Improved Particle Swarm Algorithm Of Batch Processing Machine Scheduling In Semiconductor Final Test

With the rapid development of the information industries, the semiconductor manufa-cture industry has become a strategic industry which the Chinese government focus on promoting.The semiconductor refers to the electric conductivity between the conductor and the insulator material under the normal temperature,including integrated circuit, chip, the transistor, CPU and so on. Usually, the semiconductor production line is composed of hundreds of processing steps and have different characteristics such as:batch processing, re-entry. Among them, the scheduling problem of the batch processing machine on the semiconductor production is a hot research and a challenge problem. In the process of semiconductor manufacturing, cleaning, diffusion, oxidation, corrosion, ion implantation, electron beam writing (wafer fabrication), roasted (wafer detection) and burning in are all batch-process-ing.This thesis mainly discusses the application of the crossed particle swarm algorithm to the batch-processing machine scheduling problem in the semiconductor final test, firstly, we summarized the main features and reviewed the current solution methods for the batch-processing machine scheduling problems in the semiconductor final test, then, we descri-bed the origin and basic principles of the basic particle swarm algorithm, and redesigns a crossed particle swarm optimization approach. Eventually, improved particle swarm optimization for semiconductor final test of the single batch machine problem and Parallel batch processing machine problem.Computational results showed that it is an effective and efficient method for solving batch-processing machine scheduling problems in semiconductor final test.