An intelligent warehousing system for a latex matrix ground station

Abstract

The present application relates to a kind of intelligent warehousing systems for latex matrix ground station applied to the technical field of intelligent warehousing, the system includes warehouse, multiple densely arranged shelves, multiple AGVs and central dispatch system. Formed between the shelves and the inner wall of the warehouse one-way traffic loop network by multiple channel segments, channel width and the minimum safe traffic width required for a single AGV match, central dispatch system plans path for AGV executing task, and generates reservation occupancy time window for channel segment passed by it, and then maintains global space-time table to detect space-time conflict between AGVs in real time;When detecting conflict, the system dynamically calculates the traffic priority of conflict AGV, and instructs low-priority AGV to stagger its time window through speed regulation, thereby fundamentally avoiding deadlock.The present application realizes efficient, continuous and reliable operation of the system while ensuring space utilization.

Description

Technical Field

[0001] This invention relates to an intelligent warehousing system, and more particularly to an intelligent warehousing system for a latex-based ground station applied in the field of intelligent warehousing technology. Background Technology

[0002] Latex matrix ground stations are specialized production facilities in the civil explosives industry that provide latex matrix—a core semi-finished product—for on-site mixing of emulsion explosives. Their warehousing system is crucial for the safe storage and efficient management of raw materials. Traditional latex matrix ground station warehousing relies heavily on manual forklift handling and stacking. This method is not only inefficient and prone to errors, but also poses significant safety hazards due to the extreme sensitivity of raw materials such as ammonium nitrate to impacts and friction, introducing inherent uncertainties into manual operations.

[0003] AGVs (Automated Guided Vehicles) are crucial transportation tools in Industry 4.0 smart factories. With the development of AGV technology, their application in warehousing and logistics is becoming increasingly widespread. To improve the automation and intelligence level of warehousing systems, AGVs are gradually being introduced to replace manual labor. For example, Chinese invention patent CN112184116B discloses a warehousing system based on AGV robots and its management method.

[0004] However, in space-constrained applications like latex-based ground stations where high-density storage is required, the widespread application of AGV systems faces significant challenges. Multiple AGVs operating in a confined space are highly susceptible to traffic jams (i.e., multiple AGVs blocking each other and unable to move), leading to system failure.

[0005] Existing technologies primarily employ two approaches to address this issue: First, a conservative channel exclusion strategy (such as using traffic lights) allows one AGV to monopolize the entire channel, while other AGVs must stop and wait outside the entrance. While this method avoids deadlock, it comes at the cost of significant efficiency loss, failing to fully utilize the AGVs' potential. Second, simply increasing the channel width to provide clearance, but this leads to a sharp drop in space utilization, unsuitable for applications with stringent space requirements, such as latex-based ground stations.

[0006] Therefore, existing technologies cannot simultaneously meet the triple requirements of high space utilization, high operational efficiency, and high operational reliability. How to fundamentally prevent AGV deadlock without sacrificing space utilization has become a core technological bottleneck restricting the intelligent upgrading of latex-based ground stations and even the entire high-density warehousing industry. Summary of the Invention

[0007] In view of the above-mentioned prior art, the technical problem to be solved by the present invention is: how to design an intelligent warehousing system for latex matrix ground stations that can ensure high space utilization, high operating efficiency and high operating reliability, and fundamentally prevent AGV deadlock.

[0008] To address the aforementioned issues, this invention provides an intelligent warehousing system for latex-based ground stations, comprising a warehouse, multiple shelves within the warehouse, multiple AGVs, an inbound connection area, an outbound conveyor line, a warehouse management system, and a central dispatch system. Passageways for AGVs are formed between the shelves and between the shelves and the warehouse walls. The warehouse management system assigns material inbound and outbound tasks to the AGVs. The passageways are a one-way loop network composed of multiple connected passageway segments, and the width of each passageway segment matches the minimum safe passageway width required for a single AGV to perform material inbound and outbound tasks.

