Multi-mode AI micro server based on robot operation group and control method thereof

Abstract

The invention discloses a multi-mode AI micro-server based on a robot operation group and a control method thereof, and relates to the technical field of robot operation, and the multi-mode AI micro-server comprises a data transmission interface unit which is connected with external equipment, obtains external data of the external equipment, controls the external equipment based on control data, and transmits the external data to the robot operation group; the external equipment comprises operation equipment based on a robot operation group; the gateway unit is respectively connected with the data transmission interface unit and the artificial intelligence calculation storage unit; the artificial intelligence computing storage unit is used for receiving external data and carrying out real-time calibration on the external data based on a preset control algorithm to obtain control data output, and the external data comprises demand management data, resource management data and distribution management data; and the programmable logic control unit is connected with the artificial intelligence computing storage unit and the data transmission interface unit, receives the control data, programs the control data and then sends the control data to external equipment. The problem of low transportation efficiency of dynamic path planning of an AGV logistics distribution system is solved.

Description

Technical Field

[0001] The present application relates to the field of robot operation technology, and in particular to a multimodal AI microserver based on a robot operation group and a control method thereof. Background Art

[0002] The logistics dynamic management system of the IoT manufacturing plant is aimed at the factory's operating processes and combines the IoT for intelligent management.

[0003] However, existing IoT-based logistics dynamic management systems for manufacturing plants only utilize offline path planning and lack a global coordination mechanism, which can easily lead to local congestion or resource waste. For example, these systems can cause deadlocks and collisions when multiple AGVs (Automated Guided Vehicles) collaborate. Summary of the Invention

[0004] This application provides a multimodal AI microserver based on a robot operation group and a control method thereof, in order to at least solve the problem of low transportation efficiency in dynamic path planning of AGV logistics distribution systems in related technologies.

[0005] The present application provides a multimodal AI microserver based on a robot operation group, comprising: a programmable logic control unit, a data transmission interface unit, an artificial intelligence computing and storage unit, and a gateway unit, wherein: A data transmission interface unit, connected to an external device, for acquiring external data of the external device and controlling the external device based on control data, wherein the external device includes an operating device based on a robot operating group; a gateway unit, connected to the data transmission interface unit and the artificial intelligence computing storage unit, respectively, for converting data protocols between the connected external device and the data transmission interface unit and the artificial intelligence computing storage unit, respectively, to transmit data; The artificial intelligence computing and storage unit receives external data and performs real-time calibration based on a preset control algorithm to obtain control data before outputting it. The external data includes demand management data, resource management data, and distribution management data. The programmable logic control unit is connected to the artificial intelligence computing storage unit and the data transmission interface unit respectively, and is used to receive control data, program the control data, and then send it to the external device.

[0006] In some optional embodiments, the external device includes: an intelligent building system connected to the data transmission interface unit, configured to send building data to the data transmission interface unit, and to receive building control data and operate according to the building control data; a robot system connected to the data transmission interface unit, configured to send robot data to the data transmission interface unit, and to receive robot control data and operate according to the robot control data; The plant operation system is connected to the data transmission interface unit and is used to send plant operation data to the data transmission interface unit, and receive plant operation control data and operate according to the plant operation control data.

[0007] The present application also provides a control method for a multimodal AI microserver based on a robot operation group, which is applied to the multimodal AI microserver based on a robot operation group as described above, and the method includes: Pre-calibrate the work requirements of each work step according to the work order of the work station, perform optimal planning calculation based on the pre-calibrated data in combination with real-time dynamic data, and confirm the optimal planning result of the current work step until all work steps are calibrated; Confirm the optimal planning results of all work steps and control according to the optimal planning results to deliver the required items of the workstation from the shelf to the workstation.

[0008] In some optional embodiments, the step of pre-calibrating the work requirements in each work step according to the work order of the work station includes: Confirm the required item configuration of the workstation during operation; Pre-calibrate shelf operation instructions based on the confirmed target workstation demand items; Based on the confirmed target shelf, pre-calibrate the robot configuration; Based on the confirmed target robot configuration, perform elevator and path pre-calibration; Based on the confirmed target elevator and target path, pre-calibration of operation sequence dispatch is performed.

[0009] In some optional embodiments, the step of performing optimal planning calculation on pre-calibrated data in combination with real-time dynamic data includes: Obtain the configuration of required items at the workstation, and dynamically confirm the required items at the target workstation based on real-time items; Obtain pre-calibrated shelf operation instructions and confirm the target shelf based on real-time shelf dynamics; Obtain the pre-calibrated robot configuration and confirm the target robot configuration in combination with real-time robot dynamics; Obtain pre-calibrated elevators and paths, and confirm the target elevator and path based on the real-time elevator position and path; Obtain pre-calibrated job sequence dispatch and dynamically confirm the job sequence based on the job sequence.

