Abstract
In the process of micro-EDM, tool electrode wear is inevitable, especially for complex three-dimensional cavities or microgroove structures. Tool electrode wear accumulates during machining, which will finally affect machining accuracy and machining quality. It is necessary to reduce electrode wear and compensate it through micro-EDM. Therefore, based on an established L27 orthogonal experiment, this paper uses the grey relational analysis (GRA) method to realize multi-objective optimization of machining time and electrode wear, so as to achieve the shortest machining time and the minimum electrode wear during machining under the optimal machining parameter combination. Then, the orthogonal experiment results are used as dataset of artificial neural networks (ANNs), and an ANN prediction model is established. Combined with image processing technology, the bottom profile of the machined microgroove is extracted and then an electrode axial wear compensation equation is fitted, and a fixed-length nonlinear compensation method for electrode axial wear is proposed. Finally, the GRA optimal experiment shows that machining time, electrode axial wear and radial wear are reduced by 13.89%, 3.31%, and 10.80%, respectively, compared with the H17 orthogonal experiment with the largest grey relational grade. For the study of electrode axial wear compensation methods, the consistency of the depth and width of the machined microgroove structure with compensation is significantly better than that of the microgroove structure without compensation. This result shows that the proposed fixed-length nonlinear compensation method can effectively compensate electrode axial wear in micro-EDM and improve machining quality to a certain extent.