Estimation of Concentration of Suspended Sediments with Optical-ultrasonic Hybrid System and ANFIS Modeling

Developing a robust and reliable estimation method to detect suspended sediment concentrations from various environmental and geomorphological aspects including water quality, riverbed sustainability engineering, flood management and aquatic habitats is an unavoidable necessity. In this research, a new approach has been developed using a combined optical-acoustic sensors and hybrid intelligence-based system of ANFIS modeling to predict the concentration of suspended river sediments. Also, two measurement systems were placed in a water tank in vitro, and every 50 seconds, 10 g of soil (passed through sieve 140) was added to the water as suspended sediment until the total sediment in the water was 100 grams. The operation was performed in 20 iterations and the output values ​​of the two measurement methods were given as inputs. Interface structure with only optical sensor inputs with higher efficiency coefficient of determination (R2) 0.94 and mean square error root mean square error (RMSE) 7.15 (gr) compared with the ultrasonic sensor inputs with coefficient of determination (R2) of 0.91 and root of the mean squared error was 8.72 (gr). Also, the highest efficiency of hybrid structure with two inputs of two measurement methods had coefficient of determination (R2) 0.97 and root mean square error was 5.26 (gr). According to the results, the best distance between receiver and transmitter in the ultrasonic sensor was between 8 and 15 cm and the use of hybrid system in sediment estimation was more efficient with an error of 3 and 1.5 percent less than the error of separate ultrasonic and optical systems.