On August 9th, SNEC 14th (2020) International Solar Photovoltaic and Smart Energy (Shanghai) Exhibition and Forum, CGC Beijing Accreditation Center (hereinafter referred to as "Accreditation") and Huawei Technologies Co., LTD. (hereinafter referred to as "Huawei") jointly issued the White Paper on DC Arc Intelligent Detection (AFCI) Technology of Photovoltaic Power Generation System and the White Paper on Photovoltaic Cluster IV Scanning and Intelligent Diagnosis Technology. The first AFCI Performance Highest Level certificate and the first IV Diagnostic Performance Highest Level certificate were issued to Huawei.

The photovoltaic power generation systems pull dc arc intelligent detection (AFCI) technical white paper AFCI technology are introduced in detail the development background, technical principle, technical difficulties and characteristics, the existing technology of verification and assessment results and application prospects, but for power station development enterprises and related party accurately understand and use with the function of AFCI products for reference, To guide the construction of power plants to a higher level of safety. Dc arc drawing is the biggest security threat of roof distributed PV. Conventional DC arc drawing detection scheme is easy to miss and false positive under complex noise environment. AFCI intelligent arc detection scheme, through AI and unknown spectrum self-learning in different application environments, can effectively improve noise adaptability, distinguish noise and arc characteristics, no false positives and no missed positives, and realize accurate arc and fault location.

To comprehensively assess the level of AFCI performance, the CGC scale is based on CGC/GF 175: Huawei AFCI technology and inverters embedded with AFCI function are strictly reviewed and verified in 2020 Technical Specification for Grade Evaluation of Arc Detection and Fast Shutdown Performance and UL 1699B-2018 Safety Standard for Photovoltaic DC Arc Fault Circuit Protection. The results show that Huawei inverters with AFCI function meet the requirements of UL 1699B-2018 standard, and the performance level reaches CGC/GF 175:2020 "L4", which is the highest level in the industry.

Zhou Tao, General Manager of Huawei Smart PV Distributed Business Unit, said: "Huawei AI BOOST AFCI intelligent arc detection solution effectively solves the problems of noise and environmental adaptability, and reaches the highest level in the industry in terms of coverage distance, alarm accuracy, protection speed, accurate positioning, etc. L4 highest level certificate also reflects Huawei's leading AFCI technology."

At the press conference, Ji Zhenshuang, deputy director of CGC Jianheng, said: safety is the sky, and the power station without security guarantee will return to zero. According to the investigation results of several photovoltaic fire accidents by CGC, most of the fire accidents caused by the power station itself are caused by DC arc pulling. The application of AFCI technology to implement accurate detection and fast shutdown of DC arc pull, to prevent the occurrence of non-combustion, is a realistic and objective need to improve the intrinsic safety of power station. In order to verify the performance level of AFCI technology, CGC has organized industry forces to develop the AFCI performance grade evaluation specification, centering on the application requirements of "accurate detection and rapid disposal", in CGC/GF 175: In 2020 Technical Specification for Grade Evaluation of Arc Detection and Fast Shutdown Performance, a series of indicators are specified in terms of detection range, detection accuracy and response ability. The technical review and performance verification show that Huawei AFCI technology meets or exceeds the requirements of UL 1699B-2018 in all performance indicators, which is industry-leading, especially in the detection range and adaptability.

The pv group series IV scanning and intelligent diagnosis technology white paper summarizes a new generation of intelligent IV intelligent diagnosis technology for power plant operations industry contribution, the blessing and the AI technology to help reconstruct IV photovoltaic power station operational experience, improve operational efficiency are analyzed in detail, can provide the basis for IV diagnostic technology application and depth development.

Photovoltaic power stations are characterized by multi-faceted and complex faults. How to efficiently, timely and accurately locate and eliminate faults is one of the difficult problems to be solved in power station operation and maintenance. Intelligent IV diagnosis technology makes full use of cluster current and voltage data collected by cluster inverter, combines with big data mining and AI recognition algorithm to accurately locate cluster faults. Intelligent IV diagnosis can realize one-click start, complete the full cluster detection of 100MW power station within 15 minutes, and automatically generate fault diagnosis report.

According to CGC/GF 180:2020 Technical Specification for Photovoltaic Cluster I-V Scanning and Intelligent Diagnosis Evaluation, CGC conducted scientific, systematic and rigorous technical review and performance verification on Huawei intelligent IV diagnosis technology. The results show that Huawei intelligent IV diagnosis technology has a comprehensive performance level of "L4". This is the highest level in the industry, can be said to be "ahead of the industry generation".

Kang Lei, general manager of Huawei Intelligent PV Management System, said: "Intelligent IV diagnosis, which integrates AI technology, subverts the traditional manual on-site mode and opens the zero-inspection mode. Huawei will continue to innovate, bring more black technology and create more value to customers and the industry."

At the press conference, Ji Zhenshuang, deputy director of CGC Jianheng, said: It is the general trend to comprehensively improve the intelligent level of photovoltaic power stations. Mature IV diagnosis will play a central role in the operation and maintenance of power stations, and it is indispensable in the next step of photovoltaic intelligent construction. In order to fully verify the performance level of IV diagnostic technology, CGC has organized industry forces to develop IV diagnostic technology evaluation specifications. Centering on the core application requirements of "precise scanning, complete recognition and accurate breaking", the "specifications" have determined five key evaluation indicators, such as the scanning accuracy up to standard rate, complete recognition rate and fault diagnosis accuracy rate. Through systematic and comprehensive verification, Huawei IV diagnostic technology has filled in some technical gaps, and all performance indicators have reached the industry-leading level, which has the technical foundation for large-scale application.