[0010] In some optional embodiments, the step of performing optimal planning calculation on pre-calibrated data in combination with real-time dynamic data until all operation steps are calibrated includes: When the pre-calibrated data does not match the real-time dynamic data, re-calibration is performed based on the operation requirements of the current step until the optimal planning result of the current operation step is confirmed.

[0011] In some optional embodiments, the step of obtaining a pre-calibrated elevator and path and determining a target elevator and target path in combination with the real-time elevator position and path includes: Obtain real-time elevator location and operating parameters to detect whether the elevator's carrying capacity matches the demand for items being transported. If a mismatch is detected, disassemble the demand for items and confirm the target elevator. Obtain a pre-calibrated path, and detect whether the collision probability of the pre-calibrated path is lower than a preset collision value; When a conflict value greater than or equal to a preset value is detected, the path is replanned.

[0012] In some optional embodiments, the step of controlling according to the optimal planning result to deliver the required items of the workstation from the shelf to the workstation includes: Control the factory operation system to execute the lighting operation of the target shelf; Control the robot system so that the target robot moves to the target shelf to obtain the required items at the target workstation and then moves to the target elevator; Control the intelligent building system to make the target elevator run, and control the target robot to deliver the required items to the target workstation according to the target path.

[0013] In some optional implementations, the control method further includes: Obtain transportation cost, inventory cost, and energy cost, and add them together to get the total cost; Obtaining a delivery on-time rate and a delivery utilization rate, and adding the delivery on-time rate and the delivery utilization rate to obtain a total efficiency; The total cost and the total efficiency are subtracted to obtain an operation model, and the minimum value of the operation model is confirmed as the optimal planning calculation.

[0014] The present application also provides a computer device, comprising: A memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the above-mentioned control method of the multimodal AI microserver based on the robot operation group by executing the computer instructions. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0016] Figure 1 A structural diagram of a multimodal AI microserver based on a robot operation group provided in an embodiment of the present application; Figure 2 A schematic diagram of an application of a multimodal AI microserver based on a robot operation group provided in an embodiment of the present application; Figure 3 A flowchart of a control method for a multimodal AI microserver based on a robot operation group provided in an embodiment of the present application; Figure 4 A flowchart of another method for controlling a multimodal AI microserver based on a robot operation group provided in an embodiment of the present application; Figure 5 A flowchart of another method for controlling a multimodal AI microserver based on a robot operation group provided in an embodiment of the present application; Figure 6 A flowchart of another method for controlling a multimodal AI microserver based on a robot operation group provided in an embodiment of the present application; Figure 7 This is a structural diagram of a computer device of the present application. DETAILED DESCRIPTION

[0017] The following will be combined with the accompanying drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0018] It should be noted that, in the description of this application, the terms "comprises," "includes," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or device. The terms "first," "second," etc., in this application are used to distinguish similar objects, and are not used to describe a particular order or sequence.

[0019] In order to enable those skilled in the art to better understand the present application, the present application is further described in detail below with reference to the accompanying drawings and specific implementation methods.

[0020] The logistics dynamic management system of the IoT manufacturing plant is aimed at the factory's operating processes and combines the IoT for intelligent management.

[0021] However, existing IoT-based logistics dynamic management systems for manufacturing plants only utilize offline path planning and lack a global coordination mechanism, which can easily lead to local congestion or resource waste. For example, these systems can cause deadlocks and collisions when multiple AGVs (Automated Guided Vehicles) collaborate.

[0022] In addition, the relevant technologies also have the following technical defects. First, the limitations of static distribution route planning and fixed AGV vehicles: relying on offline path planning, it is unable to cope with real-time rapid responses such as dynamic orders and equipment failures; second, multi-objective conflicts: it is difficult to balance optimization goals such as task completion time, AGV energy consumption, and path conflicts; third, multi-AGV collaboration is inefficient: lack of a global coordination mechanism, prone to local congestion or waste of resources; fourth, poor model generalization: traditional algorithm parameters are fixed and difficult to adapt to different workshop layouts and process flows; fifth, there is a lack of synergy between production operations and industrial infrastructure (factory transport freight elevators, fire elevators) and other problems.

[0023] The embodiment of the present application provides a multimodal AI micro server based on a robot operation group, such as Figure 1 and Figure 2 As shown, the multimodal AI microserver based on the robot operation group includes: a programmable logic control unit, a data transmission interface unit, an artificial intelligence computing and storage unit, and a gateway unit, among which, A data transmission interface unit, connected to an external device, for acquiring external data of the external device and controlling the external device based on control data, wherein the external device includes an operating device based on a robot operating group; a gateway unit, connected to the data transmission interface unit and the artificial intelligence computing storage unit, respectively, for converting data protocols between the connected external device and the data transmission interface unit and the artificial intelligence computing storage unit, respectively, to transmit data; The artificial intelligence computing and storage unit receives external data and performs real-time calibration based on a preset control algorithm to obtain control data before outputting it. The external data includes demand management data, resource management data, and distribution management data. The programmable logic control unit is connected to the artificial intelligence computing storage unit and the data transmission interface unit respectively, and is used to receive control data, program the control data, and then send it to the external device.

