A Model Predictive Control Approach Towards the Energy Efficiency of Submerged Dredging

Abstract

Due to the expected increase in coastal replenishment along the Dutch coastline, and the need to decrease pollutant emissions in the Netherlands and worldwide, the Autonomous Low Energy Replenishment Dredger (ALERD) is developed. The ALERD is a sustainable and cost-effective solution to perform coastal replenishment work along the Dutch coastline and will be used for submerged dredging operating fully electric. The ALERD is modelled as an underwater vehicle in six degrees of freedom, from which the forward speed, the depth, and the pitch angle are controlled using motion controllers. The ALERD is the successor of the Autonomous Underwater Maintenance Dredger (AUMD), for which it is shown that the power for propulsion can be reduced with 55%, and the power required for the dredging pumps can be reduced with 80%, due to the submerged operations. A simulation model is used to determine the required power for stability and buoyancy control of the ALERD, since this was up to this point still unknown. To be a cost-effective solution, the energy consumption for motion control should be minimized. Traditional motion control methods such as proportional-integral-derivative (PID) control are widely used for controlling underwater vehicles. The more advanced motion control method Model Predictive Control (MPC) is seen as a more promising motion control method, in terms of both trajectory-tracking and minimizing the control effort. A Model Predictive Control strategy is developed and used to minimize the energy consumption of the ALERD for stability and buoyancy control. It is shown that MPC indeed outperforms PID control on tracking the desired reference signal, while minimizing the control effort and thus the corresponding required energy. A sensitivity analysis is done using Monte Carlo Simulations, to show the effect of modelling uncertainties on the outcomes of the simulation model. There is a minimum variation in the outcomes, considering the modelling uncertainties. MPC is better capable of handling with modelling uncertainties compared to PID control, in terms of trajectory-tracking and minimizing the corresponding energy requirements. The MPC controlled ALERD is compared with the PID controlled ALERD, and a conventional dredger, and it is shown that for a similar operational profile, the MPC controlled ALERD is the most energy efficient solution. The results from the simulation model still show significant advantages of submerged dredging compared to conventional dredging, even when considering the energy demand for stability and buoyancy control, based on the used operational profile.