Abstarct: Well drilling is the primary method for accessing hydrocarbon reservoirs، and the rate of penetration (ROP), as a key performance indicator، has a direct impact on operational time and cost. Although previous studies have extensively focused on operational parameters، the role of formation geomechanical properties in ROP prediction models has received comparatively less attention. This is despite the fact that numerous studies have demonstrated the significant influence of these properties on drilling performance optimization and formation stability. The objective of this research is to develop a predictive model for ROP with particular emphasis on geomechanical characteristics. To this end، data from two vertical wells in an oil field located in southwest Iran were analyzed. A one-dimensional geomechanical model of the formation was constructed and calibrated baxsed on petrophysical logs and in-situ tests، from which the relevant geomechanical properties were extracted. Following data preprocessing and feature selection، depth، unconfined compressive strength (UCS), and Poisson’s ratio، along with weight on bit (WOB), rotary speed (RPM), and drilling fluid flow rate، were identified as the final input parameters. Support Vector Machine (SVM), Random Forest (RF), and Multilxayer Perceptron (MLP) models were developed, and their hyperparameters were optimized using Particle Swarm Optimization (PSO) and Genetic Algorithm (GA) techniques. The results indicated that the Random Forest model optimized by the Particle Swarm Optimization achieved the highest predictive accuracy for ROP estimation. Ultimately، an integrated model was developed that combines drilling operational data with geomechanical properties، enabling accurate prediction of ROP in new wells within the field.