[0024] Specifically, AI (Artificial Intelligence) refers to artificial intelligence. The programmable logic control unit serves as the core of the control system and is connected to external devices via a data transmission interface unit. It communicates with servers and external devices via a gateway unit. The AI ​​computing and storage unit communicates with the external devices via the data transmission interface unit. It calculates control data based on the external data using a preset control algorithm and then controls the external devices based on the control data. This allows for real-time calibration based on demand management data, resource management data, and delivery management data, enabling coordinated control of external devices. Items required during the operation and delivery routes are planned in real time to deliver the required items from the shelves to the workstations. This prevents deadlocks and collisions between robots during collaborative operations, significantly resolving resource conflicts. Furthermore, real-time planning prevents robots from reaching the shelf where the required item is located, only to have it removed by another workstation. This wastes energy and reduces efficiency.

[0025] Optionally, the entire chassis can be used as a mounting base for individual units. The programmable logic control unit, serving as the core of the control system, connects to various sensors and actuators (i.e., external devices) in field equipment via a data transmission interface unit. It also communicates with servers and AGVs (automated guided vehicles) via a gateway unit. The artificial intelligence computing and storage unit connects to the operational database, Supervisory Control and Data Acquisition (SCADA) system, Programmable Logic Controller (PLC), Manufacturing Execution System (MES), and intelligent building systems (such as building environment, energy, and elevator management) via data transmission interface units (such as switches and hubs) for data collection, analysis, and production management. The power supply system unit draws in external AC power, rectifies and stabilizes it, and then supplies power to other units via the server's internal busbar. In addition, reference Figure 1 The IoT lines for external devices are routed into the all-in-one server via bus ducts. Standardized connectors, such as PCIe (Peripheral Component Interconnect Express), USB-C (a USB form factor standard), and backplane buses, enable integrated electrical and communication connections. The technology complies with heat dissipation and EMC (electromagnetic compatibility) requirements.

[0026] Various sensors and actuators connected to field devices through data interface units can be converted into protocols for different interfaces. A master-slave architecture is adopted, that is, a programmable logic control unit acts as a master station, responsible for sending and receiving data frames, while field devices (such as sensors, actuators and drives) act as slave stations, which operate according to the instructions of the master station.

[0027] The multimodal AI microserver based on the robot operation group supports the hardware equipment structure with rapid installation, disassembly and function expansion. Its features include: factory equipment registration and verification mechanism, artificial intelligence automatically compares the fingerprint library of access devices, and identifies the compliance of devices.

[0028] In some feasible real-time ways, such as Figure 1 and Figure 2 As shown, the external device includes: an intelligent building system connected to the data transmission interface unit, configured to send building data to the data transmission interface unit, and to receive building control data and operate according to the building control data; a robot system connected to the data transmission interface unit, configured to send robot data to the data transmission interface unit, and to receive robot control data and operate according to the robot control data; The plant operation system is connected to the data transmission interface unit and is used to send plant operation data to the data transmission interface unit, and receive plant operation control data and operate according to the plant operation control data.

[0029] refer to Figure 1 、 Figure 2 and Figure 6 The intelligent building system primarily manages elevator location, routing, and timing. The robotics system primarily utilizes AGVs (Automated Guided Vehicles). The factory operations system primarily manages workstation requirements and shelf management. Simultaneously, real-time dynamic data from the intelligent building, robotics, and factory operations systems is captured. Using artificial intelligence and pre-defined control methods, the system generates real-time operational scenario data to plan the items needed during the operation and their delivery routes. The system also controls the coordinated operation of robots and elevators to deliver the required items from the shelves to the workstations.

[0030] An embodiment of the present application provides a control method for a multimodal AI microserver based on a robot operation group. The method is applied to the multimodal AI microserver based on a robot operation group as described above. The control method includes: Step 1: Pre-calibrate the work requirements of each work step according to the work order of the work station, perform optimal planning calculation based on the pre-calibrated data in combination with real-time dynamic data, and confirm the optimal planning result of the current work step until all work steps are calibrated; Specifically, the required items are first obtained according to the work order of the work station, and the operation steps for delivering the items are confirmed. Then, the operation requirements in each operation step are pre-calibrated. Then, the optimal planning calculation is performed in real time based on the pre-calibrated data to confirm the optimal planning result of the current operation step.

[0031] Specifically, the operation steps are divided into the first step to the nth step. The first step is pre-calibrated first, and the optimal planning calculation is performed for the pre-calibrated data of the first step based on real-time dynamic data to confirm the optimal planning result of the first step. Then, the operation requirements of the second step are pre-calibrated based on the optimal planning result of the first step, and the optimal planning calculation is performed for the pre-calibrated data of the second step based on real-time dynamic data to confirm the optimal planning result of the second step, and so on, until the calibration of the nth step is completed.

