How to optimally design industrial high-temperature heat pump for multiple objectives and working conditions?

High-temperature heat pump has aroused widespread attention for its superiority in industrial low-carbon heating. However, the inevitable fluctuation of the working conditions can significantly deteriorate its actual performance. Therefore, how to comprehensively consider multiple conditions in the design process holds great practical significance. Herein, a bi-level programming model involving multiple objectives and working conditions is proposed, within which, the lower level is based on a hybrid model resolved by the Geyser-inspired algorithm to obtain the performances of specified heat pump configuration under different conditions, and the upper level generates and ranks different configurations by the improved non-dominated sorting genetic algorithm. In the exemplary case with the payback period, carbon emission reduction being the objective function and three different conditions, an average 17.5 % increment of carbon emission reduction can be achieved by the proposed method than the conventional one under the same payback period, validating its necessity and superiority. The mean ratios of the evaporator to condenser area involved in the Pareto fronts are 0.81, 0.76 and 0.73, indicating that priority should be given to increasing the condenser area for better adaptability of variable conditions. Finally, the optimal configuration with compressor displacement, evaporator and condenser area to be 1959 m³/h, 154.5 m² and 232.0 m² is established.