International Scientific Journals
of Scientific Technical Union of Mechanical Engineering "Industry 4.0"

Existing approaches to the electric cyclic action bucket excavators’ technological operations identification are considered. The authors proposed the term “technological condition” of a single-bucket excavator as a combination of the current technological operation and data on the technical condition of the main components of the excavator. Based on the data analysed, a new method has been developed and proposed for identifying the technological state of excavators based on the mathematical apparatus and tools of neural networks and pattern recognition, which can be used in further studies. According to the methodology proposed by the authors using pattern recognition technologies based on the OpenCV library and a stereo pair formed from two Xiaomi cameras, experimental tests were carried out on a real object – an electric single-bucket excavator type “direct shovel” ECG-8I under various environmental conditions, including poor visibility. According to the test results, the proposed method showed high accuracy in technological operations identification – the identification error in the test sample did not exceed 5%, which indicates the adequacy of the constructed model. The scientific novelty of the work lies in the application of a method not previously applied in this field of technology, as well as in the proposed mathematical and simulation models. The practical novelty lies in the possibility of introducing this approach for the construction of automated management, moni toring and control systems, for solving the problems of weighing rock in an excavator bucket, determining the granulometric composition of rock in a bucket, as well as other identification problems.

In tensile machines to measure the change in length of the specimen during the stress, most time extensometer and various contact methods and tools, such as strain gauges are used, but these methods have limitations and low accuracy that cause unexpected and different results from the theoretical values for the strain. In this paper, to address this shortcoming and improve strain measurement, a non-contact method using image processing techniques in the MATLAB environment is introduced. for this purpose after capturing a move of tensile test and converting that in to some high quality pictures and using image processing we put some pattern in the first image and with using a special algorithm with name of Digital Image Correlation (DIC) detect that patterns until the last image then we can use of displacement algorithm and calculate change in the specimen’s length and after that we can also have “Strain” with using these values.