[0032] It should be noted that if the optimal planning result of the current operation step is confirmed based on the pre-calibrated data, the calibration of the current step is confirmed to be completed.

[0033] Specifically, pre-calibration is performed first, and then the optimal planning calculation is performed based on the pre-calibration data combined with the real-time dynamic data, that is, parallel calculation is performed for the calibration of each operation step.

[0034] Step 2: Confirm the optimal planning results of all operation steps, and control according to the optimal planning results to deliver the required items of the workstation from the shelf to the workstation.

[0035] Specifically, after confirming the optimal planning results of all the operation steps, the optimal planning results of all the operation steps are used as calibration results to control the robot and the elevator to deliver the required items of the workstation from the shelf to the workstation.

[0036] refer to Figure 3 Steps S3 to S6 are the specific process of controlling the optimal planning result, including: Step S3: Artificial intelligence simulation calculation confirms task or operation calibration; Step S4: Network collaborative operation control; Step S5: Shelf lighting guidance; Step S6: Robot operation driving.

[0037] Specifically, after confirming the optimal planning results for all work steps, each work process is calculated in a serial manner, and robots and elevators are controlled to deliver the required items from the shelves to the workstations. This combination of parallel and serial computing significantly improves efficiency compared to traditional, all-serial methods.

[0038] This application utilizes AI large-scale model technology to propose control models and calculation formulas for the electronics manufacturing industry and similar discrete manufacturing production scenarios. Real-time perception and analysis of environmental infrastructure (elevator operating conditions, workstation requirements) enables dynamic path planning for AGV logistics distribution systems to improve transportation efficiency. Microservers collaborate on decision-making and task scheduling, enabling real-time communication and data sharing between warehousing and logistics, improving human-machine collaboration efficiency and reducing operational difficulty and labor intensity. AI microservers are flexibly configurable and can be flexibly installed into existing enterprise control system cabinets for upgrades and retrofits. Standard control cabinets can also be quickly modified to meet enterprise needs, providing a comprehensive solution.

[0039] In some feasible implementations, the step of "pre-calibrating the operation requirements in each operation step according to the working order of the work station" in step 1 includes: Step S11: confirm the required item configuration of the workstation in the operation; Step S12: Pre-calibrate shelf operation instructions based on the confirmed target workstation required items; Step S13: Based on the confirmed target shelf, perform robot configuration pre-calibration; Step S14: Based on the confirmed target robot configuration, perform elevator and path pre-calibration; Step S15: Based on the confirmed target elevator and target path, pre-calibration of the operation sequence dispatch is performed.

[0040] Specifically, refer to Figure 3 In steps S11 to S15, the specific operation steps are to first confirm the items required by the workstation, then place the shelves where the confirmed items are located, then confirm the robot to perform the step among multiple robots, then plan the elevator and movement path of the robot, and finally confirm whether there is any conflicting operation sequence.

[0041] In some feasible implementations, the step of "performing optimal planning calculation based on pre-calibrated data in combination with real-time dynamic data" in step 1 includes: Step S21: Obtain the configuration of items required at the workstation, and dynamically confirm the items required at the target workstation based on the real-time items; Step S22: Obtain pre-calibrated shelf operation instructions and confirm the target shelf in combination with real-time shelf dynamics; Step S23: obtaining the pre-calibrated robot configuration and confirming the target robot configuration in combination with the real-time robot dynamics; Step S24: Obtain pre-calibrated elevators and paths, and determine the target elevator and target path based on the real-time elevator position and path; Step S25: Obtain the pre-calibrated job sequence dispatch, and dynamically confirm the job sequence in combination with the job sequence.

[0042] Specifically, refer to Figure 3 Steps S21 to S25 in the figure above represent steps S1 and S2 performed in parallel, forming a Y-shaped structure. Step S1 pre-calibrates each task, while step S2 uses an artificial intelligence algorithm to calculate relevant data for the task requirements. Therefore, steps S1 and S2 are performed in parallel.

[0043] For example, if the currently required item is pre-calibrated as a wrench in step S11, step S21 is to obtain the real-time wrench situation and confirm whether there is a wrench in the spare parts; if the shelf where the wrench is obtained is pre-calibrated as the 5th shelf in step S12, step S22 is to confirm whether there is a wrench in the shelf based on the real-time dynamics of the shelf, and if so, continue to execute; if it is determined in step S13 that the shelf where the wrench is obtained is the 5th shelf, the required robot is pre-calibrated, and the target robot is confirmed in step S23 based on the real-time dynamics of the robot such as the robot's position and power; if the target robot is determined in step S14, the elevator and path are pre-calibrated, and the target elevator and target path are confirmed in step S24 based on the real-time elevator position and path, specifically the currently idle elevator and the elevator close to the target robot, and the path with the shortest delivery distance; if the job sequence dispatch is pre-calibrated in step S15, the job sequence is dynamically confirmed in step S25, such as dynamic anti-collision calculation.