[0009] The central dispatch system communicates with the AGVs and is configured as follows:

[0010] Plan a travel path for each AGV performing a task, and allocate resources for the channel segments included in the path in the time dimension, generating a reservation time window for each channel segment.

[0011] Maintain a global time-space table that contains the reservation and occupancy time windows of all AGVs;

[0012] By continuously comparing the reserved time windows in the global spatiotemporal table, we can detect whether there are overlapping conflicts in spatiotemporal resources between different AGVs.

[0013] When a conflict is detected, the passage priority of the conflicting AGVs is dynamically calculated according to preset rules.

[0014] The low-priority AGVs can adjust their speed to stagger their scheduled time windows, thus avoiding conflicts between AGVs.

[0015] In the aforementioned intelligent warehousing system for latex-based ground stations, AGV deadlock can be fundamentally prevented, ensuring both high space utilization and high operational efficiency and reliability.

[0016] As a further improvement of the present invention, the reservation time window is a data structure that includes at least the start time, end time, channel segment identifier, and AGV identifier fields.

[0017] As a further improvement of the present invention, the global time-space table is a data structure in the form of a Gantt chart with time as the horizontal axis and channel segment as the vertical axis, used to store and compare all reserved time windows.

[0018] As another improvement of the present invention, multiple functional avoidance compartments are also distributed in the warehouse.

[0019] The central dispatch system is also configured as follows:

[0020] When a conflict is detected, a first deadlock prevention strategy and a second deadlock prevention strategy are generated.

[0021] The first anti-deadlock strategy is: low-priority AGVs are instructed to stagger their scheduled time windows by adjusting their speed;

[0022] The second anti-deadlock strategy is to plan a path so that the low-priority AGV can drive into the nearest and idle functional avoidance compartment to avoid the conflict point.

[0023] Calculate and compare the estimated time cost T1 of the first deadlock prevention strategy with the estimated time cost T2 of the second deadlock prevention strategy;

[0024] Select a deadlock prevention strategy with lower estimated time cost and instruct the low-priority AGV to execute it.

[0025] As a further improvement to the present invention, the functional obstacle avoidance compartment is embedded in the inner wall of the warehouse, and its size is sufficient to accommodate at least one AGV for docking.

[0026] As an additional improvement to the present invention, the functional obstacle avoidance cabin is also large enough to accommodate at least one AGV to complete a turning operation.

[0027] The central dispatch system is also configured as follows:

[0028] Develop a return-to-home strategy for each AGV performing a task;

[0029] Return-to-home strategies include:

[0030] First return strategy: Control the AGV to continue moving along the one-way traffic loop network and return to the return target point in a loop;

[0031] The second return strategy is to control the AGV to enter the nearest and available functional obstacle avoidance compartment, complete the turnaround, and then return to the return target point.

[0032] As a further improvement to the present invention, the central dispatch system is also configured as follows:

[0033] Calculate and compare the estimated time for executing the first return-to-home strategy with the estimated time for executing the second return-to-home strategy;

[0034] Choose the return strategy with the shorter time consumption as the final return strategy, and formulate the return path for the AGV based on the final return strategy.

[0035] As another improvement of the present invention, the entrance of the functional avoidance compartment is located at the end of the passage section and is connected to the passage section.

[0036] The central dispatch system is also configured as follows:

[0037] Real-time monitoring of the operating status of each AGV;

[0038] When an AGV is determined to be faulty and unable to move, it is marked as a faulty AGV and a fault cleanup procedure is executed.

[0039] As a further improvement to the present invention, the fault clearing procedure includes:

[0040] One AGV is selected from the other AGVs as the rescue AGV according to the preset rules;

[0041] Plan a rescue route to the malfunctioning AGV for the rescue AGV;

[0042] The rescue AGV is instructed to move to the location of the malfunctioning AGV and push it into the nearest available functional avoidance compartment.

