preface
Foreword
Industrial intelligent computing networkIn response to the national policy of new industrialization, Beijing Taier Yingfu Technology and Beijing National Center for Applied Mathematics jointly launched the concept on the basis of the practice of serving the digital transformation of enterprises. "Future Telescope - Industrial Intelligent Computing Network Series".This article is the last issue of the industrial intelligent computing network series, thank you for your attention!
hazards
Hazards of industrial disasters
There are two main types of sudden disasters that cause harm to the industrial field: accident disasters and natural disasters.
With the continuous advancement of industrialization and urbanization, China's population and production factors are more concentrated, the exposure, concentration and vulnerability of various disaster-bearing bodies have increased significantly, and safety risks such as urban waterlogging, fire, and gas leakage have become increasingly prominent.
With global warming, extreme weather tends to be stronger and more frequent, resulting in an increase in the possibility of floods, droughts, high temperatures, rain, snow and ice, and forest fires.
At the same time, the concealment, complexity and coupling of disaster accidents have further increased, and major disasters often lead to a series of secondary and derivative disasters and ecological environment damage, forming complex and diverse disaster chains and accident chains, which increase the complexity and difficulty of risk prevention and control and emergency response.
key technologies
Key technologies for industrial disaster scenario construction and early warning
Major emergencies in the industrial field have always been the top priority of public safety and emergency management in China. Scenario construction is an important method and starting point for carrying out "the picture before it sprouts, and the virtual one does not exist" for the expected risk based on the bottom-line thinking.
Scenario building can be used as a strategic risk management tool for the industry. On the basis of the risk research and judgment of major emergencies in the industry, the scenario list is determined and the scenario construction is carried out for each typical risk, so as to guide the emergency preparedness action, including the optimization of the emergency plan system and the planning of emergency drills.
The technical route and method of scenario construction are generally based on the screening of risk topics, and are carried out with the main line of "scenario-task-capability". Its core key technology is the disaster evolution mechanism, no two disasters are exactly the same, but the disaster law can be followed, and the underlying support of the law is the evolution mechanism, so that the scenario construction can not only "know why it is so", but also "respond to all changes with the same".
For example, the core of the evolution mechanism of natural disasters such as fires, floods, toxic gas diffusion, dam failures, and debris flows is fluid mechanics, and the means of evolution of these mechanisms are usually numerical simulation techniques with computational fluid dynamics algorithms as the core. Numerical simulation-based science** is also the only method that can provide useful information for the "four predictions" (forecasting, early warning, rehearsal, and plan) and supporting decision-making in extreme situations.
Technical route of scenario construction (citation: Wang Yongming, Theory and Practice of Scenario Construction for Major Emergencies, National Administration Press, 2020).
data tools
The industrial intelligent computing network provides data tools for the construction of disaster scenarios
Based on the industrial Internet platform, the industrial intelligent computing network integrates technologies such as identification analysis and privacy computing of industrial data to ensure data security and trustworthiness, brings together all elements of computing power, data, algorithms and application scenarios, and can provide a complete digital infrastructure for the construction and early warning of industrial emergency disaster scenarios. In particular, as a national digital infrastructure in the field of industrial Internet in China, the industrial Internet identification resolution system can ensure the security and credibility of data.
The scenario construction of industrial disaster scenarios is combined with industrial datasets to decouple data and scenarios to a certain extent, which can facilitate the rapid construction of scenarios under different working conditions in the same scenario. In order to meet the needs of accident Xi, generative AI algorithms can also be combined to quickly generate relevant scenarios with high realism, which can be supported by disaster prevention and Xi.
In addition, all kinds of real sensor data in the scene can use the inversion tools provided by the industrial intelligent computing network to generate scene data of real accidents and working conditions, and then combined with simulation technology, the corresponding scene data set can be formed, and the foundation for subsequent data analysis can be laid.
mechanism engine
The industrial intelligent computing network provides a mechanism engine for disaster scenarios
The technical core of the construction of industrial sudden disaster scenarios is the disaster evolution mechanism, and the simulation and development of the disaster evolution mechanism are the key. The simulation of the evolution mechanism of disasters itself requires the support of a large amount of computing power, data, and algorithms, and with the breakthrough of general AI models, the combination of industrial AI technology and simulation technology of disaster mechanism has become the most important technical means in the field.
The industrial intelligent computing network provides various types of mechanism models and algorithms for disaster scenarios, which is convenient for early warning and early warning. For example, the hydrodynamic model can provide early warning for scenarios such as urban waterlogging and river basin bursting under extreme weather. The correlation model of debris flow and landslide can provide algorithm support for geological disaster scenarios in mountainous areas.
Simulation of debris flow in Zhouqu, Gansu Province.
references
References
1. Wang Yongming, Theory and Practice of Scenario Construction of Major Emergencies, National Administration Press, 2020
2, Yuan Li*, Liu Wei, Zhai Jian, Improvement and Application of Titan2D, an open-source software for slope particle flow, Mountain Research, 34 (3):346-355, 2016
3， li yuan*, wei liu, jian zhai, shifeng wu, a.k. patra, e.b.pitman, refinement on non-hydrostatic shallow granular flow model in a global cartesian coordinate system , computational geosciences, 22 (2018): 87-106,