Material Removal Model of Chemical Mechanical Polishing Based on Genetic Algorithm and Neural Network Optimization

• Published in: 2022 IEEE 2nd International Conference on Mobile Networks and Wireless Communications (ICMNWC) pp. 1 - 5
• Main Authors: Feng, Chenghao, Lan, Xiaopeng
• Format: Conference Proceeding
• Language: English
• Published: IEEE 02.12.2022
• Subjects: Analytical models; Artificial neural networks; chemical mechanical polishing material; Chemicals; genetic algorithm; material removal model; optimized neural network; Rough surfaces; Surface roughness; Trajectory; Wireless communication
• DOI: 10.1109/ICMNWC56175.2022.10031833

With the further refinement of integrated circuit manufacturing, the surface flatness of chips must reach the sub nanometer level. In addition, in order to improve the storage density of computer disks, the requirements for the surface roughness of magnetic heads and disks are also getting higher and higher. The requirements for the overall surface flatness have become a challenge faced by today's electronic industry. chemical mechanical(CM) polishing (cmp) has become the only most effective processing method with its good global flattening ability. Therefore, the material removal model(RM) of CM polishing (CMP) is studied based on genetic algorithm(GA) and NN optimization. This paper briefly analyzes the research status of material removal mechanism of CM polishing, discusses the design process of GA optimized BP Neural Network(NN) model, and applies GA optimized NN to the material RM of CM polishing, which has a forward-looking significance for further understanding CM polishing.