The changing market demands pose significant disturbances to the production planning of businesses in each cycle. To address this issue, a multi-period production decision optimization model for hybrid product manufacturing/remanufacturing systems was established, considering the uncertainty of market demand evolution, under the framework of a carbon policy that combines carbon trading and carbon taxation. The objective of the model is to maximize the total profit. To solve the model, a hybrid artificial bee colony algorithm and harmony search algorithm based on chaotic mapping were designed. An adaptive harmony parameter optimization algorithm was proposed to enhance the ability of the algorithm to escape local optimal solutions. The Tent chaotic initialization and harmony algorithm population were adopted to improve the uniformity of the initial individual distribution. The artificial bee colony algorithm was introduced to reduce the adverse effects of decreasing diversity in the late stage of the algorithm, thereby improving the optimization ability of the algorithm. The rationality and effectiveness of the model and algorithm were verified through instance simulations.