Automation of experimental data processing in agronomic research

Experimental research in agronomy is the highest form of empirical methods of cognition of the influence of agrotechnological processes on quantitative and qualitative indicators of cultivated crops. The process of cognition is multi-stage and includes various forms: observation, comparison, control and measurement. The main methods of agronomic research include laboratory, vegetative, lysimetric and field, which, in combination with observations of plants and environmental conditions, represent the most important tools of scientific agronomy. Among them, the main thing is experiment in the field. Field experiment completes the exploratory research, quantifies the agrotechnical and economic effect of a new method or technology of plant cultivation and provides objective grounds for the introduction of scientific achievements in agricultural production. The analysis of modern methods of applied research indicates the fact that the construction and application of algorithms is characteristics of a set of statistical methods and methods of data processing. Meanwhile the logics of their use in the preparation of input information and its subsequent processing, as well as writing reports with the results of analysis is still not automated. In this regard, the relevance of the problem posed is due to the high frequency of errors in the calculated statistical part of evidence-based research in conditions of complex routine analytics, which takes a lot of time when processing data by representatives of other branches of science, as well as the high cost of work when contacting specialists, the cost of specialized software. In their work, the authors made an attempt to draw attention to the existing specialized software working according to algorithms that are accepted as standard in agricultural scientific institutions, for example, the algorithms of B.A. Dospekhov, which are tied to the scheme of laying field experience and repetition are considered as a factor or algorithms of N.A. Plokhinsky developed to determine quantitative characteristics based on multifactorial analysis of variance.