سیستم کارشناسی جامع / رویکرد پژوهش عملیات برای بهینه سازی عملیات خط لوله گاز طبیعی

This paper reports on a project, conducted jointly between SaskEnergy/Transgas and the University of Regina, which aims at developing an integrated decision support system for the optimization of natural gas pipeline operations. In this integrated approach, both expert systems and operations research techniques are used to model the operations of the gas pipelines. The decision support system can perform the tasks of (1) determining the state of the line pack of the pipelines and recommending the control commands to be issued, (2) determining the associated horsepower requirement, and (3) determining the specific compressor unit to be turned on or off. The first two tasks are performed by an expert system, and the third by a fuzzy programming model. The expert system has been developed on G2 and validated using a simulation program.

Pipeline systems are very important for transporting gas, oil and petroleum products, as well as water, in Canada and North America, since they are the most cost-effective ways for moving fluid products over long distances. Each major city in North America needs pipeline systems for its drinking-water distribution. In addition, according to the Canadian Association of Petroleum Producers, 360,000 m3 of crude oil and 0.5 billion m3 of natural gas are transported daily in Canada, over 300,000 km of pipeline systems. The automation of gas pipeline operations could potentially optimize the operations. This paper reports on a feasibility study into gas pipeline automation, conducted jointly between a local gas-transportation company in Regina, Saskatchewan and the University of Regina. The objective of the project is to construct an automated support system that enhances optimized operations of the gas pipelines in order to satisfy customer demand with minimal operating cost. The system consists of an expert system and a mathematical model. The integrated approach of combining two fundamentally different decision-making techniques (the expert system (ES) and operations research (OR)) has been adopted because most process-control and monitoring problems cannot be solved by either OR or ES techniques alone. The operator’s critical decision-making processes usually involve both qualitative and quantitative information. Hence, both OR and ES techniques are needed for modeling these processes. Expert systems are used in problems where no mathematical models can be formulated to provide acceptable answers, but the knowledge of an experienced human expert can give a satisfactory solution. This technique has shown exceptional performance in process control and monitoring when the working knowledge of the system is non-linear and incomplete. Operations research, on the other hand, can be used for problems in which well-constructed mathematical models are available, or can be developed. This paper proceeds as follows: Section 2 gives some background on the application problem domain, and some relevant literature. Section 3 explains the knowledge acquisition and data analysis that precedes system development. Section 4 discusses, first, the motivation for adopting the integrated approach of combining expert systems and operations research, and the rationale for using fuzzy logic for representing uncertainty. Then, the integrated expert system/operations research solution for optimizing natural gas pipeline operations is presented. Section 5 presents in detail the mathematical formulation of the compressor-selection problem. Section 6 discusses how the integrated system can function in the operational environment, Section 7 is the conclusion, and discusses some of the benefits of the integrated approach to system automation.

In order to optimize natural gas pipeline operations, an integrated decision support system combining expert systems and mathematical modeling has been developed.With the support of the GPOA system which determines whether the current line pack level is sufficient to satisfy customer demand and, if not, the horsepower requirement needed to satisfy that demand, subjectivity and inconsistency in a dispatcher’s performance can be minimized. In addition, fuzzy sets and fuzzy logic have been incorporated in a mathematical model to generate a fuzzy programming model for automating the compressor-selection process. The benefits of this integrated system, combining expert system and operations research techniques, include the following: • Fuel gas, maintenance, and call-out cost savings are realized by maintaining an optimal comfort zone. • Knowledge on optimal operation of the pipeline is documented. • Constant customer demand satisfaction is achieved. • System reliability is increased. • Consistent operation is maintained by eliminating unnecessary starting/stopping of compressors. • Training is standardized.