محمد جوادیان

This study incorporates artificial neural network (ANN) modeling to design an intelligent road signboard that uses internet of things (IoT) technology to assign the speed limit on interurban highways. The appropriate speed limit must be determined by traffic police experts based on weather conditions and times of the day. Here, an intelligent IoT-based signboard is proposed to announce speed limits on roadways considering some effective parameters, such as temperature, humidity, time, and light. The signboard receives environmental data through its sensors and uses artificial neural networks to compute the speed limit. A feed-forward neural network (FFNN) is provided as the most reliable model. A hybrid training method based on gray wolf optimization and Bayesian regularization is also developed to enhance model performance. The proposed hybrid model converges with an error of less than 3.0% to expert opinion. The model equations are extracted for use in a microcontroller that calculates a safe speed limit based on weather conditions. Additionally, the underlying IoT technology has enabled the police station to remotely monitor and control the developed system. Experimental results demonstrate the reliability of the designed signboard. In all experimental cases, the computed speed limits were in coincidence with the expert’s estimations.