[0043] Compared with the prior art, the present invention has the following beneficial effects:

[0044] 1. This invention uses a time window reservation and a global time-space table mechanism to pre-allocate and detect conflicts in the time-space resources that the AGV plans to occupy in the future before it starts executing tasks. This can proactively anticipate and resolve potential traffic deadlocks in advance, thereby fundamentally preventing AGV deadlocks. Furthermore, the dense arrangement of shelves ensures high space utilization, high operating efficiency, and high operational reliability.

[0045] 2. This invention, by introducing a functional obstacle avoidance module and a dynamic optimization algorithm, enables intelligent scheduling of the entire warehousing system. Whether it's conflict resolution during task execution or return route planning after task completion, the system can select the globally optimal solution from multiple feasible options based on real-time benefit assessment. This two-way intelligent optimization mechanism ensures that the system maintains maximum efficiency under any operating condition, minimizing AGV idle rate, waiting time, and ineffective travel, thereby greatly improving the overall throughput and economy of the system.

[0046] 3. By setting up a highly automated collaborative rescue mechanism, this invention achieves a leap from "preventing deadlock" to "deadlock resolution," transforming major system failures that traditionally require manual intervention into occasional events that the system can resolve on its own. This ensures that the unexpected failure of a single device will not lead to the long-term paralysis of the entire intelligent warehousing system, greatly improving the robustness, reliability, and availability of the system. Attached Figure Description

[0047] Figure 1 This is a structural block diagram of the intelligent warehousing system in the first embodiment of the present invention;

[0048] Figure 2 This is a logic flowchart of the central scheduling system avoiding AGV conflicts in the first embodiment of the present invention.

[0049] Figure 3 This is a structural block diagram of the intelligent warehousing system according to the second embodiment of the present invention;

[0050] Figure 4 This is a logic flowchart of the central scheduling system avoiding AGV conflicts in the second embodiment of the present invention.

[0051] Figure 5 This is a flowchart illustrating the logic of the central scheduling system determining the return path for the AGV in the second embodiment of the present invention.

[0052] Figure 6 This is a demonstration diagram of the central dispatch system executing a fault cleanup procedure in the third embodiment of the present invention. Detailed Implementation

[0053] The three embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0054] First implementation method:

[0055] Figures 1-2 This invention illustrates an intelligent warehousing system for a latex-based ground station, comprising a warehouse, multiple shelves within the warehouse, multiple AGVs, an inbound connection area, an outbound conveyor line, a warehouse management system, and a central dispatch system. Channels for AGV passage are formed between the shelves and between the shelves and the warehouse walls. The warehouse management system assigns material inbound and outbound tasks to the AGVs. The channels are a one-way traffic loop network composed of multiple channel segments. Multiple shelves are densely arranged, and the width of each channel segment only needs to match the minimum safe passage width required for a single AGV to perform material inbound and outbound tasks. Thanks to the cyclical nature of the one-way traffic loop network, after completing its work in front of the shelves, the AGV does not need to turn around in the narrow channels and can continue forward, naturally cycling along the network path to its return target point (for AGVs performing outbound tasks, the return target point is usually the outbound conveyor line; for AGVs performing inbound tasks, the return target point is usually the inbound connection area).

[0056] The central dispatch system communicates with the AGVs and is configured as follows:

[0057] Plan a travel path for each AGV performing a task, and allocate resources for the channel segments included in the path in the time dimension, generating a reservation time window for each channel segment.

[0058] Maintain a global time-space table that contains the reservation and occupancy time windows of all AGVs;

[0059] By continuously comparing the reserved time windows in the global spatiotemporal table, we can detect whether there are overlapping conflicts in spatiotemporal resources between different AGVs.

[0060] When a conflict is detected, the passage priority of the conflicting AGVs is dynamically calculated according to preset rules.