[0044] refer to Figure 4 The operation steps also include: controlling the robot to deliver the semi-finished products completed at the workstation or the items that need to be temporarily stored to the shelves. During this process, the target items completed at the workstation in the operation are confirmed, and the target items include semi-finished products or items that need to be temporarily stored; based on the confirmed target items, the shelf operation guidance is pre-calibrated; based on the confirmed target shelf, the robot configuration is pre-calibrated; based on the confirmed target robot configuration, the elevator and path are pre-calibrated; based on the confirmed target elevator and target path, the operation sequence dispatch is pre-calibrated.

[0045] In some feasible implementations, the step of "combining real-time dynamic data with pre-calibrated data to perform optimal planning calculations until all operation steps are calibrated" in step 1 includes: Step (1): When the pre-calibrated data does not match the real-time dynamic data, re-calibrate according to the operation requirements of the current step until the optimal planning result of the current operation step is confirmed.

[0046] Specifically, during the real-time dynamic calculation process, if the pre-calibrated data cannot be confirmed or is about to be used in other workstation operations, it is re-calibrated until the optimal planning result of the current operation step is confirmed.

[0047] For example, if the current pre-calibrated robot's battery life is insufficient to support delivery energy consumption or the robot is about to be used for another workstation, a new robot will be recalibrated. If the current pre-calibrated freight elevator is currently being used for another workstation, a new freight elevator will be recalibrated until the optimal freight elevator planning result is confirmed.

[0048] In some feasible implementations, step S24 includes: Step (1): Obtain the real-time elevator position and operating parameters to check whether the elevator's carrying capacity matches the demand for running items; Step (2): When a mismatch is detected, the target elevator is confirmed after disassembling the running item requirements; Step (3): obtaining a pre-calibrated path and detecting whether the conflict probability of the pre-calibrated path is lower than a preset conflict value; Step (4): When a conflict value higher than or equal to the preset conflict value is detected, the path is replanned.

[0049] Specifically, refer to Figure 4 The server inputs data from the physical layer to obtain robot status, shelf inventory scans, elevator operation data, and elevator operating parameters. After data preprocessing, the server calculates the path conflict probability and generates a navigation path when the conflict probability is below a preset conflict threshold (15%). Dynamic path replanning is performed when the conflict probability is above the preset conflict threshold (15%). By matching transportation demand with elevator (freight elevator) capacity, the server decomposes tasks using a task decomposition strategy when the elevator is overloaded. When the elevator is not overloaded, transportation instructions are generated.

[0050] In some feasible implementations, the step of "controlling according to the optimal planning result to deliver the required items of the workstation from the shelf to the workstation" in step 2 includes: Step (1): Control the factory operation system to execute the lighting operation of the target shelf; Step (2): Control the robot system so that the target robot moves to the target shelf to obtain the required items at the target workstation and then moves to the target elevator; Step (3): Control the intelligent building system to make the target elevator run, and control the target robot to deliver the required items to the target workstation according to the target path.

[0051] Specifically, the operation of each step is carried out in a serial manner. First, the factory operation system executes the lighting operation of the target shelf. Then, the robot system enables the target robot to move to the target shelf to obtain the required items at the target workstation and then move to the target elevator. The intelligent building system is controlled to make the target elevator run, and the target robot is controlled to deliver the required items at the target workstation to the target workstation according to the target path to complete the tasks of all steps.

[0052] It should be noted that during the control process, if there is an emergency, the emergency situation will be uploaded to the emergency center for separate processing. If normal timing control is possible, the normal timing control situation will be uploaded to the background.

[0053] In some feasible implementations, the control method of the multimodal AI microserver based on the robot operation group further includes: Step 1: Obtain transportation cost, inventory cost, and energy cost, and add them together to get the total cost; Specifically, the formula for the total cost is as follows: w1·(∑i transportation cost i + ∑j inventory cost j + ∑k energy consumption cost k).

[0054] Step 2: Obtain the delivery on-time rate and the delivery utilization rate, and add the delivery on-time rate and the delivery utilization rate to obtain the total efficiency; Specifically, the formula for total efficiency is as follows: w2·(∑m punctuality rate m+∑n utilization rate n).

[0055] Step 3: Subtract the total cost from the total efficiency to obtain the operation model, and confirm that the minimum value of the operation model is the optimal planning calculation.

[0056] Specifically, the optimal planning calculation model is minZ = w1·(∑i transportation cost i + ∑j inventory cost j + ∑k energy cost k) - w2·(∑m on-time rate m + ∑n utilization rate n). This results in a planning calculation that achieves the highest efficiency and utilization rate at the lowest cost. Transportation includes the distance of the route, energy consumption includes the up and down movement of the elevator, and inventory includes item management. On-time rate refers to the on-time rate of delivery of requested items, and utilization rate refers to the utilization rate of the robot and elevator. It should be noted that the linear calculation portion accounts for the fixed costs of transportation, including robot depreciation, while the nonlinear calculation portion includes the cubic relationship between energy cost and speed due to resistance during operation. Furthermore, an integer constraint requires that the number of shelf adjustments must be an integer; specifically, a maximum of a preset number of adjustments per month (five).

