Fuzzy logic model development for troubleshooting at degumming and bleaching process

Abstract

In palm oil industries, troubleshooting to overcome plant or equipment failure is always performed by engineers or plant worker. Failures at the plant affect the sustainable development for industry. Failures can increase the operating cost, product waste and emotional stress of the plant workers. A more systematic troubleshooting method has been introduced to support troubleshooting work to reduce mistakes and increase efficiency. In this paper, a fuzzy logic model for troubleshooting cases that represent failures to obtain product quality at degumming and bleaching processes in a palm oil refinery is presented. Fuzzy logic model has been widely used in complex chemical processes due to its ability to capture human knowledge and combine it with numerical information. Fuzzy logic consists of three major elements which are fuzzification, fuzzy inference system (FIS) and defuzzification. In fuzzification step, crisp value data which is obtained from Distributed Control System (DCS) and technical documentation are converted into fuzzy value by mapping it to the membership function that represent linguistic labels such as "low", "normal" and "high". Fuzzy if-Then rules are used to map the fuzzy input to fuzzy output using Mamdani approach in fuzzy inference system. The fuzzy ifthen rules are developed based on expert knowledge and experience in palm oil industry. Lastly, the fuzzy value is defuzzified to obtain crisp value by using Centre of Gravity method. The model is developed using fuzzy logic toolbox in MATLAB software. The fuzzy logic model was tested with cases of bleaching temperature failures at degumming and bleaching processes. Based on the cases presented in this paper, the results show that the model is capable to diagnose failures and suggest action.