APPLICATION OF PARTICLE SWARM OPTIMIZATION TO TOOL SIZE DETERMINATION FOR NARBS PROFILE MILLING

摘要

Particle swarm optimization (PSO) method is a newly-developed evolutionary technique for global minimum/maximum solutions. Because it is simple to implement, this method has attracted many researchers' attention and has been widely applied to engineering problems. In industry, mechanical parts often contain profile features such as profile pockets, sides, and islands, which are usually designed with planar non-uniform rational B-splines (NURBSs) as their profile curves and are side milled along these curves. For finish machining of these profile features without gouging, a cutting tool should be selected with its radius less than the minimum radius of their profile curve. To find this radius, a conventional method is to calculate the curve curvature at every point; however, this method is quite time-consuming. An established method applies local optimization techniques to the profile curve for this radius, but often ends up a local minimum radius, which can cause gouging in some other places. To overcome these difficulties, this proposed research applies the PSO optimization method to find the global minimum radius efficiently. First, the mathematical formulae of the curvatures of a NURBS are derived, and the tool selection criteria are rendered. Second, the PSO method is introduced. Then it is applied to an example part with a profile pocket for its minimum radius, together with the conventional and the established methods. Finally, their results are compared to demonstrate the accuracy and efficiency of the PSO method. Therefore, this proposed work provides an effective approach to determining cutters' size for side milling of profile features and can facilitate the production planning for these parts.