With the development of manufacturing workshops toward double-floor and high-density configurations, traditional single-floor parallel row ordering methods are inadequate for addressing the complex constraints introduced by elevators and longitudinal transportation aisles in inter-floor material handling. To address this issue, a double-floor k-parallel row ordering problem with transportation aisles is proposed. First, a collaborative layout structure integrating facilities, elevators, and longitudinal transportation aisles is established. A material handling distance metric based on physically feasible transportation paths is developed, and a mixed-integer linear programming (MILP) model is formulated to minimize the weighted material handling distance between facilities by jointly optimizing row assignment, intra-row ordering, elevator entrance and exit selection, and aisle interference avoidance. Then, to improve the solution efficiency for medium- and large-scale instances, an incremental evaluation-guided local search memetic algorithm is proposed. The algorithm incorporates material flow information into a randomized variable neighborhood descent procedure to guide neighborhood exploration and employs a row-level incremental fitness evaluation strategy to eliminate redundant computations, thereby improving both search efficiency and solution quality. Computational results demonstrate that the proposed algorithm obtains the same optimal solutions as the exact method for small-scale instances and outperforms the conventional genetic algorithm for medium- and large-scale instances. A case study of a crane manufacturing enterprise shows that the optimized layout reduces the material handling cost by 22.42%, demonstrating the effectiveness and practical applicability of the proposed method.