Hybrid FLC-LMS Algorithm for Predicting Sediment Volume in the River

Problems such as flooding, various diseases from microbial pathogens, rivers becoming slum and unsightly, reduced availability: The data found that 82% of the 550 rivers in Indonesia were polluted and in critical condition. The high level of pollution caused various of clean water and many more. Sediment was one of the causes of that problem. There have been many studies detecting objects in the water, but no one predicted the volume of objects in the water. The purpose of this research is to make a device to detect the volume of sediment in the rivers to improve the performance of river monitoring. A device was equipped with an infrared sensor. Raising and lowering the sensor under the device was controlled by the front DC motor with a Fuzzy Logic Controller (FLC) algorithm using two input and one output. Error (E) and change of error (CE) were inputs for FLC and change of control was the output. The fuzzy membership function depended on the distance of the sensor to the sediment. The number of membership functions used was less than the previous research, so the calculation was much simple. This research used hybrid FLC and Adaptive LMS predictive algorithm. Fuzzy Logic Controller was for controlling the Front DC motor when raising and lowering the sensor and Adaptive LMS predictive algorithm was for predicting the volume and minimalizing the error calculation. PID control was also used as a comparison. The results showed that using the FLC algorithm can find sediment in less than four seconds with an error rate of 0.0005, while the PID sensor found sediment in more than 4 seconds with an error rate of 0.001. In addition, the Adaptive LMS predictive algorithm minimized the error calculation with an accuracy level of 99.66 percent.