A fuzzy logic approach for diagnosing and planning the leveling network of the Brazilian Geodetic System

Abstract

The leveling network consists of benchmarks that represent the vertical component of a geodetic system. This network is essential for the planning and execution of engineering projects. Its maintenance and densification require detailed planning to ensure the efficient allocation of resources. In this sense, to optimize the diagnosis and planning of geodetic campaigns, assisting decision-making in the face of budgetary and operational limitations, a methodology based on fuzzy logic was proposed, which deals with uncertainties and subjectivities in the decision-making process. The fuzzy process, implemented using the Mamdani method, encompassed fuzzification, inference, and defuzzification, leading to a hierarchical prioritization of field activities into high, medium, and low categories. Notably, the results identified 9,155 destroyed benchmarks and 26,322 benchmarks classified as not found, amounting to 49.6% of the leveling network. This situation could be more critical, as 19,388 of the 22,524 benchmarks marked as 'good' in the geodetic database and classified as high-priority for physical verification have not been visited in over 20 years. The approach provided a systemic vision that will support the planning and optimization of resources, strengthening strategic decision-making.