[0057] For example, transport timeliness is calculated as the standard deviation of order completion times, equipment utilization is calculated as AGV effective travel time / total travel time, energy efficiency is calculated as (total transport volume × distance) / total power consumption, elevator collaboration is calculated as appliance full load factor × time window matching, and abnormality recovery time is calculated as the average time from conflict detection to route planning. On-time performance is considered transport timeliness. Utilization includes equipment utilization.

[0058] The specific solution process is as follows: Data preprocessing and hierarchical modeling, specifically input data, workstation demand matrix, AGV initial position, elevator status, and ring path topology; then hierarchical modeling, the upper layer (task allocation): use integer programming to assign tasks to AGVs and determine shelf adjustment plans; the middle layer (path planning): generate cross-paths based on improved artificial intelligence algorithms and integrate elevator reservation mechanisms; the lower layer (speed optimization): dynamically adjust speed to balance energy consumption and punctuality.

[0059] The improved genetic algorithm solution includes coding design, where the chromosome contains the task assignment sequence (binary), the number of shelf adjustments (integer), and the speed segment value (real number); the fitness function is the inverse of the objective function Z, and a penalty term is added to deal with constraint violations (such as elevator overload penalty); crossover and mutation are performed, with two-point crossover used in the task assignment part and uniform mutation used in the shelf adjustment part. The speed gene is perturbed using a simulated annealing strategy to avoid local optimality.

[0060] Dynamic scheduling and replanning involve real-time monitoring, updating AGV positions, elevator queue lengths, and workstation demand changes every 5 seconds. A trigger mechanism initiates local replanning (based on rolling horizon optimization) when a path conflict or elevator wait time exceeds 30 seconds occurs.

[0061] The following Table 1 shows the specific data of the calibrated optimal planning results:

[0062] Table 1 Specifically, the work scenario operation process is realized based on the objective function combined with artificial intelligence data aggregation functions, such as data collection and classification methods.

[0063] The following Table 2 is a comparison of economic benefits with traditional solutions:

[0064] Table 2 Through path planning, global optimization, and dynamic adjustment, the performance standard quantification system is further developed. Efficiency indicators quantified by performance standards include single-machine performance: such as task completion time (standard value ≤5 minutes / order), idle rate (target ≤15%); cluster performance: throughput per unit time (such as a certain warehouse item / hour), task balance (variance ≤0.2); and data synchronization consistency (5G network packet loss rate ≤0.01%). AI guides AGV operation routes based on efficiency indicators quantified by performance standards, namely the S6 stage. In the S6 stage, the AI ​​algorithm server achieves standardized control of AGV operations through the following processes: initial path: assigning an energy-saving and efficient baseline route to each AGV; AI server real-time correction: monitoring AGV position through edge computing nodes to calculate the multi-machine collaborative operation system.

[0065] The control method of the multimodal AI microserver based on the robot operation group also includes: Step 1: Check whether the robot's power level is lower than the preset value; Step 2: If yes, control the robot to move to the charging position for charging.

[0066] refer to Figure 5 By detecting the robot's power level in real time, when it is detected that the power level is less than 20%, the robot is controlled to move to the charging position for charging.

[0067] Specifically, refer to Figure 5 , Figure 5 This is the control logic flow chart of the robot. After it is in the standby state, if it receives a task instruction, it will detect the validity of the task. When it is detected to be valid, it will use the path planning engine to detect whether the path conflicts. When a conflict is detected, it will re-plan the path and move to reach the target point. If it does not reach the target point, it will correct the positioning to the target point in real time and then perform loading and unloading operations. If it is in the interrupted state, it will detect whether the battery level of the robot is lower than the preset value (20%). If so, it will execute charging navigation, control the robot to move to the charging position, and automatically charge and connect for charging. After detecting that charging is completed, it will control the robot to move to the standby position for standby.

[0068] In some feasible implementations, a control method for a multimodal AI microserver based on a robot operation group includes: Step 1: Scan the inventory of items on the shelf and check whether the inventory is below the inventory safety threshold; Step 2: When it is detected that the inventory level is lower than the safety threshold, the replenishment priority is calculated and replenishment is controlled according to the replenishment priority.

[0069] Specifically, through inventory safety value detection, when insufficient inventory is detected, replenishment priorities are calculated and replenishment instructions are triggered. When sufficient inventory is available, the current inventory level is maintained. Then, control instructions are fused and sent to the robot, elevator controller, and cargo adjustment device for control, providing real-time feedback on the current status.

[0070] It is worth noting that the replenishment instruction can be sent to a robot that replenishes goods, or it can be sent to a user so that the user can replenish the goods.