[0061] Low-priority AGVs can stagger their scheduled time windows by adjusting their speed (e.g., slightly accelerating or slightly decelerating) to avoid conflicts between AGVs.

[0062] The reservation occupancy time window is a data structure that includes at least the start time, end time, channel segment identifier, and AGV identifier fields. The global time-space table is a Gantt chart-style data structure with time as the horizontal axis and channel segment as the vertical axis, used to store and compare all reservation occupancy time windows.

[0063] The rules preset for dynamically calculating the priority of conflicting AGV passage are based on one or more of the following factors: the urgency of the task being performed by the AGV, the load status of the AGV, and the distance already traveled by the AGV.

[0064] The urgency of the tasks performed by the AGV is determined by the task priority label assigned by the warehouse management system. Urgent outbound tasks usually have the highest priority.

[0065] AGV load status: Heavy-load (carrying materials) AGVs have higher priority than unloaded AGVs because of their large running inertia, high start-stop energy consumption and high safety risks.

[0066] AGV distance traveled: For AGVs that have traveled a long distance, their priority can be appropriately increased to avoid excessive delays in their tasks.

[0067] The central dispatch system assigns weights to each factor based on real-time conditions and calculates the final priority value. The one with the highest value is defined as a high-priority AGV, and the rest are defined as low-priority AGVs.

[0068] To further illustrate the dynamic calculation process of passage priority, a specific embodiment is provided here.

[0069] The preset rule is a weighted scoring method. The central scheduling system presets a basic weight coefficient for each factor affecting priority, and can dynamically fine-tune it according to the real-time load of the system.

[0070] In this embodiment, the initial basic weight coefficients for each factor are set as follows:

[0071] Task urgency weighting coefficient W 紧急 =0.5;

[0072] AGV Load Condition Weighting Coefficient W 载重 =0.3;

[0073] AGV travel distance weighting coefficient W 距离 =0.2;

[0074] At the same time, scoring criteria are set for each factor:

[0075] Mission urgency level S 紧急 Emergency missions are worth 100 points, and ordinary missions are worth 50 points.

[0076] AGV Load Status S 载重 : 100 points for heavy load, 50 points for unloaded;

[0077] The AGV has traveled a distance of S 距离 The linear mapping method is used to map the actual driving distance to a score range of 0-100. The longer the distance, the higher the score.

[0078] The formula for calculating the priority value P is:

[0079] P=(W 紧急 ×S 紧急 )+(W 载重 ×S 载重 )+(W 距离 ×S 距离 );

[0080] For example: Suppose that the two AGVs that are in conflict within the system are AGV-1 and AGV-2.

[0081] AGV-1 performs emergency outbound tasks (S 紧急 =100), under heavy load (S 载重 =100), the distance traveled is rated as 80 points;

[0082] AGV-2 performs routine tasks (S 紧急 =50), in an unloaded state (S 载重 =50), the distance traveled is rated as 30 points.

[0083] Calculate their priority values ​​respectively:

[0084] P1=(0.5×100)+(0.3×100)+(0.2×80)=50+30+16=96;

[0085] P2=(0.5×50)+(0.3×50)+(0.2×30)=25+15+6=46;

[0086] After comparison, P1 > P2, therefore the central dispatch system determined that AGV-1 has a high priority and AGV-2 has a low priority. The system then instructed AGV-2 to perform avoidance by adjusting its speed or entering the avoidance compartment.

[0087] It is understood that the above weighting, scoring criteria, and calculation formulas are merely one feasible embodiment and are not intended to limit the present invention. Those skilled in the art can adjust the weights of each factor and the scoring mechanism according to actual application scenarios.

