基于改进粒子群算法的数字光刻成像研究

发布时间：2018-01-15 13:30

本文关键词：基于改进粒子群算法的数字光刻成像研究　出处：《南昌航空大学》2017年硕士论文　论文类型：学位论文

【摘要】：集成电路是当今智能化产业的基础,而光刻技术是集成电路发展的主要推动力。相较于传统的掩模光刻技术,数字光刻技术具有灵活、速度快、无需物理掩模等优点。然而,由于光学邻近效应的存在,掩模图形投影到硅片上时会出现畸变,造成掩模失真。本文围绕提高数字光刻质量进行了深入研究,主要内容如下:(1)研究基于像素的数字光刻成像方法。为了提高数字光刻的质量,我们就必须找出更好的掩模。运用传统的成像算法,不仅计算复杂、成像精度低,并且传统的成像算法把掩模作为整体进行优化,不利于分析每个像素对光刻成像的影响。基于像素的数字光刻技术使用聚焦光学器件投影图像,因此单个像素的强度分布可以被认为是一个点扩散函数,其可以由高斯分布描述。通过仿真实验,证明了该算法具有更快的成像速度和更高的成像精度。(2)提出了一种基于双重循环增加粒子多样性的改进粒子群算法。粒子群算法具有收敛快、易实现的优点。但是,传统的粒子群算法对于复杂的问题很有可能会陷入局部极值,从而导致过早收敛,无法获得最优解,而导致陷入局部极值的主要原因之一是算法后期粒子多样性的缺失,为了避免陷入局部极值,我们设置两重循环来增加粒子的多样性。第一重循环控制粒子群重新初始化的次数,以此来增加粒子群的多样性,相应的扩大搜索范围;第二重循环控制每次粒子群更新之后的粒子的更新次数。将改进的粒子群算法与其它几种粒子群算法在多个测试函数上进行了实验对比,结果证明该算法具有更好的优化性能。(3)利用改进的粒子群算法对数字光刻掩模进行优化。由于基于像素的数字光刻成像算法会使与照明图像无关的像素也获得光强,进而会导致成像图形边缘比较模糊。基于此,本文对掩模上像素灰度值进行优化。利用改进的粒子群算法先对灰度因子进行初始化,将其作为当前全局最优。其次对灰度因子进行更新,并将成像结果与原始掩模进行对比,若图形误差小于当前全局最优对应的图形误差,则全局最优更新,否则保持不变,通过不断迭代,最终找到合适的灰度因子,从而改变成像图像光强分布,提高成像质量。最后对其在不同的掩模图形上进行了实验,证明了算法的优化性能。
[Abstract]:Integrated circuit is the basis of intelligent industry nowadays, and lithography technology is the main driving force of the development of integrated circuit. Compared with traditional mask lithography technology, digital lithography technology is flexible and fast. However, because of the optical proximity effect, the mask pattern will be distorted when projected onto the silicon wafer, resulting in mask distortion. In this paper, the improvement of digital lithography quality is studied in depth. In order to improve the quality of digital lithography, we must find a better mask. Using traditional imaging algorithm, not only computational complexity. The traditional imaging algorithm optimizes the mask as a whole, which is not conducive to analyzing the effect of each pixel on the lithography. The pixel based digital lithography technology uses focused optical devices to project images. Therefore, the intensity distribution of a single pixel can be considered as a point diffusion function, which can be described by Gao Si distribution. It is proved that the algorithm has faster imaging speed and higher imaging accuracy.) an improved particle swarm optimization algorithm based on double cycles to increase particle diversity is proposed. The particle swarm optimization algorithm has fast convergence. However, the traditional particle swarm optimization algorithm is likely to fall into local extremum for complex problems, which leads to premature convergence and can not obtain the optimal solution. One of the main reasons for falling into local extremum is the loss of particle diversity in the later stage of the algorithm, in order to avoid falling into local extremum. We set up double cycles to increase the diversity of particles. The first cycle controls the times of particle swarm reinitialization to increase the diversity of particle swarm and expand the search range. The second cycle controls the number of particle updates after each PSO update. The improved PSO algorithm is compared with other PSO algorithms on several test functions. The results show that the algorithm has better optimization performance. The improved particle swarm optimization algorithm is used to optimize the digital lithography mask. Because of the pixel based digital lithography imaging algorithm, the pixels independent of the illumination image can also get light intensity. Based on this, the gray value of pixels on the mask is optimized, and the improved particle swarm optimization algorithm is used to initialize the gray level factor. It is regarded as the current global optimal. Secondly, the gray factor is updated, and the imaging results are compared with the original mask. If the image error is smaller than the current global optimal corresponding graphics error, then the global optimal update. Otherwise, we can find the appropriate gray factor through iteration, which can change the intensity distribution of the imaging image and improve the image quality. Finally, the experiment is carried out on different mask images. The optimization performance of the algorithm is proved.
【学位授予单位】：南昌航空大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP18;TN405

【参考文献】