The Chinese conglomerate has announced how it will integrate Artificial Intelligence (AI) into its stringed devices this year. The approach includes converting inverters into smart PV controls, developing AI inference modules, and creating an AI training and inference platform. Machine learning is also incorporated into operation and maintenance, grid management and PV system planning.
May 7, 2020
After developing its first artificial intelligence (AI) -based solar inverter last year, the Chinese conglomerate Huawei announced how it intends to further integrate AI into its solar business.
According to Huawei, there will be "extensive integration of cloud, AI and 5G technologies" this year.
An AI roadmap document produced by the company expanded the AI integration into its string inverters by converting them into smart PV controllers, developing AI inference modules, and creating an AI training and inference platform.
The manufacturer said its string inverters will be enhanced for high-precision, real-time data acquisition with real-time control of energy yield optimization at the string level, DC arc detection and response to grid control – including real-time inference and execution and self-regulation functions.
According to the inverter manufacturer, the aforementioned AI inference devices will improve the acquisition of device data and derive AI models for optimal power generation in real time.
Huawei added that an AI training and inference platform aims to further refine AI algorithm models without changing existing equipment.
Testing
Using string inverters equipped with neural network-based AI training and modeling would enable increased power generation, the manufacturer said, citing tests carried out in PV systems in China's Anhui province. According to the Huawei AI roadmap, energy yield increased by an average of 1.31% in 183 days. In two projects of the Chinese General Nuclear Power Group Huanghen Hydropower, the energy yield improved by 0.5 to 1%.
The inverter maker said AI algorithms are also used to prevent fires, especially those caused by DC arcing caused by poor contact. According to Huawei, the AFCI (Arc-Fault Circuit Interrupters) solution is already able to modulate the operation of PV systems if a fire hazard is identified. The technology is also said to be able to learn arch features using AI models based on a database of more than 1 million arch features.
The Huawei roadmap document includes plans to improve AI capability for O&M, grid management and PV plant design. For the former, Huawei said that error processing tools built into AI technology could replace the O&M experts in the near future.
"As repetitive and dangerous tasks are handled by AI systems, these technologies free people from everyday work, reducing manpower while improving efficiency, speed and accuracy," the company said.
This content is protected by copyright and may not be reused. If you would like to work with us and would like to reuse some of our content, please contact: editors@pv-magazine.com.