Establishment of data-driven multi-objective model to optimize drilling performance

Drilling parameters optimization has consistently generated research interest over the years because of the costsaving benefits associated to improve drilling efficiency. However, several physics-based and data-driven models have been developed for drilling parameters optimization, and the majority of the data-driven models are based on regression methods. Obtaining highly accurate and optimized drilling parameters with rapid as well as costeffective simulation runs is difficult to achieve. To accurately and rapidly predict drilling parameters, a multiobjective optimization model was proposed in this study. In the proposed model, the rate of penetration (ROP), unit drilling cost (UDC), and mechanical specific energy (MSE) were considered as the objective functions, while the weight on bit (WOB) and rotations per minute (RPM) were chosen as the optimization variables. Meanwhile, a method for ROP prediction based on improved back propagation neural network (BP) is also presented. Field data presented in this study indicate that when drilling is free of drilling complications, this multi-objective optimization model could optimize WOB and RPM with higher ROP and lower MSE, and UDC.