Container storage planning and agent behavioral rule designing for an automated port transportation system

Transcription

1 Container storage planning and agent behavioral rule designing for an automated port transportation system Satoshi HOSHINO, Jun OTA, Akiko SHINOZAKI, Hideki HASHIMOTO The University of Tokyo Mitsubishi Heavy Industries, TD. Abstract In this paper, we construct a horizontal AGV transportation system as an automated port container terminal; then, establish the system that would be more cost effective. Although it is difficult to control the horizontal AGV transportation system for the automation, it has been made clear that the horizontal system is more cost effective than the conventional vertical system. From this point of view, we consider the system management, that is, the container storage planning based on the stored container information and cooperative agent s behavior designing as the multi-agents system for a realization of the effective transportation system. Key Words: AGV Transportation System, System Management, Container Storage Planning, Cooperative Agent Behavior Designing. 1. 1) AGV Automated Guided Vehicle AGV (I) (II) (III) (I) 2) (II) 3) 4) AGV AGV 4) Fig.1 AGV AGV Container yard area [m] Quay area Container ship QCC 1 [m] Transportation area Container yard area AGV (oad) AGV (Empty) Transportation area 32 [m] RTGC 1st location 2nd location Stopping AGV 3rd location Working path Fig.1 Horizontal AGV Transportation System RTGC Rubber Tired Gantry Crane (III) (III) AGV RTGC AGV I nth location 3 [m]

2 ane (ocation) 3 [m] Work path 4 Rows x Bays x 4 Tires = 3 [TEU] 1 container Bay (Handling point) = [feet] Row Fig.2 Container storage location 2. AGV 2 1 Fig.1 AGV RTGC step 1. Quay area QCC Quay Container Crane Container ship AGV step 2. Container storage location step 3. AGV Transportation area step 4. Working pathagv step. AGV RTGC step 6. RTGC RTGC step 7. AGV step Fig [TEU] 4 Rows + Bays + 4 Tires TEU Twenty-foot Equivalent Units 2 3 AGV RTGC AGV RTGC Tire Table 1 Specifications of the AGV, RTGC, and QCC AGV oad / Empty Max. traveling velocity [m/s].6 / 6.94 Acceleration [m/s 2 ]. /. Deceleration [m/s 2 ].63 /.63 RTGC Max. traveling velocity [m/s] 2. Acceleration [m/s 2 ].1 Deceleration [m/s 2 ].4 Storing time [s] Transferring time [s] QCC oading time [s] 6 AGV RTGC = Table1 AGV RTGC QCC ) AGV oad Empty RTGC QCC 2 1 [Berth] QCC Fig.1 1 [Berth] [m] 3 QCC RTGC 2 AGV 1 3 QCC Row Bay ) AGV 7)

3 Container transfer ( -->) Sensing and (a) The AGV A measures the relative distances of the RTGC A and RMGC B AGV_C Call (a) The AGV B calls The RTGC A Container transfer ( -->) AGV_C Call Call (b) The RTGC A can cross the RTGC B with moving (b) The AGV A calls the RMGC B based on the relative distances Fig.3 Distance-based selection (An AGV calls an RTGC based on each relative distance when the AGV arrives at the destination point) Fig.4 Two RTGCs of different sizes cooperate in one location without work space Work space of Work space of (i)agv RTGC (ii)agv RTGC AGV RTGC AGV RTGC 3 2 RTGC 1 Fig.3 AGV A RTGC Fig.3(a) 2 RTGC AGV A RTGC A RTGC B AGV A RTGC B Fig.3(b)RTGC B RTGC B (a) AGV (b) AGV A RTGC A 2 RTGC RTGC Work space Fig.4(a) AGV B RTGC RTGC B AGV A Call (a) The AGV A calls the RTGC B based on the work space Work space of Work space of (b) The RTGC B comes to the point Fig. Work-space-based selection (An AGV calls an RTGC based on the work space of the RTGC) RTGC A AGV B RTGC A RTGC B Fig.4(b) 2 RTGC (a) (b) AGV 3 3 RTGC 2 Fig. 1 2 RTGC AGV RTGC RTGC

