On May 22, the "2018 First International Conference on New Energy Vehicles and Power Batteries (CIBF Shenzhen)" was held at the Shenzhen Convention and Exhibition Center. Liu Peng, Deputy Secretary-General of the National New Energy Vehicle Big Data Alliance, delivered a speech at the "New Energy Vehicle Special Session – Understanding the Transformation Path of Vehicle-Electric Partnership under New Policies" forum. The following is the content of his speech:
Liu Peng, Deputy Secretary-General of the National New Energy Vehicle Big Data Alliance
Today I will mainly talk about the issue of electric vehicle power batteries from the perspective of big data, focusing on four aspects.
New energy vehicle big data
Currently, our platform monitors 883,000 vehicles. Since January 1, 2017, new energy vehicles must meet certain requirements to be connected to the platform, including being vehicles sold within the last six months. Of these 880,000 vehicles, an average of 498,000 are online daily, with 160,000 charging sessions and an average daily mileage of 12 million kilometers, averaging 20 kilometers per day.
Our platform monitors real-time location information and battery baseline data during vehicle operation, including detailed records of vehicle speed, battery readiness, battery voltage, and temperature. Using this data, we can understand the vehicle's usage status, as well as the battery and motor's status, from different perspectives. Currently, based on the GB/T32960 standard, we monitor a total of 73 battery-related data items. For lithium batteries, this includes current, voltage, and detailed information on individual cells; for fuel cells, it includes basic information. Based on this information, we can accurately describe the user's needs and propose breakthroughs and management directions for subsequent battery research and development.
Power battery fault diagnosis
As the primary energy source for electric vehicles, the power battery system has become one of the main sources of failure in electric vehicles due to factors such as technology, operating conditions, and usage environment. Safety has now become the primary issue facing the development of electric vehicles. Power battery safety has become an international challenge and a hot research topic.
Based on big data, we utilize the national regulatory platform to conduct comprehensive analysis of massive amounts of data. The relevant evaluation standards and analysis algorithms are now quite mature. Through high-precision and high-efficiency fault early warning analysis algorithms, we can quickly and in real-time respond to determine the probability of fault occurrence, including fault level, fault category, fault frequency, fault vehicle model, and fault cause.
Based on big data, we can also combine the basic safety thresholds provided by car manufacturers to conduct vertical and horizontal data comparison and mining, and establish a high-safety remote fault diagnosis system for power battery systems; by comparing the changes in typical data entropy values, we can build a fault early warning system for power battery systems.
Through our fault diagnosis system, we classify and analyze vehicle operating status models, and conduct systematic statistical and refined analysis of power battery system fault types, fault locations, and fault occurrence frequencies to improve the efficiency and accuracy of electric vehicle fault diagnosis.
Power Battery Traceability Management
How can we effectively conduct supervision based on existing big data to support subsequent services?
There's a lot of talk about energy conservation and environmental protection these days, especially regarding batteries, which face many challenges, including how to recycle, manage, and share information. All of these require attention.
If we can standardize and summarize the information on battery production and replacement during use, we can determine how to handle each individual battery cell and how to assess its value, which is very beneficial for subsequent utilization and dismantling.
In addition, the monitoring platform can also interact and analyze data from multiple levels, including battery, vehicle, user, and recycling.
Power Battery Value Assessment
How should the pricing of battery reuse and recycling be determined? This requires the support of big data to effectively assess the battery's energy state, power state, and applicable fields. These are the key areas we will focus on in our future research. The core approach is still based on big data analysis to assess the battery's health status, which will then support the application of subsequent battery systems.
Taking the battery capacity degradation rate over 10,000 kilometers as an example, we combined user operating data and used mileage within the 50%-80% range to compare the battery degradation rate over 10,000 kilometers. This data range has the largest volume and the most usage intervals, and it reflects the degradation process throughout the entire process, which is also what users are most concerned about.
We analyzed the results of four models from two manufacturers. The best-performing model had a degradation rate of 1.4% per 10,000 kilometers, while the worst had 8.2%. This degradation rate is not solely due to battery issues, but also to problems with the management system, which directly affects the battery's lifespan and performance in the vehicle.
Regarding the actual driving range and the nominal range, we compared the actual distance traveled per kilowatt-hour consumed with the nominal value, and the result is the index value. This index is also affected by driving habits; many people use air conditioning, which leads to changes in user feedback data. In other words, this result is related to both the battery itself and environmental adaptability.
In addition, there is the distribution of SOC values at the start of charging. This data can reflect drivers' confidence in the range and passenger capacity of new energy vehicles, as well as the construction status of charging stations and other infrastructure.
