague
    (1980)
  • E. O'Shima

    Safety supervision of valve operations

    J. chem. Engng Japan

    (1978)
  • J.R. Rivas et al.

    Synthesis of failure-safe operations

    AIChE Jl

    (1974)
  • V.A. Ivanov et al.

    On algorithmization of the startup of chemical productions

    Engng Cybernet

    (1980)
  • V.A. Ivanov et al.

    Design principles for chemical production startup algorithms

    Automn remote Control

    (1980)
  • A. Kinoshita et al.

    An algorithm for synthesis of operational sequences of chemical processing plants

    (1981)
  • D.F. Rudd et al.
    (1973)
There are more references available in the full text version of this article.

Cited by (42)

  • Model-based and rule-based synthesis of operating procedures for planning severe accident management strategies

    2020, Progress in Nuclear Energy
    Citation Excerpt :

    Originated from the planner Stanford Research Institute Problem Solver (Strips) (Fikes and Nilsson, 1971), plenty of OPS systems represent actions with two basic elements, i.e. preconditions and effects. Various planning algorithms were developed to solve different inference issues (Rivas and Rudd, 1975; Fusillo and Powers, 1987). Considering actions may not have a fixed set of effects as argued by Strips-like systems, some OPS approaches adapt the action synergy to synergize effects of multiple actions (O’Shima, 1978; Foulkes et al., 1988).

  • Time-optimal symbolic control of a changeover process based on an approximately bisimilar symbolic model

    2019, Journal of Process Control
    Citation Excerpt :

    Therefore the operations can be provided as some preconditions to be respected by the plant's actuators. Some examples of using this approaches are [5–10]. On the other hand, trajectory-based approaches incorporate also quantitative specifications and optimization of cost functions with respect to process dynamics.

  • Generation of operating procedures for a mixing tank with a micro genetic algorithm

    2013, Computers and Chemical Engineering
    Citation Excerpt :

    Such representation approach works well for static environments, but is inappropriate for many real problems where the duration of an action should be considered. One example of a state-planning approach is the method proposed by Fusillo and Powers (1987) who implement a planner in which the problem is stated as a state-space graph where operations are used to move between states. In their work, a state is defined as a vector that combines qualitative values of process variables with positions of valve actuators and pump switches.

  • An automata based method for online synthesis of emergency response procedures in batch processes

    2012, Computers and Chemical Engineering
    Citation Excerpt :

    Subsequent studies towards the design and verification of procedural controllers under normal plant conditions have also been carried out extensively (Chen & Chen, 1994; Hamid, Sin, & Gani, 2010; Kaspar & Ray, 1992; Kim & Moon, 2009; Naka, Lu, & Takiyama, 1997; Panjapornpon, Soroush, & Seider, 2006; Sanchez & Macchietto, 1995). It can be observed that this issue has been tackled with numerous different modeling/reasoning mechanisms, e.g., the AI-based linear and nonlinear planning strategies (Fusillo & Powers, 1987; Lakshmanan & Stephanopoulos, 1988; Viswanathan, Johnsson, Srinivasan, Venkatasubramanian, & Arzen, 1998), the mathematical programming models (Crooks & Macchietto, 1992; Galán & Barton, 1997; Li, Lu, & Naka, 1997), the symbolic model verifiers (Kim, Kim, & Moon, 2009), and various different qualitative models such as the state graphs (Hoshi, Nagasawa, Yamashita, & Suzuki, 2002; Ivanov, Kafarov, Perov, & Reznichenko, 1980; Kinoshita, Umeda, & O'Shima, 1982) and Petri nets (Chou & Chang, 2005; Hashizume, Yajima, Ito, & Onogi, 2004; Lai, Chang, & Hwang, 2007; Wang, Chou, & Chang, 2005; Yamalidou & Kantor, 1991). Generally speaking, although these different approaches were effective for synthesizing the normal operating procedures, very few of them can be applied to generate proper emergency response strategies.

  • Error-free scheduling for batch processes using symbolic model verifier

    2009, Journal of Loss Prevention in the Process Industries
View all citing articles on Scopus
View full text