[0088] Unlike existing technologies that passively avoid conflicts (such as waiting or detouring) only after they occur, this invention uses a pre-allocation time window and a global time-space table mechanism to pre-allocate and detect conflicts in the time-space resources that the AGV plans to occupy for a period of time before it starts executing its tasks. This proactively anticipates and resolves potential traffic deadlocks in advance, thereby fundamentally preventing AGV deadlocks. Furthermore, the dense arrangement of the racks ensures high space utilization, high operating efficiency, and high operational reliability.

[0089] For industrial environments requiring continuous, stable, and safe operation, such as latex-based ground stations, the highly reliable logistics support provided by this invention not only improves production efficiency but also greatly enhances the safety and stability of the entire production system, thus having significant practical and economic value.

[0090] Second implementation method:

[0091] Figures 3-4 This invention illustrates an intelligent warehousing system for a latex-based ground station. Unlike the first embodiment, the warehouse also features multiple functional avoidance compartments distributed throughout. These compartments are embedded in the warehouse walls and are large enough to accommodate at least one AGV for temporary docking.

[0092] The central dispatch system is also configured as follows:

[0093] When a conflict is detected, a first deadlock prevention strategy and a second deadlock prevention strategy are generated.

[0094] The first anti-deadlock strategy is: low-priority AGVs are instructed to stagger their scheduled time windows by adjusting their speed;

[0095] The second anti-deadlock strategy is to plan a path so that the low-priority AGV can drive into the nearest and idle functional avoidance compartment to avoid the conflict point.

[0096] Calculate and compare the estimated time cost T1 of the first deadlock prevention strategy with the estimated time cost T2 of the second deadlock prevention strategy;

[0097] Select a deadlock prevention strategy with lower estimated time cost and instruct the low-priority AGV to execute it.

[0098] The estimated time cost refers to the additional time cost incurred in implementing deadlock prevention strategies. Its calculation method is as follows:

[0099] First, estimate the total time T required for the AGV to complete its current task as planned without conflict. 00 T 00 It can be obtained by dividing the total path length by the average speed of the AGV and adding the fixed pickup and drop-off time;

[0100] Subsequently, the total time T required to complete the current task under the first deadlock prevention strategy was estimated. 01 And the total time T required to complete the current task under the second deadlock prevention strategy. 02 .

[0101] Finally, the estimated time cost of the two deadlock prevention strategies is calculated:

[0102] T1=T 01 -T 00 ;

[0103] T2=T 02 -T 00 .

[0104] If T1≤T2, then the first anti-deadlock strategy is selected, and the low-priority AGV is instructed to adjust its speed to avoid conflicts between AGVs.

[0105] If T2 < T1, then the second anti-deadlock strategy is selected. The central dispatch system then controls the low-priority AGV to enter the nearest and idle functional avoidance compartment to avoid the conflict point. After the low-priority AGV enters the functional avoidance compartment, the central dispatch system will switch the status of the functional avoidance compartment to the occupied state. After the high-priority AGV passes the conflict point, the low-priority AGV will be instructed to leave the functional avoidance compartment so that it can continue to execute the unfinished task. At the same time, the status of the functional avoidance compartment will be switched to the idle state.

[0106] This implementation introduces a multi-strategy dynamic optimization mechanism, enabling the system to move away from relying on a single conflict resolution method and intelligently select the globally optimal solution based on real-time operating conditions (such as traffic flow, obstacle avoidance cabin position, and AGV speed status). This not only further improves the system's throughput efficiency but also significantly enhances its robustness in handling complex operating conditions.

[0107] The space of the obstacle avoidance compartment is large enough to accommodate at least one AGV to complete a turning operation;

[0108] The central dispatch system is also configured as follows:

[0109] Develop a return-to-home strategy for each AGV performing a task;

[0110] Return-to-home strategies include:

[0111] First return strategy: Control the AGV to continue moving along the one-way traffic loop network and return to the return target point in a loop;

[0112] The second return strategy is to control the AGV to enter the nearest and available functional obstacle avoidance compartment, complete the turnaround, and then return to the return target point.

