OPTIMIZATION METHODS AND INFORMATION TECHNOLOGIES IN ACTIVE EXPERIMENT.

Despite the fact that the effectiveness of experiment planning methods, especially when solving applied problems, has been proven more than once, the ideas of a multifactorial experiment are very slowly being introduced into science and engineering practice. The reasons for this are: the complexity of organizing the experiment; the parameters of the systems under study are of a complex dynamic nature and are subject to significant influences of changes in environmental conditions; the apparent complexity of the planning and calculation matrix discourages researchers with insufficient mathematical background; many manuals on the application of a multifactorial experiment are written at a level inaccessible to engineers, given the training in the disciplines of mathematical and computer cycles in technical universities. The experiment occupies the main place among the methods of obtaining information about the internal relationships of phenomena in nature and technology. As the processes and phenomena under study become more complex, the costs of the equipment and the experiment increase. During the tests, a large amount of experimental data is collected that requires processing and analysis. At the same time, the duration of the analysis, comprehension of the test results and their accounting for adjusting the characteristics of new products is very significant. The systems approach involves considering all elements of an active experiment as a single system. From these positions, it is necessary to present the general properties of the experiment as an object of study and give recommendations on the choice of mathematical techniques and methods that the experimenter can use when choosing decisions during the preparation of the experiment, its conduct and processing of the results. It is very important to choose methods and tools for processing experimental data. Despite the fact that, since 2002, many works have been proving the effectiveness of using the Excel analysis package for processing experimental data, the scientific and educational literature still uses coding of variables, randomization of experiments, and the choice of critical values of the criteria for assessing the quality of mathematical models from tables. In addition, optimization methods are not used in the works for forecasting. In this paper, in order to accelerate the process of introducing effective computer processing methods into science and engineering practice, the implementation of the methodology for processing active experiment data using Excel is presented. [ABSTRACT FROM AUTHOR]