Synchrotron tomography experiments are transitioning into multifunctional,cross-scale,and dynamic characterizations,enabled by new-generation synchrotron light sources and fast developments in beamline instrumentation...Synchrotron tomography experiments are transitioning into multifunctional,cross-scale,and dynamic characterizations,enabled by new-generation synchrotron light sources and fast developments in beamline instrumentation.However,with the spatial and temporal resolving power entering a new era,this transition generates vast amounts of data,which imposes a significant burden on the data processing end.Today,as a highly accurate and efficient data processing method,deep learning shows great potential to address the big data challenge being encountered at future synchrotron beamlines.In this review,we discuss recent advances employing deep learning at different stages of the synchrotron tomography data processing pipeline.We also highlight how applications in other data-intensive fields,such as medical imaging and electron tomography,can be migrated to synchrotron tomography.Finally,we provide our thoughts on possible challenges and opportunities as well as the outlook,envisioning selected deep learning methods,curated big models,and customized learning strategies,all through an intelligent scheduling solution.展开更多
Purpose The control network is a critical infrastructure that supports the stable operation of the accelerator.Each device accessing the control network has different device information,such as IP address,MAC address,...Purpose The control network is a critical infrastructure that supports the stable operation of the accelerator.Each device accessing the control network has different device information,such as IP address,MAC address,connected switches and ports,device location and purpose.Accurately maintaining the mapping relationship between these device information facilitates network management.It helps inventory assets,fault location and provide visibility into dynamic changes of device on the network.However,existing tools cannot fully satisfy these demands.They only map some information and lack details important for accelerator facilities like device location and purpose.Additionally,they only reflect the current status rather than indicating dynamic changes across all devices over time.As intelligent devices proliferate,the scale of the control network is rapidly expanding,posing greater challenges in maintaining mapping relationships.Methods This paper proposes a device information-centered Accelerator Control Network Management System(ACNMS).It establishes a device information management framework and allows network administrators to perceive the dynamic changes of devices on the network.The system adopts a layered architecture.Back-end modules implement the core logic of all functions.The graphical user interface presents data and provides a management portal.Results The system test on the control network of the National Synchrotron Radiation Laboratory demonstrates that it can meet the functional design objectives.The application scenarios of the ACNMS are further expanded through system integration and combination with network automation.Conclusion The ACNMS has proven to be an efficient network management tool that significantly improves the operation and maintenance efficiency of the accelerator control network.展开更多
基金This work was funded by the National Science Foundation for Young Scientists of China(grant 12005253)the Strategic Priority Research Program of the Chinese Academy of Sciences(XDB 37000000)+1 种基金the Innovation Program of the Institute of High Energy Physics,CAS(E25455U210)the Hefei Science Center,Chinese Academy of Sciences(award 2019HSC-KPRD003).All authors gratefully acknowledge support from the BL13HB and BL16U2 beamlines of the Shanghai Synchrotron Radiation Facility(SSRF)and BL07W beamline of the National Synchrotron Radiation Laboratory(NSRL)。
文摘Synchrotron tomography experiments are transitioning into multifunctional,cross-scale,and dynamic characterizations,enabled by new-generation synchrotron light sources and fast developments in beamline instrumentation.However,with the spatial and temporal resolving power entering a new era,this transition generates vast amounts of data,which imposes a significant burden on the data processing end.Today,as a highly accurate and efficient data processing method,deep learning shows great potential to address the big data challenge being encountered at future synchrotron beamlines.In this review,we discuss recent advances employing deep learning at different stages of the synchrotron tomography data processing pipeline.We also highlight how applications in other data-intensive fields,such as medical imaging and electron tomography,can be migrated to synchrotron tomography.Finally,we provide our thoughts on possible challenges and opportunities as well as the outlook,envisioning selected deep learning methods,curated big models,and customized learning strategies,all through an intelligent scheduling solution.
基金supported by Hefei Advanced Light Facility(HALF),a major national science and technology infrastructure in China.
文摘Purpose The control network is a critical infrastructure that supports the stable operation of the accelerator.Each device accessing the control network has different device information,such as IP address,MAC address,connected switches and ports,device location and purpose.Accurately maintaining the mapping relationship between these device information facilitates network management.It helps inventory assets,fault location and provide visibility into dynamic changes of device on the network.However,existing tools cannot fully satisfy these demands.They only map some information and lack details important for accelerator facilities like device location and purpose.Additionally,they only reflect the current status rather than indicating dynamic changes across all devices over time.As intelligent devices proliferate,the scale of the control network is rapidly expanding,posing greater challenges in maintaining mapping relationships.Methods This paper proposes a device information-centered Accelerator Control Network Management System(ACNMS).It establishes a device information management framework and allows network administrators to perceive the dynamic changes of devices on the network.The system adopts a layered architecture.Back-end modules implement the core logic of all functions.The graphical user interface presents data and provides a management portal.Results The system test on the control network of the National Synchrotron Radiation Laboratory demonstrates that it can meet the functional design objectives.The application scenarios of the ACNMS are further expanded through system integration and combination with network automation.Conclusion The ACNMS has proven to be an efficient network management tool that significantly improves the operation and maintenance efficiency of the accelerator control network.
