Improving accuracy and efficiency of the machined PEEK denture based on NSGA-II integrated GABP neural network

Abstract

Objectives

The polymer polyetheretherketone (PEEK) is gradually being used in dental restorations because of its excellent mechanical properties, chemical resistance, fatigue resistance, thermal stability, radiation translucency and good biocompatibility. To process PEEK dentures with lower surface roughness as quickly as possible, the non-dominated sorting genetic algorithm-II (NSGA-II) integrated genetic algorithm back propagation (GABP) neural network was proposed, which can adjust the combination of process parameters for milling PEEK dentures.

Methods

The PEEK machining was conducted using a four-axis dental milling machine at different process parameters. The surface roughness of PEEK dentures was characterized using surface roughness profiler and scanning electron microscopy (SEM). The optimum machining performance of milling PEEK dentures was investigated using a multi-objective optimization model named as NSGA-II integrated GABP neural network algorithm. The surface roughness (Ra) and material removal rate (MRR) were used as optimization objectives.

Results

The multi-objective optimization model effectively improved surface roughness and machining efficiency for milling PEEK dentures. The validation experiments showed that the surface roughness of all PEEK dentures was less than 0.2$\mu m$, which was within the range of surface roughness set in this paper. The GABP surface roughness prediction model had an average error of 6 %. For the same surface roughness value, the optimized milling parameters all had a greater material removal rate.

Significance

The research results can improve current PEEK denture CAD/CAM technology by providing appropriate milling parameters using NSGA-II integrated GABP algorithm.