Dataset of Ship Block Scheduling

1. Research Assumptions and Data Content Overview
This dataset supports research on ship block construction scheduling optimization, aiming to solve multi-objective scheduling problems in complex and dynamic production environments using deep reinforcement learning (DDQN, EDDQN) and heuristic algorithms. The data covers the entire process from initial berth information, block details to scheduling results of various algorithms, and can be used to reproduce the simulation experiments and performance comparative analysis presented in the paper.
2. Data Structure and File Description
2.1 Top-level Directory Structure
Raw_input_Data: Contains all raw input data, which is the fundamental information for the scheduling problem.
Figure6&7_raw_data: Stores the original result data used to generate Figures 6 and 7 in the paper, for comparing the performance curves of different algorithms.
运行结果 (Run_Results): Contains complete scheduling result files from all algorithms (reinforcement learning and heuristic).
Figure10_raw_data.xlsx: The raw data used to generate Figure 10 in the paper.
2.2 Raw Input Data (Raw_input_Data)
Three core Excel files define the basic parameters of the scheduling problem:
初始胎位信息.xlsx (Initial_Berth_Information.xlsx): Records the initial status, available time, resource capacity, and other information of each berth.
分段信息表.xlsx (Block_Information_Table.xlsx): Contains detailed information of all blocks to be scheduled, such as block ID, process requirements, processing time, resource demand, priority, etc.
胎位信息.xlsx (Berth_Information.xlsx): Defines the physical attributes, maximum load, processable block types, and other constraints for each berth.
2.3 Figure Source Data (Figure6&7_raw_data)
DDQN_results.xlsx: Scheduling results of the Deep Q-Network (DDQN) algorithm under different instances, including convergence speed, energy consumption, completion time, and other metrics.
EDDQN_results.xlsx: Corresponding results of the improved Deep Q-Network (EDDQN) algorithm, used for comparison with DDQN.
2.4 Run Results (运行结果 / Run_Results)
Contains scheduling result files for multiple algorithms, with file names clearly identifying the algorithm type and problem scale:
DDQN_scheduling_results872 个分段_s...: Scheduling results based on the DDQN algorithm.
EDDQN_scheduling_results872 个分段_s...: Scheduling results based on the EDDQN algorithm.
scheduling_results_heuristic_earliest_...: Scheduling results based on the "EST" heuristic rule.
scheduling_results_heuristic_longest_...: Scheduling results based on the "LPT" heuristic rule.
scheduling_results_heuristic_resource_...: Scheduling results based on the "RUB" heuristic rule.
scheduling_results_heuristic_shortest...: Scheduling results based on the "SPT" heuristic rule.