COMPUTER TECHNOLOGIES OF AIRFIELD FLEXIBLE PAVEMENT CLASSIFICATION RATING DETERMINING USING ARTIFICIAL INTELLIGENCE

Abstract

This paper is dedicated to the specifics of using artificial intelligence for determining the classification rating of airfield flexible pavements. The International Civil Aviation Organization (ICAO) has adopted a new method called Aircraft Classification Rating / Pavement Classification Rating (ACR/PCR), which will replace the Aircraft Classification Number / Pavement Classification Number (ACN/PCN) method. ACR is the aircraft classification rating, defined as twice the permissible wheel load, expressed in hundreds of kilograms. PCR is the classification rating of airfield pavement, defined as the ACR of the “critical” or reference aircraft at its maximum permissible takeoff weight. Due to the introduction of ICAO’s new ACR/PCR method, there is a need to adapt the Ukrainian methodology to meet these new requirements. Using the ICAO ACR computer program, the ACR classification rating for various values of standard load was determined. The paper examines the use of artificial intelligence (AI), specifically ChatGPT and Llama, for regression analysis in determining the classification rating of airfield flexible pavements. Empirical formulas for determining the airfield flexible pavement classification rating (PCR) for four categories of subgrade strength have been obtained. AI chatbot ChatGPT and on-premises AI Llama can be used for regression analysis. When determining unknown parameters of a regression model using the on-premises AI Meta Llama, the same level of accuracy is achieved as when using OpenAI ChatGPT. However, on-premises AI ensures data confidentiality and security, operates independently of the Internet, and, as a result, can function autonomously even in remote locations.