PDF《Based Framework LPS》

1 Production is the quantity of soil dredged per unit of time. operators. Zhou et al. [14], for example, address this issue by proposing a number of required competences and a system for certification for CSD operators. Others, for example [1] and [9], follow a long tradition of training dredger masters by using pur- pose-built simulators. Other researchers have addressed these issues by exploring how computers can provide assistance in dredging operations. Tang et al. [11] argue that if dredging processes can be monitored by computer software, the dredging state can be evaluated more accurately and, in turn, adjustments can be made more effectively. Similarly, Cox et al. [1] argue that automatic monitoring can free dredging operators from the tedious, prolonged and tiring task of watching many different gauges and apparatuses. Furthermore, Ni et al. [9] suggest that automatic monitoring together with fault detection can facilitate early diagnosis and repair of faults, and even possi- bly precautionary adjustments, before costly deterioration. Our contribution is along these latter lines. In particular we share the objectives of Wang and Tang [13], in providing computerized expert assistance to dredging operators. Fig. 2. A cutter dredger Cutter Head In this paper, which extends [10], we explore how LPS can be used to provide an executable computerized model of CSD operations. We provide a schema for the modeling and a brief outline of the logic-based formalization. This is our first attempt at this application, and the model has been tested only in simulation. To provide a model of CSD there is a need for setting the optimal ranges of various operational parameters, such as ideal ranges of speeds for the cutter head swing and rotation for different types of soil, and the optimal ranges of production. We base our parameters on the work of Li and Xu [8]. They have used data mining techniques on actual dredging data to determine the primary dredging parameters for a balanced optimiza- tion of high production and low energy consumption. In the short to medium term, we see two potential applications for our work. Firstly it can be used as an online advice and guidance system for dredging operators, to help reduce the complexity of their operations and decision making. Secondly it can be used as a training system for would-be operators. In the long term it can be used to automate parts of the dredging operation. Fig. 3. Network of pipes from the dredger head towards discharge Fig. 4. Panel of Monitors Fig. 5. Panel of Monitors and Controllers P1:dredge pump P2:discharge pump Suction pipe Discharge pipe Discharge pipe

2 A Schema of Intelligent Cutter Suction Dredging Using LPS LPS seems particularly well suited to the task of modeling intelligent dredging for several reasons. It allows the representation of the state of the dredging task in terms of the task's operational parameters, and it provides a language that can model both processes for proactive behavior and event-driven production system-type rules for reactive behavior. Thus it can model "normal" operations when everything is going well, and it can model how an abnormality and operational problem can be identified and what steps need to be taken to rectify it. Moreover, the LPS model is executable, in the sense that given periodic input of the dredger sensor readings and monitors, it outputs the next course of actions with their suitable operational parameters. A schema for modeling CSD in LPS is presented in Figure 6. This includes two parts. On the left there is knowledge for intelligent decision-making in dredging using data mining and statistical methods [8]. A small part of this knowledge is summa- rized in Table 1. This shows suitable ranges of some CSD parameters optimal for high production and low energy consumption. These ranges have been extracted for differ- ent types of soil, for example sand, rock and clay. The table focuses on parameters for sand dredging. This data informs the rest of the schema on the right side of Figure