Instruments and Systems: Monitoring, Control, and Diagnostics Annotation << Back Using Machine Learning in Fire Robot Control Tasks I.N. POZHARKOVA, S.G. TSARICHENKO, S.G. NEMCHINOV The article examines the effectiveness of using various machine learning methods for solving problems of controlling fi re robots, in particular, predicting the movement trajectory of a fi re extinguishing agent from a fi re monitor, including under disturbing infl uences. The results of predicting the characteristic points parameters, the upper and lower boundaries of the water jet without taking into account disturbing infl uences based on algorithms trained using the results of fi re monitors testing at the experimental site are presented. A method for predicting the trajectory of a fi re extinguishing agent jet from the fi re robot monitor, taking into account the infl uence of wind and air temperature in real time, which allows it to be used to generate appropriate control actions without delay, which can lead to a critical decrease in fi re extinguishing effi ciency, is described. The results of a comparison of a water jet trajectory from a fi re monitor with a side wind and negative air temperature, built on the basis of the described method and obtained experimentally, are presented. Keywords: fi re robots, jet trajectory prediction, machine learning, multilayer perceptron, neural network, computational fl uid dynamics, CFD. DOI: 10.25791/pribor.12.2023.1460 Pp. 19-25.