Development of an Intelligent System for Diagnosing Rumen Acidosis in Cows. Part 1: Formalization of Tasks

Introduction. Improvement of the diagnostic methods for detecting bovine ruminal acidosis is a relevant problem due to the prevalence of this disease resulting in cattle murrain. The most future-oriented methods of early diagnostics of this pathology are prophylaxis campaigns or individual examinations of animals using the online services in the frame of veterinary telemedicine technologies. However, in Russia, the efficient use of these technologies is hindered by the absence of well-developed methods of assessing the pathology based on the mathematical tools integrated into the advanced intelligent systems. In our study, we attempted to create an intelligent system for diagnosing rumen acidosis in cows. The first stage of the research consisted of formalization of tasks based on the concept of fuzzy logic. Materials and Methods. The study was conducted at the Ural State Agrarian University and the Northern Trans-Ural State Agricultural University from 2022 to 2025. For diagnosing rumen acidosis, a minimum list of diagnostic parameters was determined as input variables for the fuzzy system (based on the analysis of data obtained from the specialized reference books and scientific publications). Due to the possible difficulty of combining several parameters exceeding reference values, a method for representing uncertainties — fuzzy set theory — was used to correctly solve the problem in the given settings. A hierarchical structure was chosen to build the intelligent system, which makes it strictly logical and transparent for analysis. Scilab open source software with embedded text editor Scinotes were used to develop the application. Results. A set of necessary diagnostic parameters including pH of rumen contents, fat content in milk, pulse rate, and respiratory rate was justified. A substantive formulation and formalization of the problem of fuzzy inference for assessing rumen acidosis pathology in cows was completed. The structure of the hierarchical-type intelligent system was developed. All variables, including intermediate ones, were formalized within the fuzzy model of the intelligent diagnostic system. Discussion and Conclusion. In the first part of the study, we have established the formal models of all variables (input, output, and intermediate) for solving the task of diagnosing rumen acidosis in cows. In case of emergence of new parameters directly related to acidosis, they can also be integrated into the intelligent system. Based on the formalized data obtained at this stage of the research, in the second part of the study, it is planned to apply computer technologies to create the intelligent system for assessing rumen acidosis pathology in cows.