A dynamic programming algorithm for optimal control of tidal range schemes

Optimising the operation and design of tidal range schemes is a long-standing and critical problem in their design which also affects their financial feasability. The optimisation process is very time consuming considering its complexity and the number of variables which leads to significant simplifications in the process. Enhancing the efficiency and accuracy of optimisation methods is essential for maximising their energy output. In this study, we propose an efficient tabular dynamic programming for tidal range scheme optimisation by refining the algorithm. Dynamic programming is capable of finding the global optimal solution given that the assumptions used to formulate the optimisation problem are valid. The algorithm refined in this study was tested with two different schemes, namely the West Somerset Lagoon and the North Wales Lagoon in the UK, and demonstrated accurate and efficient prediction of the scheme’s operation. The computational time required by the dynamic programming algorithm was significantly lower than the most promising optimisation approaches used for tidal range schemes, namely Genetic Algorithm, showing potential for real-time application. The refined algorithm includes provisions to simulate tidal range schemes with block formulation to account for the fact that turbines and sluice gates are usually installed in different blocks and are subjected to tidal phase differences. Results demonstrated that optimisation including the multi-block formulation lead to electricity generation up to 6% more annually. • First dynamic programming approach for multi-block tidal range scheme operations. • Operation strategy guaranteed globally optimal. • Tested on West Somerset and North Wales lagoons, with real-time potential.