Assessment of river water quality using an expert system

This study is to develop an expert system to assess water quality for user to derive at a decision equivalent to an expert decision. The ES-RWQ was developed as simple yet reliable tool to assist user in assessing the status of the water quality and relating it to the source contributing to the water quality problem. The main objective is to build a prototype of ES-RWQ with an assessment method, and the ability to derive to a recommendation on the water quality problem. While the conventional models can provide an inside to the quality problem but being complex issue in itself, an effective method of extracting value added information from all sources to facilitate decisions on the implementation of cost-effective pollution prevention and control measurements still requires expert consultation. Therefore a re-look at the process of managing water quality are critical and timely to ensure the strategies and actions planned will lead to measurable water quality improvement. Reliable assessment tools are needed to effectively communicate the water quality data so that the data become an important part of finding solutions and decision making process. Expert system (ES-RWQ) aim to assist decision maker and water quality managers to decide on the most appropriate decision/action to be taken when confronting a situation which requires immediate action. ES-RWQ was developed using the visual basic programming language as the tool that consists of user interface, knowledge engine, and inference engine. The user interface can be construct using menu driven or natural language as the communication mode between user and the system. Load duration curve has been identified as one of the assessment tool that offers a practical approach to watershed management. The duration curve has an added value due to its ability to perform a quick and reliable statistical analysis of data targeting the sources as well as linking it to the potential implementation efforts to the hydrologic condition of the watershed. The assessment of water quality was based on pollutant loading model establishing the load capacity of the river to determine specific limit for thr river in receiving pollutant. Selection of tools was based on the combination of prediction reliability with ease of use and reduced requirement of field data. A prototype developed using Microsoft Visual Basic and validated by selecting two sub-catchment from Melaka and one sub-catchment of upper Sungai Langat, Selangor. Four variables Biochemical Oxygen Demand (BOD), Chemical Oxygen Demand (COD), Ammoniacal Nitrogen (AN) and Total Suspended Solid (TSS) of different river classification standard were also selected to validate the ES-RWQ. The 20 years of flow data, the variables, and other information on the sub catchment were input into the prototype, for it to execute the knowledge and the inferences engine to ensure the system produced the expected result. ES-RWQ, produced an output of a load duration curve as the assessment for the water quality status, with it graphical representation, map extract from the geographical information system and recommendation to control pollution source identified. When the information of sub-catchments were executed by the system the output perform as per the specification design for the system to function. Test run with the user agencies was conducted with three rivers; Sg. Pinang, Pulau Pinang, Sg. Jelai, Pahang and Sg. Kanowit, Sarawak. The ES-RWQ was agreed by user agencies as an assessment tool and loading capacity as new approach to determine the status of water quality and identify sources of pollution. With effective knowledge management tool such as ES-RWQ will enhance decision making process and employing computer based-technology to capture knowledge and human expert knowledge as an added toolfor river water quality management.

Text FPAS_2010_1_.pdf Download (279kB)

Actions (login required)

View Item