PDF《Based Framework LPS》

Intelligent Cutter Suction Dredging Using the Logic Based Framework LPS F.

Sadri Department of Computing, Imperial College London, UK [email protected] Abstract. LPS (Logic-based Production System) is a framework that combines logic programs with reactive rules and a destructively- updated database. The logic programs provide proactive behavior and allow definitions of processes, and the reactive rules provide reactive behavior. This paper describes a first attempt in using LPS to model the operations of cutter suction dredging (CSD). It is the result of a year-long consultation with experts from the Dredging Engineering Research Centre at Hohai University. LPS was chosen for this appli- cation because its combination of proactivity and reactivity was thought to be a good match for CSD operations. These require pro- cesses for normal operations, as well as constant monitoring to identi- fy any operational problems that may be arising and taking reactive correction steps. Keywords: Reactive rules, Process modelling, Artificial intelligence, Executable model

1 Introduction LPS (Logic-based Production System) [2,3,4,5,6,7] is a logic-based state transition framework inspired by logic programming and artificial intelligence. It combines logic programs with reactive rules and a destructively-updated database. The logic programs provide goal-driven proactive behavior and definitions of processes and the reactive rules provide event-driven reactive behavior. LPS has both operational and declarative semantics and the operational semantics has been proved sound in general and complete in certain special cases. LPS has been implemented in XSB Prolog and in Java, and has been used for a variety of small trial applications, including stock control, teleo-reactive robotics, workflow and gaming. This paper describes a first attempt in using LPS to model the operations of cutter suction dredging. It is the result of a year-long consultation with experts from the Dredging Engineering Research Centre at Hohai University, and uses their data [8] on dredging parameters. Dredging engineering plays an important role in port construction, flood control and drainage, reclamation projects, and other aspects of environmental manipulation and protection. There are different types of dredgers which operate differently and are suitable for different types of soil [12]. In this paper we concentrate on cutter suction dredgers (CSDs), e.g. in Figure1,which are some of the most widely used types of dredgers. They have a cutting device at the inlet of the suction pipe. The cutting de- vice, exemplified in Figure 2, loosens the water bed by rotation and swinging from side to side, and moves the soil towards the suction mouth where the slurry is then sucked up the suction pipe and transported through a network of pipes, such as in Figure 3, and deposited where required. Dredging using CSDs involves major challenges, one of the greatest of which is the toll it takes on the environment due to high emission and high energy consump- tion, aggravated by inefficiency and low production1 . Operating a CSD requires ex- pertise. Due to the complexity of the dredging environment, operators need to contin- ually monitor and adjust the running state of dredging equipment to prevent pipe blockage and to achieve high production and low energy consumption. The dredging equipment is complex, and operators need to keep an eye on a large set of operation parameters. Fig. 1. A cutter-suction dredger [12] Figure

4 shows one panel of monitors. A dredging operator typically needs to keep an eye on several such panels to check, for example, the flow of the slurry along the network of pipes, the production, the density of the slurry at various points in the pipeline and other parameters. In addition he needs to operate the dredger through control panels such as the one in Figure 5, including control rods and buttons. The efficiency and effectiveness of the dredging operation is highly dependent on the experience of the operator [13,14]. An experienced operator will notice a de- veloping problem quickly, and will know the best steps to rectify the problem before it develops into a costly situation, both in terms of time and resources. Dredging is a growing activity, and it requires a substantial increase in the number of well-trained