Development of a Fast Camera Diagnostic Platform for Real-time Control Applications

快速相机是研究聚变装置中等离子体物理的重要诊断仪器。目前基于相机的系统被用来研究湍流、等离子体成分、密度、排气温度以及聚变反应堆控制的所有关键方面。目前，快速相机平台要么必须从头开始建造，要么从供应商那里购买。从头开始构建相机平台需要大量的开发时间，但会产生专门针对用例的系统。从供应商那里购买摄像头可以快速实施系统，但由于专有软件平台的原因，可能会限制功能和未来的升级途径。许多快速摄像头系统使用称为GigE Vision的通信协议来连接到控制PC。该协议可以免费许可用于研究目的，但目前还不存在开放源代码的实现。因此，GigE视觉系统通常只提供附带的专有控制和采集软件。使用GigE VISION的开源和用户友好平台将有助于快速创建可用于各种可用相机的高度定制的仪器。现场可编程门阵列(现场可编程门阵列)是一种集成电路，能够在裸逻辑级别对其进行编程，以专门用于任何任务。由于其通用性，基于现场可编程门阵列的系统非常适合构建定制采集和实时处理系统，而无需经历开发专用集成电路(ASIC)的电气工程方面。为一个FPGA编写的固件可以应用并集成到其他FPGA平台中，并作为开放源码软件提供。该项目的目的是开发一个使用GigE视觉协议的灵活的快速摄像头诊断平台。在现场可编程门阵列上实现GigE VISION协议将允许开发用于实时处理应用的低延迟快速相机仪器。该诊断平台可应用于聚变研究的许多领域，如灯丝跟踪、等离子体整形、实时保护或等离子体边缘物理研究。虽然鼓励设计的原理证明应适合于不同诊断的要求，但该项目的建议实施是新版本的高分辨率边缘半径光学测量(HOMER)诊断。与其当前状态相比，HOMER诊断具有增加功能的潜力，其实施将展示该平台的多功能性。建议探索的新想法是实时H模事件触发，用于等离子体边缘轮廓研究的实时边缘电子或中性气体密度测量，或用于实时等离子体边界成像和实时保护的2D相机实现。该项目将在一系列计算和实验领域培养技能，并提供现场可编程门阵列开发、等离子体控制方面的专业知识和等离子体物理方面的广泛知识。

Fast cameras are an important diagnostic instrument for studying plasma physics in fusion devices. Current camera based systems are used to study turbulence, plasma composition, density, exhaust temperature, and shaping all crucial aspects for the control of a fusion reactor. Currently fast camera platforms must either be built from scratch or purchased from a supplier. Building a camera platform from scratch requires a lot of development time but results in specialised system specific for the use case. Buying a camera from a supplier can result in implementing a system quickly but can limit functionality and future upgrade paths due to proprietary software platforms.Many fast camera systems use a communication protocol called GigE Vision to connect to a controlling PC. This protocol can be licensed for free for research purposes but there currently does not exist an open source implementation. Therefore GigE vision systems are generally only available with accompanying proprietary control and acquisition software. An open-source and user friendly platform using GigE vision would help in quickly creating highly customised instrumentation that can be used for a wide range of available cameras. Field Programmable Gate Arrays (FPGA) are integrated circuits capable of being programmed at the bare logic level to specialise them to any task. Because of their generic nature, FPGA based systems are ideal for building custom acquisition and real-time processing systems without going through the electrical engineering aspects of developing an Application Specific Integrated Circuit (ASIC). Firmware written for an FPGA can be applied and integrated into other FPGA platforms and made available as open-source software. The aim of this project is to develop a flexible fast camera diagnostic platform using the GigE vision protocol. Implementing the GigE vision protocol on FPGA will allow the development of low latency fast camera instrumentation for real-time processing applications. The diagnostic platform could be applied to many areas of fusion research, such as filament tracking, plasma shaping, real-time protection, or plasma edge physics studies. While it is encouraged that the designed proof-of-principle should be suitable to accomodate requirements from different diagnostics, the suggested implementation for this project is a new version of the Hi-Res Optical Measurement of Edge Radius (HOMER) diagnostic. The HOMER diagnostic has potential for increased functionality, compared to its current state, the implementation of which will demonstrate the versatility of the platform. Suggested novel ideas to explore are real-time H-mode event triggering, real-time edge-electron or neutral gas density measurements for plasma edge profile studies, or a 2D camera implementation for real-time plasma boundary imaging and real-time protection. The project will develop skills in a range of computational and experimental areas, as well as provide expertise in FPGA development, plasma control, and a broad knowledge in plasma physics.