Neural network modeling of safety system for construction equipment operation in permafrost zone

4th International Scientific Conference "Arctic: History and Modernity"

生态环境科学;历史学

Idrisova, J.I.^1^2^3 ; Kaverzneva, T.T.^1 ; Rumyantseva, N.V.^2 ; Skripnik, I.L.^3

Peter the Great Saint-Petersburg Polytechnic University, St-Petersburg, Russia^1

St Petersburg Academic University, St-Petersburg, Russia^2

St. Petersburg University of the Ministry of Emergency Situations, St-Petersburg, Russia^3

关键词: Climatic conditions;    Computing Experiments;    Construction equipment operations;    Engineering problems;    Initial information;    Neural network model;    Periodic components;    Temperature fluctuation;   

Others  :  https://iopscience.iop.org/article/10.1088/1755-1315/302/1/012128/pdf
DOI  :  10.1088/1755-1315/302/1/012128

学科分类：环境科学（综合）

来源: IOP

【 摘 要 】

The problems of neural network modeling of working conditions securing system while operating constructional equipment in permafrost zone are considered. Determining the temperature fields distribution on the soil depth on air temperature in regions with harsh climatic conditions will allow us to solve a number of engineering problems directed to creation of safe working conditions at construction equipment operation. As a first approximation temperature on the surface and in the depth of the upper layer of earth changes under the periodic law, following the change of air temperature during the year. With soil depth increasing the amplitude of temperature fluctuations decreases, but a random component related to weather conditions (the impact of which could be significant) is added to the periodic component caused by change of a season. In the article the mathematical model in the form of Stefan's problem in which boundary conditions on earth surface are replaced with results of measurements is considered. Methods of neural network creation of this problem solution and results of computing experiments are given. The received results show that neural networks are the flexible tool, allowing to consider featuring of a task and all available information. Thus accuracy of results corresponds to accuracy of initial information. The additional information can be effectively used for specification of the required decision.