Research On Multi-objective Optimization Design Technology Of Lithography Machine Tolerance

Integrated circuit manufacturing industry is a key industry related to social economic development and national security.Lithography tools is the core equipment of large-scale integrated circuit manufacturing,and tolerance optimization design is a very important work throughout the whole life cycle of lithography tools development and manufacturing.The rationality of tolerance selection not only affects the performance,reliability,service life and other key indicators of the product,but also determines the processing technology and production cost of the product to a large extent.Based on the existing tolerance design research,this paper mainly studies and analyzes the tolerance optimization design model construction and model solving algorithm.The main research contents are as follows:(1)In this paper,the theories and methods of tolerance design were analyzed,and a new tolerance optimization design model was proposed,which took economic machining accuracy and assembly function requirements as constraints,and comprehensively weighed machining cost,quality loss cost and quality assurance cost.The model can assign different weights to processing cost,quality loss cost and quality assurance cost according to the characteristics of products or industries,and the distribution of weight coefficients can be calculated by AHP.(2)Quantum-behaved particle swarm optimization(QPSO)is an improved particle swarm optimization(PSO)with global convergence,fewer control parameters and strong optimization performance.However,due to the gradual reduction of particle diversity in the later stage of search,the global search ability is reduced,which makes it easier to fall into the local optimal solution.In view of the randomness and ergodicity of chaotic search,it is an effective method to increase the population diversity of swarm intelligence algorithm.In this paper,the ergodicity of Logistic map and Tent map are compared and analyzed.An improved Tent map with more uniform ergodic performance is proposed and introduced into the iterative process of QPSO algorithm,and a quantum-behaved particle swarm optimization algorithm based on the multi-channel Tent chaotic search(MTQPSO)is obtained.MTQPSO algorithm uses standard algorithm test function to verify the simulation experiment,and the results show that compared with PSO algorithm,QPSO algorithm and CQPSO algorithm,MTQPSO algorithm has better searching performance on single peak and multi-peak function.(3)Taking the mirror group structure of lithography tools as the object of tolerance optimization design,the tolerance design was carried out by using the traditional tolerance design method and the new tolerance optimization design model.The tolerance optimization model is solved by PSO algorithm,QPSO algorithm,CQPSO algorithm and improved QPSO algorithm.By comparison and analysis of the simulation results,the applicability and excellence of the tolerance multi-objective optimization design model and the improved algorithm proposed in this paper are verified in solving the tolerance optimization design problem.