In view of the production decision-making problems of manufacturing/remanufacturing hybrid systems under the carbon cap-and-trade mechanism, a production decision-making optimization model of multi-cycle manufacturing/ remanufacturing hybrid systems is proposed to maximize the total profit and minimize carbon emission considering the impact of customer demand differences and input costs and benefits of carbon emission reduction technology. According to the characteristics of the model and the thought of multi-population co-evolution, a multi-population hybrid cuckoo search(MPHCS) algorithm is designed to solve the problem. A custom preference random walk strategy is proposed to increase the influence of excellent individuals on global search, so as to guide population evolution and realize information sharing among subpopulations. The adaptive formula is used to update the step size to prevent falling into local optimum. Moreover, the chaos search based on Logistic mapping is introduced to carry out fine search on the high-quality solution region found by the global search of the MPHCS algorithm and improve local search ability. Finally, an example is given to illustrate the rationality and effectiveness of the proposed model and algorithm.