基于遗传算法的复杂绘制系统性能瓶颈识别

Identify Performance Bottlenecks on Algorithm Level Based on Genetic Algorithm

研究在绘制系统满足指定渲染效果的条件下,针对该系统的实时性能,完成基于遗传算法的绘制系统性能瓶颈的识别方法。即:在建立绘制系统中绘制算法参数与绘制时间关系模型的基础上,利用遗传算法完成绘制特征空间的全局最优划分以及对应空间下性能瓶颈的识别。从而进一步指导实时绘制系统的开发者,在满足绘制效果的前提下,如何调整绘制算法的参数,以获得高效的绘制效率。

Studies how to identify the performance bottleneck of rendering system based on genetic algorithm under the condition that rendering sys. tem satisfies the specified rendering effect and aims at the real-time performance of the system. That is, on the basis of establishing the re. lationship model between rendering algorithm parameters and rendering time in the system, the global optimal partition of rendering feature space and the identification of performance bottlenecks in corresponding space are completed by genetic algorithm. Therefore, it can guide the developers of real-time rendering system to adjust the parameters of rendering algorithm to achieve high rendering efficiency.