论文部分内容阅读
本文基于人工神经网络的非线性和容错性，对集成电路生产工艺进行了 分析和优化。主要内容有：１．使用人工神经网络方法建立模型，确定生产线上 工艺参数和成品率之间的映射关系，构造以工艺参数为输入，成品率为输出的多 维函数曲面；２．对该多维函数曲面进行搜索，找出成品率最高的最优点；３．以 该最优点的工艺参数值为依据，确定工艺参数的规范值，对工艺参数提出优化建 议，提高成品率。结论：神经网络提出的优化建议是合理的，并已用于集成电路 生产线。
Based on the nonlinearity and fault tolerance of artificial neural network, this paper analyzes and optimizes the production process of integrated circuits. The main contents are: 1. The artificial neural network method is used to establish the model to determine the mapping relationship between the process parameters and the yield of the production line, and the multi-dimensional functional surface with the process parameters as the input and the yield as the output is constructed.2. The multi-dimensional function surface search, find the highest yield the best point; 3. Based on the process parameter value of the optimal point, the specification value of the process parameter is determined, the optimization suggestion is given to the process parameter, and the yield of the product is improved. Conclusion: The neural network’s proposed optimization is reasonable and has been used in integrated circuit production lines.