In this paper, a new aggregation operator, called the continuous triangular fuzzy generalized ordered weighted averaging (CTF-GOWA) operator, is proposed for aggregating triangular fuzzy numbers (TFNs). In order to adapt to the uncertainty and complexity of prediction, differential polymerization method is used and a special form of CTF-GOWA operator is generated. In particular, the special cases of control parameters and basic unit monotone functions introduced in information aggregation is discussed and its credibility is verified. In addition, the CTF-GOWA operator is applied to the aggregation of multiple TFNs, and various aggregation operators are obtained, including the weighted CTF-GOWA (WCTF-GOWA) operator, the ordered WCTF-GOWA (OWCTF-GOWA) operator, and the combined CTF-GOWA (CCTF-GOWA) operator. Later, the induced variables are added to the OWCTF-GOWA operator to obtain the induced OWCTF-GOWA (IOWCTF-GOWA) operator, which is combined with the cosine similarity degree to construct a combined prediction model about the TFNs. The proposed model has the advantages of less computation, high accuracy, and more adaptability to the actual situation. Finally, the validity of the proposed model is verified by illustrative examples. Sensitivity and comparative analyses are also implemented to assess the stability and validity of our method.