The hourly cooperation of production units and storage units in integrated energy system (IES) improves its performance on energy generation and supply, but increases the complexity in the capacity design. Conventional methods are flawed in terms of inappropriate operation mode and inaccurate performance evaluation. To improve the reliability of IES design, this research adopts operation trajectory prediction, energy dispatch schedule, and real time correction into operation mode, as well as annual continuous active dispatch optimization into capacity design method. IES components are divided into production unit, storage unit, and backup units. The proposed operation mode operates storage units actively to shift peak demand and resist uncertainty. The proposed design method optimizes both the capacity and cooperation of all units through an annual continuous operation, which improves the stability and reliability of the system performance. Compared with conventional methods, IES designed by the proposed method reduces energy consumption by 2.22%, annual cost by 1.40%, and carbon emission by 3.87%. In addition, this IES presents stability and robustness under different renewable energy and customer demand uncertainties. The proposed operation mode and design method contribute to accurate and speedy optimization, and provide reliable suggestions and feedbacks for IES investment and management. •Storage units are operated actively to shift peak demand and resist uncertainty.•Both capacity and cooperation of production units and storage units are optimized.•Stability and reliability of performance are improved in operation and design.•Accurate reference and feedback are provided for IES investment and management.