Research On Job Scheduling Optimization Of Automatic Storage And Retrieval System At Airfreight Station

Storage is an important part of air cargo, and storage operation optimization is an important way to improve the overall efficiency of air cargo transportation. As the main carrier of storage link at air freight station, automated storage and retrieval system(AS/RS) is gradually becoming a bottleneck of air cargo transportation with the increasing demand of air cargo throughput. Different from the ordinary AS/RS whose storage and retrieval operations carried out concentratively in the gateway of the aisle, the multiple-points operation mode of AS/RS at air freight station has the characteristics of concurrency and distribution, and its scheduling process has the characteristics of high density and high flexibility. The operation strategy of AS/RS at air freight station designed with the idea of separate optimization is not compatible with the growing demand of throughput. Discussing the operation optimization of AS/RS at air freight station plays an important role in perfecting the AS/RS scheduling strategy and improving the potential of throughput. The research content of this dissertation comes from the project named “AS/RS equipment monitor system(EMS) for air cargo terminal” presided by the author. Aiming at the warehouse operation process in EMS, this dissertation starts from three key links namely storage location assignment, out point assignment and elevation transfer vehicle(ETV) scheduling, and makes a deep research on the optimization problems of AS/RS under different operation modes at air cargo station.The system assigned the destination address of storage automatically is one of the main storage operation mode of EMS. The nearest neighbor assignment policy and the first come first service(FCFS) scheduling strategy are operation strategies at storage location assignment phase and scheduling phase respectively. In this operation process, the correlation between storage location assignment phase and scheduling phase is cut apart by separate optimization strategy, which in turn affects and limits the further improvement of operation efficiency of AS / RS at air freight station. To overcome the defects of the operation strategy in the process mentioned above, an integrated optimization model is presented by adopting integrated optimization method and analyzing the effect of storage location assignment and job scheduling to working efficiency comprehensively. A two-phase tabu search algorithm is presented to solve this problem according to its characteristics, and the effectiveness of the proposed algorithm is verified by numerical examples with different scale. The comparison of simulation results of integrated optimization method and isolated optimization method shows that the integrated optimization method is superior to isolated optimization method in solution performance and computation time, and the maximal improvement rate is 11.09% and 23.98% respectively.In the mode of two ETV parallel operation, the destination addresses of the two modes, namely storage according to designated destination address and retrieval according to the number of unit load device(ULD), are all assigned by the users. The generated instructions are completed by two ETV according to FCFS scheduling strategy. Since instruction grouping and sorting under this scheduling strategy haven’t consider the influence of double plate and anti-conflict avoidance to operating efficiency, the improvement of operating efficiency of AS/RS and the potential digging of throughput are restricted. In connection with the two kinds of operating modes mentioned above, under double ETV parallel operation mode, the problem of instruction grouping and sorting is optimized. The effect of double ULD transport and anti-collision avoidance to working efficiency is analyzed. A double ETV scheduling optimization model, whose objective is the minimal completion time of command sequences, is established, and an improved ant colony algorithm is designed to solve the model. The comparison of simulation results of different instruction sequence scale shows that the completion times of instruction sequences obtained by the optimization method presented in this paper are all superior to the completion times obtained by FCFS scheduling strategy, and the improvement rate is 37.35% to 53.61%.In EMS, retrieval of the entire inventory of ULD from the same flight is achieved according to the flight number. In this activity, the nearest neighbor strategy is adopted to determine the retrieval point, while FCFS scheduling strategy is adopted in the mode of two ETV parallel operation. Aiming at the disadvantages of operation strategy designed with the idea of separate optimization in the retrieval process carried out according to the flight number, basing on the optimization problem of double ETV scheduling, the effect of out point selection to double ULD transport, to anti-collision avoidance and to working efficiency is introduced. From the perspective of joint optimization, an integrated optimization model is abstracted to optimize the port assignment and double ETV scheduling, and the solving method of the model are given. The comparison of simulation results of different instruction sequence scale shows that, the integrated optimization method is superior to isolated optimization method, and the average improvement rate and the maximal improvement rate is 7.35% and 9.84% respectively.In the aspect of operation mode, based on the imbalance of the job tasks in different periods of air freight station, an energy saving operation mode that interchange efficiency with energy consumption is put forward. Under the double ETV parallel operation mode, the effect of double ULD transport and anti-collision avoidance to working efficiency and energy consumption is analyzed. A multi-objective optimization model of double ETV scheduling, whose objectives are the minimal energy consumption and the minimal completion time of command sequences, is established under the energy saving mode. An improved non-dominated sorting genetic algorithm is designed to solve the model, and the correctness of the model and the validity of the algorithm are verified by experimental simulations. In the Pareto optimal solution set obtained finally by simulation shows that, comparing with scheduling scheme with the minimal energy consumption, the working efficiency of scheduling scheme with the minimal command sequences completion time is improved by 27.17%, and the energy saving of scheduling scheme with the minimal energy consumption is improved by 17.89% comparing with scheduling scheme with the minimal command sequences completion time.