Academician of the Academy of Engineering Chaiten: Large data for the study of industrial complex system dynamic behavior mechanism to provide

Chaiten, academician of Chinese Academy of Engineering, said at the China computer Conference in Saturday that large industrial data should open up possible avenues for the study of the dynamic behavior of complex systems with unclear mechanisms in the industrial field, while the study of dynamic behavior of complex systems would lead to higher productivity and other advantages.

Monitor complex systems for sensors, video monitoring and other data collectors bring different types of data groups, before large data applications, in addition to structured data, other types of data is difficult to use machine analysis to mining the application value, and the current "large data application technology, modeling technology and simulation technology, such as the prediction of dynamic performance is possible." "Chaiten said.

He cited, for example, the wind turbine maker Vestas's crossover analysis of weather data and turbine instrumentation data, and improvements in the layout of wind turbines, which increased the power output of wind turbines and extended service life.

Chaiten says industrial data is similar to the traditional concept of large data, but it is different. In his view, the major data features in the industrial field include the characteristics of gross quantity, continuous sampling, multi-source, low value density and dynamic data. He stressed that the industrial field of large data "large" is relative to the concept of industrial processes resulting from data at multiple GB levels. Take 100,000 common sensors for example, about 8Bytes per second production data per hour about the production of 3GB data, "in fact, the capacity is not large."

Chaiten prediction, large data application will promote the dynamic model construction, safe operation and monitoring, multi-objective optimization control method in industry field.

Press: Guivit/Sun Fangfang/He Ging revisers: Li Ming selected

(Responsible editor: The good of the Legacy)