Process modeling of force behavior in the automatic bovine cortical bone milling process using adaptive neuro-fuzzy inference system (ANFIS)

The automatic cortical bone milling process is being utilized in surgeries and orthopedic operations including knee joint replacement, ontological, spinal cord and hip replacement The behavior of forces generated by the cutting tool during the cortical bone milling process should be prudently evaluated. In this article, an adaptive neuro-fuzzy inference system is utilized to model the effect of important parameters in the cortical bone milling process including the rotational speed, feed rate, depth of cut and tool diameter so as to predict the cutting forces. To model the process force behavior, experimental tests are conducted on the fresh cow femur. Next, the results of performed experiments are used to train and test the employed inference system. In this model, the most influential parameters of automatic cortical bone milling process including the rotational speed, feed rate, tool diameter and depth of cut are taken as the input parameters, while the cutting forces in the feed direction, normal to the feed direction and normal to the bone surface as well as the resultant force are considered as the output. With regard to the network training section, the values of root mean square error and mean absolute error criteria are quite small, being close to zero. These criteria also yield relatively small values (less than 0.5) for the date related to the network test section. Overall, the average network errors for estimating the network cutting forces in the training part and test part are equal to 0.37% and 8.7%, respectively.