In the present era of smart computing and digital transformation multiprogramming multitasking, multi-user operating environment is playing an important role where rapid arrival, termination and switching of multiple processes is crucial for performance. CPU Utilization is one of the key performance parameter and CPU scheduling crucial for CPU utilization. On the way of improving performance, substantial research and efforts are being done by community researchers and operating system processing is becoming fast day by day. Major performance evaluations criteria like waiting time, response time turnaround time number of context switching are sophisticatedly enhanced to achieve better processing. Round-Robin (RR) scheduling is a well-known important scheduling algorithm where quantum time slot pointedly affect performance. In case of very large quantum time it works like First come first serve or if it is too small then RR works like a process sharing algorithm which increases the system overheads in view of context switching. Thus setting up an effective quantum time is a major challenge for researchers. In this work a clustering inspired novel intelligent approach has been proposed to significantly enhance the round robin algorithm performance. An intelligent sub ready queue by using standard deviation and optimal threshold is created. Dynamic quantum is generated for each cluster and performance is evaluated on sample data in terms of waiting time, response time, turnaround time and context switching. The results of the proposed intelligent approach are compared with other existing algorithms and are found better.

在当今的智能计算和数字转换多程序多任务处理时代，多用户操作环境扮演着重要角色，其中快速到达，终止和切换多个进程对于性能至关重要。CPU利用率是关键性能参数之一，而CPU调度对于CPU利用率至关重要。在提高性能的方式上，社区研究人员正在进行大量研究和努力，并且操作系统处理正变得越来越快。主要性能评估标准（例如等待时间，上下文切换的响应时间周转时间）得到了全面增强，以实现更好的处理。循环调度（RR）调度是一种众所周知的重要调度算法，其中量子时隙会明显影响性能。在非常大的量子时间的情况下，它的工作方式类似于“先到先服务”，或者如果它太小，则RR就像一种进程共享算法，鉴于上下文切换，它会增加系统开销。因此，设置有效的量子时间是研究人员面临的主要挑战。在这项工作中，提出了一种启发聚类的新颖智能方法，以显着提高循环算法的性能。通过使用标准偏差和最佳阈值创建智能子就绪队列。将为每个群集生成动态量子，并根据等待时间，响应时间，周转时间和上下文切换来评估示例数据的性能。将所提出的智能方法的结果与其他现有算法进行比较，发现效果更好。