[0113] The central dispatch system is also configured as follows:

[0114] Calculate and compare the estimated time T for executing the first return-to-home strategy. r1 (i.e., the time taken to return cyclically along the loop network) and the estimated time T for executing the second return strategy. r2 (That is, the total time spent going to the functional avoidance cabin, turning around, and returning);

[0115] Choose the return strategy with the shorter time consumption as the final return strategy, and formulate the return path for the AGV according to the final return strategy. After the AGV completes the picking or placing operation in front of the target shelf, it will return along this return path.

[0116] If T r1 ≤T r2 The first return-to-home strategy is selected as the final return-to-home strategy.

[0117] If T r2 <T r1 If so, then the second return strategy is selected as the final return strategy.

[0118] When the central dispatch system plans the travel path for each AGV performing a task, the travel path includes the path to the work point and the return path after completing the task. In the first implementation, the central dispatch system plans the travel path for the AGV according to the cyclical return strategy, which is the first return strategy mentioned above. However, in this implementation, the system calculates and compares the estimated time of the first and second return strategies based on real-time conditions to select the final return strategy with the shorter time and formulate the corresponding return path for the AGV. This dynamic return strategy selection mechanism allows the AGV to flexibly adjust its return method according to the actual scenario.

[0119] For example, after an AGV completes its task, if there is an available functional obstacle avoidance bay nearby, and the central dispatch system calculates that the estimated time for the second return strategy is less than that for the first return strategy, then the central dispatch system will select the second return strategy to plan the return path for the AGV. This allows the AGV to quickly turn around and return to its return target point more efficiently, reducing unnecessary looping time within the channel and improving overall operational efficiency.

[0120] Conversely, if the estimated time for the first return strategy is shorter after calculation, the AGV will continue to return along the one-way loop network, ensuring that the system operates optimally under different circumstances. This intelligent decision-making mechanism makes the intelligent warehousing system of this embodiment more flexible and efficient in handling the AGV return problem, further improving the intelligence and overall performance of the entire warehousing system.

[0121] This implementation method, by introducing a functional avoidance cabin and a dynamic optimization algorithm, enables intelligent scheduling of the entire warehousing system. Whether it's conflict resolution during task execution or return route planning after task completion, the system can select the globally optimal solution from multiple feasible options based on real-time benefit assessment. This two-way intelligent optimization mechanism ensures that the system maintains maximum efficiency under any operating condition, minimizing AGV idle rate, waiting time, and ineffective travel, thereby greatly improving the overall throughput and economy of the system.

[0122] The third implementation method:

[0123] Please see Figure 6 Unlike the second implementation, the entrance to the functional avoidance compartment is located at the end of the passage section and is connected to the passage section. The AGV can drive directly into the functional avoidance compartment by traveling in a straight line along the passage section.

[0124] The central dispatch system is also configured as follows:

[0125] Real-time monitoring of the operating status of each AGV;

[0126] When an AGV is determined to be faulty and unable to move, mark it as a faulty AGV and execute the fault cleanup procedure.

[0127] The troubleshooting procedure includes:

[0128] One AGV is selected from the other AGVs as the rescue AGV according to the preset rules;

[0129] Plan a rescue route to the malfunctioning AGV for the rescue AGV;

[0130] The rescue AGV is instructed to move to the location of the malfunctioning AGV and push it into the nearest available functional avoidance compartment.

[0131] The specific process by which the central dispatch system determines an AGV malfunction and executes the fault clearing procedure is as follows:

[0132] 1. Fault detection and diagnosis:

[0133] The central dispatch system maintains a periodic heartbeat communication mechanism with each AGV. By default, the central dispatch system receives a heartbeat signal from each AGV every first preset time interval (e.g., 1 second). If an AGV fails to report a heartbeat for a second preset time interval (e.g., 3 seconds), or actively uploads a fault code containing information such as drive error, power failure, or emergency stop triggering, the central dispatch system immediately determines that the AGV is a "faulty AGV" and marks its current location as a high-priority temporary obstacle point, prohibiting all other AGVs from reserving the time window of the passage segment where that location is located.

