This paper explores intelligent traffic light management advancements, focusing on controlling intersection traffic opening times. The decision-making process is influenced by factors such as traffic density. The information for these decisions is gathered from sensors placed on the streets, whose accuracy can vary. Data collected are processed to aid control agents in decision-making. The paper proposes an intersection control algorithm that operates under the assumption of lacking sensorisation. To balance raw sensor data, control nodes implement a reinforced learning algorithm to select the most suitable combination of sensors to improve traffic parameters. The paper also introduces a method for calculating traffic density by combining sensors with imprecise data. This research contributes to intelligent traffic management by providing a novel approach to intersection control and traffic density calculation.