Optimization of Nonlinear Controller for Cutting Force Compensation in Machine Tool

Ting, Terence Sy Horng (2023) Optimization of Nonlinear Controller for Cutting Force Compensation in Machine Tool. Final Year Project (Bachelor), Tunku Abdul Rahman University of Management and Technology.

Text Full Text - TERENCE TING SY HORNG.pdf Restricted to Registered users only Download (6MB)

Abstract

Based on the progress of science and technology, the accuracy and precision of the machine tools are more valued in the manufacturing industries to produce the final workpiece with accurate measurements. During the milling process, the feed of the machine tool against the workpiece can produce external disturbance forces, such as cutting force and friction force. The disturbance forces can reduce tracking accuracy and affect positioning accuracy. In this research, cutting force data from different spindle speeds, which are 1000 rpm, 1500 rpmand 2000 rpm, is set as the disturbance force. To overcome these problems, the goal of this study is to compensate for the cutting force in the machine tool by designing a suitable nonlinear controller. The sliding mode controller (SMC) is selected, and its tuning parameters are found by applying different optimization algorithms, such as Genetic Algorithm (GA), Particle Swarm Optimization (PSO), and Ant Colony Optimization (ACO). The chattering effect is a main drawback for SMC that results of discontinuous signals that can seriously impair the accuracy of machine tool tracking. The tuning parameters are obtained from the SMC without cutting force data by applying optimization algorithms, which is the robust method. The tuning parameters that obtained are applied and the cutting force data with motor constant is included in the SMC. The results obtained for SMC without the cutting force data are proved to reduce the chattering effect, while the results obtained for SMC with the cutting force data are proved to compensate for the cutting force. For the SMC without cutting force data, the Particle Swarm Optimization (PSO) is chosen as the most effective tuning algorithm because of its greatest chattering suppression capability among three different tuning algorithms, with a reduction of 61.53% in tracking error results, 69.868% in FFT results, 67.391% in ITAE results, and 71.03% in RMSE results when compared to the benchmark. However, the other two tuning algorithms, GA and ACO, also show improved effects on chattering suppression. Besides that, the Genetic Algorithm (GA) shows the greatest and stable cutting force compensation, with an average reduction of 48.608% in tracking error results, 61.562% in ITAE results, and 62.057% in RMSE results when compared to the benchmark. In short, the Genetic Algorithm (GA) has achieved all the objectives of this research by showing good reduction on chattering effect and cutting force compensation.