[0134] 2. Dynamic selection rules for rescue AGVs:

[0135] First priority: Idle and unloaded AGVs. AGVs that are currently idle and not carrying any materials should be prioritized for rescue missions to avoid disrupting other operations.

[0136] Second priority: AGVs that are currently idle but have tasks. If there are no idle AGVs, then an AGV that is currently executing a task but is in an idle state (e.g., has completed delivery and is returning to pick up goods) will be selected. The central scheduling system will interrupt its current low-priority task and record the interruption point for later recovery.

[0137] Third priority: Distance and capacity matching. Among the AGVs that meet the above conditions, the AGV with the shortest path to the fault point is selected. At the same time, the central dispatch system will verify the rated thrust and weight of the rescue AGV to ensure that it is physically capable of pushing the faulty AGV.

[0138] After the selection is completed, the central dispatch system issues an emergency rescue instruction to the rescue AGV.

[0139] 3. Rescue route planning and highest priority:

[0140] The central dispatch system plans a rescue path for the rescue AGV, directly leading to the malfunctioning AGV and from there to the nearest available functional avoidance compartment. This path is assigned the highest global priority. The central dispatch system will forcibly cancel all currently reserved time windows on this path and instruct existing AGVs on the path to implement avoidance strategies (such as entering the avoidance compartment in advance or slowing down and waiting) to make way for the rescue mission, ensuring an unobstructed rescue path.

[0141] 4. Promote collaboration between operation and safety:

[0142] Docking: Each AGV is equipped with a buffer anti-collision mechanism or a docking traction mechanism (the buffer anti-collision mechanism and the docking traction mechanism are existing technologies, and their specific structures will not be described in detail here). After the rescue AGV travels to the location of the faulty AGV, it uses the buffer anti-collision mechanism or the docking traction mechanism to physically dock with the faulty AGV.

[0143] Propulsion: After docking is completed, the rescue AGV will propel the faulty AGV forward along the rescue path at a low speed of a preset safe speed (e.g., 0.2 m / s).

[0144] Placement: Securely push the malfunctioning AGV into the nearest available obstacle avoidance compartment. Once the rescue AGV confirms that the malfunctioning AGV has completely entered the obstacle avoidance compartment, it reports mission completion to the central dispatch system.

[0145] 5. System Recovery:

[0146] After the malfunctioning AGV is cleared, the central dispatch system removes the temporary obstacle marker from its location and releases the occupied passageway resources. Subsequently, the central dispatch system restores normal dispatch order, and the rescue AGV can resume its interrupted task or enter idle mode according to instructions. Simultaneously, the central dispatch system will issue an alarm, prompting relevant technicians to promptly proceed to the functional avoidance compartment to maintain the malfunctioning AGV. Furthermore, if the malfunctioning AGV is already performing a task, the central dispatch system will also send a signal to the warehouse management system, instructing the warehouse management system to reassign the unfinished task of the malfunctioning AGV to other AGVs.

[0147] This implementation method achieves a leap from "deadlock prevention" to "deadlock resolution" by setting up a highly automated collaborative rescue mechanism. It transforms major system failures that traditionally require manual intervention into occasional events that the system can resolve on its own. This ensures that the unexpected failure of a single device will not lead to the long-term paralysis of the entire intelligent warehousing system, greatly improving the system's robustness, reliability, and availability, and meeting the stringent requirements of continuous and stable operation in high-demand industrial environments such as latex-based ground stations.

[0148] In light of current practical needs, the above-described embodiments of this invention are not limited to these specific implementations. Any changes made within the scope of knowledge possessed by those skilled in the art, without departing from the concept of this invention, still fall within the protection scope of this invention.