Image credit: International Energy Information Network

8月9日，在SNEC第十四届(2020)国际太阳能光伏与智慧能源(上海)展览会暨论坛上，CGC北京鉴衡认证中心（以下简称“鉴衡”）与华为技术有限公司（以下简称“华为”）联合发布《光伏发电系统直流拉弧智能检测（AFCI）技术白皮书》、《光伏组串IV扫描与智能诊断技术白皮书》，并向华为颁发首张AFCI性能最高等级认证证书、首张IV诊断性能最高等级认证证书。

《光伏发电系统直流拉弧智能检测（AFCI）技术白皮书》详尽地介绍了AFCI技术的开发背景、技术原理、技术难点和特点、现有技术的验证和评估结果及应用前景，可为电站开发企业及相关方准确了解和使用带AFCI功能的产品提供参考，引导电站建设走向更高安全等级。

直流拉弧是屋面分布式光伏最大的安全威胁，常规直流拉弧检测方案在复杂的噪声环境下容易漏报、误报。AFCI智能电弧检测方案，通过AI和不同应用环境下未知频谱自学习，可有效提升噪声适应性，并区分噪声和电弧特征，不误报和不漏报，还可实现电弧及故障精准定位。

为全面评估AFCI性能水平，CGC鉴衡依据CGC/GF 175：2020《电弧检测及快速关断性能等级评价技术规范》、UL 1699B-2018《光伏直流电弧故障电路保护安全标准》，对华为AFCI技术及嵌入AFCI功能的逆变器进行了严格的评审和验证。结果表明：华为带AFCI功能的逆变器满足UL 1699B-2018标准要求，性能水平均达到CGC/GF 175：2020 “L4”，此等级为目前行业最高等级。

华为智能光伏分布式业务部总经理周涛讲道：“华为AI BOOST AFCI智能电弧检测方案，有效地解决了噪声、环境适应性方面的难题，在覆盖距离、告警准确性、保护速度、精准定位等方面达到行业最高水平，L4最高等级证书也正体现了华为在AFCI技术上的领先。”

发布会上，CGC鉴衡副主任纪振双表示：安全为天，没有安全保障的电站一切归0。根据CGC鉴衡对多起光伏火灾事故的调查结果，电站本身原因导致的火灾事故，大多为直流拉弧引起。应用AFCI技术，实施直流拉弧的精准检测和快速关断，防患于未燃，是提高电站本质安全度的现实和客观需要。为验证AFCI技术的性能水平，CGC鉴衡组织行业力量制定了AFCI性能等级评价规范，围绕“精准检测、快速处置”的应用需求，在CGC/GF 175：2020《电弧检测及快速关断性能等级评价技术规范》中，就检测范围、检测精度、响应能力等方面规定了系列指标。通过技术评审和性能验证，华为AFCI技术各项性能指标均达到或优于UL 1699B-2018的要求，处于行业领先水平，特别在检测范围和适应性方面。

《光伏组串IV扫描与智能诊断技术白皮书》总结了新一代智能IV诊断技术对于电站智能化运维的行业贡献，并就AI加持IV技术如何助力重构光伏电站运维体验、提升电站运维效率进行了详细分析，可为IV诊断技术应用及深度开发提供依据。

点多面广、复杂多样是光伏电站的故障特点，如何高效、及时、精准地定位并消除故障，是电站运维需要破解的难题之一。智能IV诊断技术充分利用组串式逆变器采集的组串电流及电压数据，结合大数据挖掘及AI识别算法，精准定位组串故障。智能IV诊断可实现一键式启动，15分钟内完成100MW电站全量组串检测，并自动生成故障诊断报告。

CGC鉴衡依据CGC/GF 180:2020《光伏组串I-V扫描与智能诊断评价技术规范》，对华为智能IV诊断技术进行了科学、系统、严谨的技术评审和性能验证，结果表明：华为智能IV诊断技术，综合性能水平达到“L4”。这是目前业界最高水准，可以说“领先行业一代”。

华为智能光伏管理系统业务总经理康磊提到：“融合了AI技术的智能IV诊断，颠覆了传统的人工上站方式，开启了零巡检的模式。华为将持续创新，给客户和行业带来更多黑科技，创造更多价值。”

发布会上，CGC鉴衡副主任纪振双表示：全面提升光伏电站智能化水平是大势所趋，成熟的IV诊断在电站运维中将起到中枢作用，在下一步光伏智能化建设中不可或缺。为全面验证IV诊断技术的性能水平，CGC鉴衡组织行业力量制定了IV诊断技术评价规范，围绕“扫得精、识得全、断得准”的核心应用需求，“规范”中确定了扫描精度达标率、识全率、故障诊断准确率等5项关键评价指标。通过系统、全面地验证，华为IV诊断技术填补了某些方面的技术空白，各项性能指标均达到行业领先水平，已具备大规模应用的技术基础。