4 RTGC RTGC Fig.(a) AGV A RTGC RTGC A RTGC B AGV A RTGC B RTGC B RTGC A Fig.(b) 3 4 AGV RTGC Table1 RTGC AGV RTGC RTGC Fig.6 RTGC 1 2 RTGC RTGC 1 AGV A RTGC RTGC RTGC B Fig.6(a) RTGC B AGV Fig.6(b) RTGC 2 AGV A RTGC RTGC B Fig.6(c) RTGC B Fig.6(d) RTGC RTGC B RTGC B (a) (b) AGV RTGC A RTGC RTGC AGV RTGC AGV RTGC AGV AGV AGV Fig.7 Initial task A A Row Bay AGV RTGC 4 2 RTGC 4 1 RTGC Table1 AGV RTGC Sensing and based on Distance A and B (a) The AGV measures the relative distances of the RTGC A and RMGC B; then, selects and calls the RTGC B Call (b) The RTGC arrives at the transferring point before the AGV arrives Work space of Call Work space of (c) The AGV selects and calls the RTGC B based on the work space Work space of Work space of Call (d) The RTGC arrives at the transferring point before the AGV arrives Fig.6 An AGV selects and calls an RTGC based on each relative distance or work space when the AGV comes into the working path AGV (1)RTGC (2) RTGC Fig.7 RTGC A RTGC B (1)

5 Table 2 Implementation conditions of the constructed systems Selection type RS DS DS DS DS WS WS WS WS Select & Call timing DP DP WP DP WP DP WP DP WP Container storage planning Condition numbder 1) 2) 3) 4) ) 6) DS: Distance-based selection, WS: Work-space-based selection, RS: Random selection DP: on the Destination Point, WP: on the Work Path, : Random storage, : Storage planning Order Task A C B [Initial task] 1 2 J B Work space of RTG_A 3 D 4 6 F E C 7 D 8 K 9 I F E A 11 H 12 G G H [Replanned task] Order Task A H B Work space of RTG_B J [Planned task] Order 1 2 Task A B 4 C I I J D K 3 C 9 E K D E F G H F 11 G 12 Fig.7 Container storage planning Planned task RTGC A (2) RTGC Planned Task Replanned task Table2 RTGC DS WS DP WP 8 1) AGV RTGC RTGC RTGC DS RTGC WS RTGC Table2 4) 6) 2 RTGC Fig.2 RTGC 16 [m] [Bay] 2) AGV= RTGC= 6 [TEU] 9 I J 11 K 12 Throughput [TEU/hour] 1 [TEU/hour] 2 Fig.8 AGV RTGC Fig.8(c) Construction cost = a AGV s + b RT GCs a b AGV RTGC a : b =1:2 1) 1 Fig.8(a) Fig.8(c) Fig.8(e) RTGC AGV Fig.8(b) Fig.8(d) 1 1 AGV 1 3 Fig.9 Fig.9 1) [point] max. 46 min. 38 max. 17 [%] 2) 4) 4) RTGC

6 (a) Condition # 1) (b) Condition # 2) (c) Condition # 3) (d) Condition # 4) (e) Condition # ) (f) Condition # 6) Fig.8 and RTGCs which constructs a system for a required throughput Construction cost [point] Condition # 1) Condition # 2) Condition # 3) Condition # 4) Condition # ) Condition # 6) Fig.9 Comparison results of the total construction costs 2) 3) 1 1 AGV Fig.8(b) Fig.8(c) RTGC AGV RTGC AGV RTGC ) 6) AGV RTGC AGV RTGC 6) 6. AGV 1) P. Ioannou et al. : Advanced Material Handling: Automated Guided Vehicles in Agile Ports, CCDoTT Technical Report, Center for Advanced Transportation Technologies, (1). 2) S. Hoshino et al. : Optimal Design Methodology for an AGV Transportation System by Using the Queuing Netowork Theory, 7th International Symposium on Distributed Autonomous Robotic Systems, pp. 391, 4. 3) S. Hoshino et al. : Comparison of an AGV Transportation System by Using the Queuing Network Theory, in Proc. IEEE/RSJ Int. Conf. Intell. Robots and Systems, pp , (4). 4), : AGV,, 1G, (4) ) MITSUBISHI HEAVY INDUSTRIES, TD. : Advanced Technology Cargo Handling Systems, Products Guide, (4). 6), :,, Vol. 16, No. 4, pp , (1998) 7), : AGV, C, 67, 68, pp , (1)