Claims

1. An intelligent storage system for a latex matrix ground station, comprising a warehouse, a plurality of shelves arranged in the warehouse, a plurality of AGVs, an inbound docking area, an outbound conveying line, a warehouse management system and a central dispatching system, channels for the AGVs to pass through are formed between the shelves and between the shelves and the inner walls of the warehouse, the warehouse management system is used to assign the AGVs with material inbound and outbound tasks, characterized in that: The channel is a one-way traffic loop network composed of multiple channel segments connected together, and the width of the channel segment matches the minimum safe traffic width required by a single AGV to perform material in-and-out warehouse tasks; The central scheduling system is in communication connection with the AGVs, and is configured to: plan a travel path for each AGV performing a task, and allocate resources in the time dimension for the channel segments contained in the path to generate a reserved occupancy time window for each channel segment; maintain a global space-time table containing the reserved occupancy time windows of all AGVs; detect whether there is an overlap conflict in the space-time resources between different AGVs by continuously comparing the reserved occupancy time windows in the global space-time table; when a conflict is detected, dynamically calculate the traffic priority of the conflicting AGV according to a preset rule; instruct the low-priority AGV to stagger its reserved time window through speed adjustment to avoid conflicts between AGVs; The reserved occupancy time window is a data structure that at least includes a start time, an end time, a channel segment identifier, and an AGV identifier field; The global space-time table is a Gantt chart form data structure with time as the horizontal axis and channel segments as the vertical axis, used to store and compare all reserved occupancy time windows; The warehouse is also distributed with multiple functional avoidance cabins; The central scheduling system is also configured to: generate a first anti-lock strategy and a second anti-lock strategy when a conflict is detected; The first anti-lock strategy is to instruct the low-priority AGV to stagger its reserved time window through speed adjustment; The second anti-lock strategy is to plan a path for the low-priority AGV to drive into the nearest and idle functional avoidance cabin to avoid; Calculate and compare the estimated time cost T1 of the first anti-lock strategy and the estimated time cost T2 of the second anti-lock strategy; Select the anti-lock strategy with smaller estimated time cost and instruct the low-priority AGV to execute.

2. An intelligent warehousing system for a latex substrate ground station as claimed in claim 1, wherein, The functional avoidance cabin is embeddedly installed in the inner wall of the warehouse, and its space size can at least accommodate one AGV for parking.

3. An intelligent warehousing system for a latex substrate ground station as claimed in claim 2, wherein, The space size of the functional avoidance cabin can also at least accommodate one AGV to complete the turning operation; The central scheduling system is also configured to: develop a return strategy for each AGV performing a task; The return strategy includes: The first return strategy: control the AGV to continue to move along the one-way traffic loop network and return to the return target point in a loop; The second return strategy: control the AGV to drive into the nearest and idle functional avoidance cabin, complete the turning, and then return to the return target point.

4. An intelligent warehousing system for a latex substrate ground station as claimed in claim 3, wherein, The central scheduling system is also configured to: calculate and compare the estimated time cost of executing the first return strategy and the estimated time cost of executing the second return strategy; select the return strategy with smaller time cost as the final return strategy, and develop a return path for the AGV according to the final return strategy.

5. An intelligent warehousing system for a latex substrate ground station as claimed in claim 4, wherein, The entrance of the functional avoidance cabin is located at the end of the channel segment and is connected with the channel segment; The central scheduling system is also configured to: monitor the running state of each AGV in real time; when it is determined that a certain AGV has failed and cannot move, mark this AGV as a failed AGV and execute a fault cleaning program.

6. An intelligent warehousing system for a latex substrate ground station as claimed in claim 5, wherein, The fault cleaning program includes: According to a preset rule, one AGV is selected from other AGVs as a rescue AGV; A rescue path to the fault AGV is planned for the rescue AGV; The rescue AGV is instructed to move to the fault AGV and push it into the nearest idle function avoidance cabin.

Images