[0071] Furthermore, shelf layout optimization can be implemented by relocating frequently visited shelves near elevators each month. For example, the top five most frequently visited shelves can be relocated near elevators, reducing cross-floor transport distances by ≥20%. A task allocation strategy prioritizes urgent orders to AGVs with a battery level greater than a preset level. For example, orders with a remaining time of less than 10 minutes are assigned to AGVs with a battery level greater than 80%. An elevator reservation mechanism allows AGVs to request an elevator 30 seconds in advance, and the system allocates a time window. For example, elevator A is available between 10:00:00 and 10:00:15. Optimizing dynamic routes, a clockwise and counterclockwise dual-channel design allows for dynamic direction switching during congestion. Specifically, this involves using lidar to monitor traffic flow and setting virtual passes at key nodes (such as intersections) to avoid deadlocks, reducing collision rates by 90%. Speed ​​control and energy management ensure a speed of 1.8 m / s on straight sections (maximizing on-time performance) and 0.8 m / s on curves and congested areas (reducing energy consumption and collision risk). Compared to a 1.5 m / s constant speed solution, energy consumption is reduced by 27%, and on-time performance only decreases by 3%. Factory implementation has verified that the AGV idle rate has dropped from 35% to below 10%. Cost optimization: The transport demand matching module reduces ineffective elevator starts and stops, reducing energy consumption by under 20%. Enhanced reliability: Response time for handling inventory anomalies has been reduced from 25 minutes with manual inspections to 2 seconds with automatic system triggering.

[0072] An embodiment of the present invention further provides a computer device having the above-mentioned logistics dynamic management system for the Internet of Things manufacturing plant.

[0073] See also Figure 7 , Figure 7 is a structural diagram of a computer device provided by an optional embodiment of the present invention, such as Figure 7As shown, the computer device includes: one or more processors 10, memory 20, and interfaces for connecting various components, including high-speed interfaces and low-speed interfaces. Various components utilize different buses to communicate with each other and can be installed on a common mainboard or installed in other ways as needed. The processor can process the instructions executed in the computer device, including instructions stored in the memory or on the memory to display the graphical information of the GUI on an external input / output device (such as, a display device coupled to the interface). In some optional embodiments, if necessary, multiple processors and / or multiple buses can be used together with multiple memories and multiple memories. Equally, multiple computer devices can be connected, and each device provides part of the necessary operations (for example, as a server array, a group of blade servers, or a multi-processor system). Figure 7 A processor 10 is taken as an example.

[0074] The processor 10 may be a central processing unit, a network processor, or a combination thereof. The processor 10 may further include a hardware chip. The hardware chip may be an application-specific integrated circuit, a programmable logic device, or a combination thereof. The programmable logic device may be a complex programmable logic device, a field programmable gate array, a general purpose array logic, or any combination thereof.

[0075] The memory 20 stores instructions that can be executed by at least one processor 10, so that the at least one processor 10 executes the method shown in the above embodiment.

[0076] The memory 20 may include a program storage area and a data storage area, wherein the program storage area may store an operating system and application programs required for at least one function; the data storage area may store data created based on the use of the computer device, etc. In addition, the memory 20 may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some optional embodiments, the memory 20 may optionally include a memory remotely located relative to the processor 10, and these remote memories may be connected to the computer device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0077] The memory 20 may include a volatile memory, such as a random access memory; the memory may also include a non-volatile memory, such as a flash memory, a hard disk or a solid-state drive; the memory 20 may also include a combination of the above types of memory.

[0078] The computer device further includes an input device 30 and an output device 40. The processor 10, the memory 20, the input device 30 and the output device 40 may be connected via a bus or other means. Figure 7 The bus connection is taken as an example.

[0079] The input device 30 can receive input digital or character information and generate key signal input related to user settings and function control of the computer device. Examples include a touch screen, keypad, mouse, trackpad, touchpad, pointing stick, one or more mouse buttons, trackball, joystick, etc. The output device 40 may include a display device, auxiliary lighting devices (e.g., LEDs), and tactile feedback devices (e.g., vibration motors). Such display devices include, but are not limited to, liquid crystal displays, light emitting diodes, monitors, and plasma displays. In some optional embodiments, the display device may be a touch screen.

[0080] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the above description has generally described the components and steps of each example according to their functions. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0081] The above is a detailed introduction to a multimodal AI microserver based on a robot operation group and its control method provided by this application. This article uses specific examples to illustrate the principles and implementation methods of this application. The description of the above embodiments is only used to help understand the method of this application and its core idea. It should be pointed out that for ordinary technicians in this technical field, without departing from the principles of this application, several improvements and modifications can be made to this application, and these improvements and modifications also fall within the scope of protection of the claims of this application.

