With a high proportion of new energy access, the deployment of energy storage can assist the power system to cut peaks and fill valleys and smooth out fluctuations. However, current energy storage systems are costly and require government support. To this end, a profitability strategy for energy storage to maximise operating profit in an electricity market consisting of the grid is proposed, storage operators and customers. A profitability strategy that takes into account incentives in combination with an intelligent algorithm that provides different weighted reward allocations to the storage system operator for each peak hour is proposed. On the one hand, the algorithm is based on deep learning of least square support vector machine to establish price and load forecasting models. On the other hand, deep reinforcement learning is used to determine the optimal charging and discharging strategy considering the peak state of power grid, user load demand and the profits of energy storage system operators. Finally, a case study is conducted to verify that the strategy can significantly improve the profitability of the energy storage system operator and reduce the pressure on the grid.