Search results

(1 - 2 of 2)
Data-Driven Methods for Soft Robot Control and Turbulent Flow Models
Title
Data-Driven Methods for Soft Robot Control and Turbulent Flow Models
Creator
Lopez, Esteban Fernando
Date
2022
Description
The world today has seen an exponential increase in its usage of computers for communication and measurement. Thanks to recent technologies,...
Show moreThe world today has seen an exponential increase in its usage of computers for communication and measurement. Thanks to recent technologies, we are now able to collect more data than ever before. This has dawned a new age of data-driven methods which can describe systems and behaviors with increasing accuracy. Whereas before we relied on the expertise of a few professionals with domain-specific knowledge developed over years of rigorous study, we are now able to rely on collected data to reveal patterns, develop novel ideas, and offer solutions to the world’s engineering problems. No domain is safe. Within the engineering realm, data-driven methods have seen vast usage in the areas of control and system identification. In this thesis we explore two areas of data-driven methods, namely reinforcement learning and data-driven causality. Reinforcement learning is a method by which an agent learns to increase its selection of ideal actions and behaviors which result in an increasing reward. This method was applied to a soft-robotic concept called the JAMoEBA to solve various tasks of interest in the robotics community, specifically tunnel navigation, obstacle field navigation, and object manipulation. A validation study was conducted to show the complications that arise when applying reinforcement learning to such a complex system. Nevertheless, it was shown that reinforcement learning is capable of solving three key tasks (static tunnel navigation, obstacle field navigation, and object manipulation) using specific simulation and learning hyperparameters. Data-driven causality encompasses a range of metrics and methods which attempt to uncover causal relationships between variables in a system. Several information theoretic causal metrics were developed and applied to nine mode turbulent flow data set which represents the Moehlis model. It was shown that careful consideration into the method used was required to identify significant causal relationships. Causal relationships were shown to converge over several hundred realizations of the turbulent model. Furthermore, these results match the expected causal relationships given known information of self-sustaining processes in turbulence, validating the method’s ability to identify causal relationships in turbulence.
Show less
Intelligent Job Scheduling on High Performance Computing Systems
Title
Intelligent Job Scheduling on High Performance Computing Systems
Creator
Fan, Yuping
Date
2021
Description
Job scheduler is a crucial component in high-performance computing (HPC) systems. It sorts and allocates jobs according to site policies and...
Show moreJob scheduler is a crucial component in high-performance computing (HPC) systems. It sorts and allocates jobs according to site policies and resource availability. It plays an important role in the efficient use of system resources and users satisfaction. Existing HPC job schedulers typically leverage simple heuristics to schedule jobs. However, the rapid growth in system infrastructure and the introduction of diverse workloads pose serious challenges to the traditional heuristic approaches. First, the current approaches concentrate on CPU footprint and ignore the performance of other resources. Second, the scheduling policies are manually designed and only consider some isolated job information, such as job size and runtime estimate. Such a manual design process prevents the schedulers from making informative decisions by extracting the abundant environment information (i.e., system and queue information). Moreover, they can hardly adapt to workload changes, leading to degraded scheduling performance. These challenges call for a new job scheduling framework that can extract useful information from diverse workloads and the increasingly complicated system environment, and finally make well-informed scheduling decisions in real time.In this work, we propose an intelligent HPC job scheduling framework to address these emerging challenges. Our research takes advantage of advanced machine learning and optimization methods to extract useful workload- and system-specific information and to further educate the framework to make efficient scheduling decisions under various system configurations and diverse workloads. The framework contains four major efforts. First, we focus on providing more accurate job runtime estimations. Estimated job runtime is one of the most important factors affecting scheduling decisions. However, user provided runtime estimates are highly inaccurate and existing solutions are prone to underestimation which causes jobs to be killed. We leverage and enhance a machine learning method called Tobit model to improve the accuracy of job runtime estimates at the same time reduce underestimation rate. More importantly, using TRIP’s improved job runtime estimates boosts scheduling performance by up to 45%. Second, we conduct research on multi-resource scheduling. HPC systems are undergoing significant changes in recent years. New hardware devices, such as GPU and burst buffer, have been integrated into production HPC systems, which significantly expands the schedulable resources. Unfortunately, the current production schedulers allocate jobs solely based on CPU footprint, which severely hurts system performance. In our work, we propose a framework taking all scalable resources into consideration by transforming this problem into multi-objective optimization (MOO) problem and rapid solving it via genetic algorithm. Next, we leverage reinforcement learning (RL) to automatically learn efficient workload- and system-specific scheduling policies. Existing HPC schedulers either use generalized and simple heuristics or optimization methods that ignore workload and system characteristics. To overcome this issue, we design a new scheduling agent DRAS to automatically learn efficient scheduling policies. DRAS leverages the advance in deep reinforcement learning and incorporates the key features of HPC scheduling in the form of a hierarchical neural network structure. We develop a three-phase training process to help DRAS effectively learn the scheduling environment (i.e., the system and its workloads) and to rapidly converge to an optimal policy. Finally, we explore the problem of scheduling mixed workloads, i.e., rigid, malleable and on-demand workloads, on a single HPC system. Traditionally, rigid jobs are the main tenants of HPC systems. In recent years, malleable applications, i.e., jobs that can change sizes before and during execution, are emerging on HPC systems. In addition, dedicated clusters were the main platforms to run on-demand jobs, i.e., jobs needed to be completed in the shortest time possible. As the sizes of on-demand jobs are growing, HPC systems become more cost-efficient platforms for on-demand jobs. However, existing studies do not consider the problem of scheduling all three types of workloads. In our work, we propose six mechanisms, which combine checkpointing, shrink, expansion techniques, to schedule the mixed workloads on one HPC system.
Show less