Claims

1. A multimodal AI microserver based on a robot operation group, characterized in that: include: Programmable logic control unit, data transmission interface unit, artificial intelligence computing storage unit, gateway unit, among which, A data transmission interface unit, connected to an external device, for acquiring external data of the external device and controlling the external device based on control data, wherein the external device includes an operating device based on a robot operating group; a gateway unit, connected to the data transmission interface unit and the artificial intelligence computing storage unit, respectively, for converting data protocols between the connected external device and the data transmission interface unit and the artificial intelligence computing storage unit, respectively, to transmit data; The artificial intelligence computing and storage unit receives external data and performs real-time calibration based on a preset control algorithm to obtain control data before outputting it. The external data includes demand management data, resource management data, and distribution management data. The programmable logic control unit is connected to the artificial intelligence computing storage unit and the data transmission interface unit respectively, and is used to receive control data, program the control data, and then send it to the external device.

2. The multimodal AI microserver based on a robot operation group according to claim 1, characterized in that: The external device includes: an intelligent building system connected to the data transmission interface unit, configured to send building data to the data transmission interface unit, and to receive building control data and operate according to the building control data; a robot system connected to the data transmission interface unit, configured to send robot data to the data transmission interface unit, and to receive robot control data and operate according to the robot control data; The plant operation system is connected to the data transmission interface unit and is used to send plant operation data to the data transmission interface unit, and receive plant operation control data and operate according to the plant operation control data.

3. A control method for a multimodal AI microserver based on a robot operation group, characterized in that: The method is applied to the multimodal AI microserver based on a robot operation group as claimed in claim 1 or 2, and the method includes: Pre-calibrate the work requirements of each work step according to the work order of the work station, perform optimal planning calculation based on the pre-calibrated data in combination with real-time dynamic data, and confirm the optimal planning result of the current work step until all work steps are calibrated; Confirm the optimal planning results of all work steps and control according to the optimal planning results to deliver the required items of the workstation from the shelf to the workstation.

4. The control method of the multimodal AI microserver based on the robot operation group according to claim 3 is characterized in that: The step of pre-calibrating the work requirements in each work step according to the work order of the work station includes: Confirm the required item configuration of the workstation during operation; Based on the confirmed target workstation demand items, pre-calibrate the shelf operation instructions; Based on the confirmed target shelf, pre-calibrate the robot configuration; Based on the confirmed target robot configuration, perform elevator and path pre-calibration; Based on the confirmed target elevator and target path, pre-calibration of operation sequence dispatch is performed.

5. The control method of the multimodal AI microserver based on the robot operation group according to claim 4 is characterized in that: The step of performing optimal planning calculation based on pre-calibrated data in combination with real-time dynamic data includes: Obtain the configuration of required items at the workstation, and dynamically confirm the required items at the target workstation based on real-time items; Obtain pre-calibrated shelf operation instructions and confirm the target shelf based on real-time shelf dynamics; Obtain the pre-calibrated robot configuration and confirm the target robot configuration in combination with real-time robot dynamics; Obtain pre-calibrated elevators and paths, and confirm the target elevator and path based on the real-time elevator position and path; Obtain pre-calibrated job sequence dispatch and dynamically confirm the job sequence based on the job sequence.

6. The control method of the multimodal AI microserver based on the robot operation group according to claim 5 is characterized in that: The step of performing optimal planning calculation on pre-calibrated data in combination with real-time dynamic data until all operation steps are calibrated includes: When the pre-calibrated data does not match the real-time dynamic data, re-calibration is performed based on the operation requirements of the current step until the optimal planning result of the current operation step is confirmed.

7. The control method of the multimodal AI microserver based on the robot operation group according to claim 6 is characterized in that: The steps of obtaining the pre-calibrated elevator and path and determining the target elevator and target path in combination with the real-time elevator position and path include: Obtain real-time elevator location and operating parameters to detect whether the elevator's carrying capacity matches the demand for items being transported. If a mismatch is detected, disassemble the demand for items and confirm the target elevator. Obtain a pre-calibrated path, and detect whether the collision probability of the pre-calibrated path is lower than a preset collision value; When a conflict value greater than or equal to a preset value is detected, the path is replanned.

8. The control method of the multimodal AI microserver based on the robot operation group according to claim 3 is characterized in that: The step of controlling according to the optimal planning result to deliver the required items of the workstation from the shelf to the workstation includes: Control the factory operation system to execute the lighting operation of the target shelf; Control the robot system so that the target robot moves to the target shelf to obtain the required items at the target workstation and then moves to the target elevator; Control the intelligent building system to make the target elevator run, and control the target robot to deliver the required items to the target workstation according to the target path.

9. The control method of the multimodal AI microserver based on the robot operation group according to claim 8, characterized in that: The control method further includes: Obtain transportation cost, inventory cost, and energy cost, and add them together to get the total cost; Obtaining a delivery on-time rate and a delivery utilization rate, and adding the delivery on-time rate and the delivery utilization rate to obtain a total efficiency; The total cost and the total efficiency are subtracted to obtain an operation model, and the minimum value of the operation model is confirmed as the optimal planning calculation.

10. A computer device, characterized in that: include: A memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the control method of the multimodal AI microserver based on the robot operation group according to any one of claims 3 to 9 by executing the computer instructions.

Images