<![CDATA[A new method to automate the batch process of removing toxic chromium ions from wastewater using Oempster-Shafer's (OS) evidence theory is described. The removal of the toxic chromium ions from wastewater is a good example of a process where conventional output or state feedback controllers cannot be simply applied because the concentration of the ion cannot be easily measured online or estimated from other measured parameters. The batch process of removing toxic chromium ions by adding a reducing agent involves reduction and oxidation (redox) reactions which are usually monitored using the oxidation reduction potential (ORP) probe. However, the relationship between ORP and concentration of chromium ions is difficult to establish, hence, a reliable online control is seldom achieved using output feedback control. The approach here is to treat the sequence of ORP values obtained at each sample interval as partial evidences with different degrees of belief to indicate whether the removal process has been completed or not. Using OS's theory of evidence these partial evidences are fused or aggregated to give a more reliable and robust real time control decision. In this paper, a modification is proposed to overcome deficiencies in the OS's combination rule in combining sequences of evidence from the same source. The algorithm based on this evidence theory has been tested in the laboratory, and the results obtained show that the algorithm is robust with respect to noise and process variation.Keywords: Chromium removal process, Dempster-Shafer's evidence theory, evidence aggregation, information fusion, and oxygen reduction potential (CRP).]]>
Copyrights © 2005
