To address the problems of finished product inventory accumulation, prolonged makespan, and excessive energy consumption caused by demand uncertainty in tire manufacturing, this paper proposes a prediction driven dual line cooperative batch scheduling optimization method. The production process is divided into a pre-production stage on the rubber compound line and a customized manufacturing stage on the tire line, where a rolling forecasting model drives the rubber compound line to prepare materials in advance according to predicted demand. A multi-objective batch scheduling model is constructed with minimizing makespan, total inventory cost, and energy consumption cost as optimization objectives, and dynamic coordination between the two lines is achieved through joint decoding, inventory recursion, and tri-state feedback correction. On this basis, a dual population non-dominated sorting genetic algorithm is proposed for solving. Simulation experiments demonstrate that the proposed algorithm outperforms several mainstream multi-objective algorithms in terms of convergence and solution set distribution; compared with the no prediction mode, makespan, total inventory cost, and energy consumption cost are reduced by $29.3\%$, $33.3\%$, and $14.1\%$, respectively, validating the effectiveness and practicability of the prediction driven dual stage cooperative scheduling approach.