Promoting grid innovation and upgrading is what Intel has been doing all along.
”
Recently, you may also be paying attention to the frequent "power ration incidents" across the country. Electricity is not only related to the national economy and people's livelihood, it is also the engine that drives the world and accelerates innovation. In fact, in order to alleviate the energy gap while addressing global climate change, China's renewable energy generation has been at the forefront of the world. As of the end of September, the installed capacity of wind power in China increased by 32.8% year-on-year, and the installed capacity of solar power generation increased by 24.6% year-on-year[1]. Renewable energy generation is also a global megatrend.
Changes in the way of power generation promote the evolution of the flow of electrical energy from one-way to two-way. Of course, this requires intelligent transformation of the power grid, including transmission and distribution stations, while the maintenance and upgrading of traditional power equipment requires a lot of cost and time.
Intel is working with eco-partners to modernize substations using smart edge architectures with innovative technologies such as the Internet of Things, 5G, and big data, ensuring the digital transformation and grid modernization of grid companies from the new infrastructure level. Let the smart grid become a strong driving force for China's economic development.
Let's take a look at how the power infrastructure under Intel technology support implements three major upgrades: intelligent, software-defined, and data-driven.
1
Detection voltage sag AI edge enablement
Voltage sag is a serious power quality problem, which will cause equipment failure, and if it is serious, it will lead to equipment damage and cause major economic losses. However, the traditional detection method has the problem of endpoint effect, which reduces the measurement accuracy.
Based on edge computing and cloud computing architecture, Intel proposes real-time voltage sag monitoring and governance schemes. The solution's edge-side devices use 2nd-generation Intel Xeon Scalable processors and FPGAs to analyze the collected data using artificial intelligence models to quickly and accurately locate the start and end of voltage sags with lightweight inference, and calculate voltage amplitude and frequency. At the same time, the performance of the edge-side heterogeneous platform is further improved.
Grid voltage analysis and control system based on Intel cloud computing
The solution's back-end cloud platform also uses Intel Xeon Scalable processors, supported by tool frameworks such as Analytics Zoo and OpenVINO Toolkit, Pytorch optimized for Intel architecture, and the edge side constitutes an end-to-end AI detection solution, and at the same time, a variety of deep learning and machine learning algorithm models are also deployed, which are faster detection speed, higher accuracy, and high load requirements for model inference. And push the results to the grid voltage control system.
Based on Intel's edge computing and cloud computing, this solution has the characteristics of short latency, high performance, and more accurate measurement results, thereby improving the efficiency of voltage sag management.
2
5G MEC combines hard and soft to eliminate the hidden danger of equipment overheating
In the past, under blue skies, power engineers could often be seen working on tall power towers. Although it is indeed very handsome, such a situation is undoubtedly dangerous. Today, the power equipment fault diagnosis system based on the Intel 5G MEC architecture and the OpenVINO toolkit is deployed, which can quickly analyze temperature changes and accurately determine the fault point, thereby eliminating the safety hazards of power equipment and timely discovering, processing, and preventing the occurrence of major accidents.
Intel-based fault diagnosis system architecture
In this system, the 5G MEC edge cloud based on Intel hardware and software is composed of 2nd generation Intel Xeon Scalable processors, Intel FlexRAN reference architecture, OpenVINO toolkit, and open edge service software OpenNESS, which is responsible for high-performance workloads such as data analysis, AI model building and inference, and model push. The flexible and open Intel FlexRAN reference architecture enables a common x86 architecture server platform to achieve the performance of traditional access network BBU, providing high-throughput, low-latency 5G wireless connectivity in different application scenarios. The OpenNESS platform can obtain a unified application development and hosting environment in the heterogeneous network environment of public network and private network, making it easier for power companies to migrate and deploy intelligent applications to the edge.
Intel products and technologies help build 5GMEC solutions
3
The open source edge insight platform gives the world's problems nowhere to hide
In power systems, there is another world problem: small current ground faults. Due to the vague current characteristics and the wide variety of fault types, the traditional detection method has low accuracy, which is difficult to meet the requirements of power grid automation and high reliability.
To accelerate the rapid deployment of industrial applications on edge platforms, Intel developed the Intel Industrial Edge Insights Software Platform Framework and open sourced it. It integrates time-sensitive data injection and analysis components, which can be deployed on the edge device closest to the data source, including the four open source components of TICK, completing the four key steps of data collection, data storage, data processing, and data display, and realizing the landing and productization of the algorithm.
Industrial Edge Insights Software Platform Software Block Diagram
Practical application shows that on the basis of ensuring 100% judgment accuracy, this method increases the judgment speed from 100 milliseconds to 10 milliseconds, which greatly improves the judgment efficiency, not only prevents the spread of grounding accidents, enhances the safety level of new energy projects, but also maintains photovoltaic power plants to continue to generate electricity, and improves economic benefits.
4
Work together to accelerate digital transformation and build a power grid ecosystem
As the energy market embraces digital transformation, power systems will evolve from analog architectures to new digital power system standards. Driven by China's new infrastructure construction and "double carbon" economy, strong smart grids and energy Internet of Things are booming. Intel will continue to be committed to promoting the accelerated transformation of China's power industry, catalyzing the organic integration of the power world and the digital world, and working with partners to build a new ecosystem across the power value network.
END