Multifunctional movable tool rack linkage method based on Internet of Things

Abstract

The invention relates to a multifunctional movable tool rack linkage method based on the Internet of Things, and the method comprises the steps: building the communication connection between tool racks through the Internet of Things, generating a corresponding rack position adjustment scheme when the tool racks detect a high-frequency connection region, transmitting an adjustment instruction to the adjacent tool racks, and controlling the adjacent tool racks to carry out the adjustment. In the process of adjusting the structure of the tool rack, the position change of a tool is tracked in real time by utilizing a computer vision method, and the handover fluency improvement rate is calculated by comparing the tool track change before and after adjustment, so that the effect of rack position adjustment is evaluated; the tools are classified and coded according to the multifunctional attributes of the tools, and when the tool rack is adjusted in a linkage mode, the tools with complementary functions are preferentially placed at the adjacent positions.

Description

Technical Field

[0001] The present application relates to the field of information technology, and in particular to a multifunctional activity tool rack linkage method based on the Internet of Things. Background Art

[0002] In the multifunctional active tool rack linkage system based on the Internet of Things, the impact of tool handover frequency on space division is a key technical issue. When the tool handover frequency in a certain area is high, the tool rack in that area needs to respond quickly and adjust the rack structure to meet the needs of high-frequency handovers. Therefore, a handover heat map generated according to the tool handover frequency is required.

[0003] However, how to dynamically adjust the rack structure between adjacent tool racks based on the handover heat map and ensure that the adjustment process does not affect the smoothness of tool handover is a complex technical challenge.

[0004] Specifically, the tool rack must first be able to monitor and analyze the frequency and location distribution of tool handovers in real time and generate a handover heat map. Then, the tool rack must identify high-frequency handover areas based on the heat map and determine whether the rack structure needs to be adjusted. If adjustment is required, the tool rack must also consider the direction of adjustment and how to ensure the accessibility and stability of the tools during the adjustment process.

[0005] In addition, since the adjustment of the tool rack may affect the layout and tool access of adjacent tool racks, real-time collaboration and optimization are required between multiple tool racks to ensure the operating efficiency and smooth handover of the entire system. This requires real-time information exchange and decision-making consultation between tool racks, and adjustment of their respective rack structures and tool allocations based on global optimization goals. Summary of the invention

[0006] In order to solve the above problems existing in the prior art, the purpose of this application is to provide a multifunctional active tool rack linkage method based on the Internet of Things.

[0007] The present application discloses a method for linking a multifunctional active tool rack based on the Internet of Things, comprising the following steps:

[0008] Step S101, by installing a radio frequency identification sensor on a tool rack in a target area, real-time data on tool handover frequency is acquired to identify the number of times a tool on the tool rack in the target area is used. If the number of times a tool is used within a preset time period is higher than a preset number of times threshold, the corresponding tool rack is determined to be a high-frequency handover area;

[0009] Step S102, using a clustering algorithm to perform spatial division on the high-frequency handover area, determining the cluster center of the high-frequency handover area, determining the position coordinates of the sub-areas in the high-frequency handover area according to the cluster center, and obtaining the tool rack structure model according to the position coordinates of the sub-areas;

[0010] Step S103, counting the number and direction of tool handovers in the sub-areas of the high-frequency handover area within a preset time period, generating a handover heat map, and determining an adjustment plan for adjacent tool racks using a fuzzy reasoning method based on the heat distribution in the map and in combination with the tool rack structure model;

[0011] Step S104, establishing a communication connection between tool racks through the Internet of Things, and when a tool rack detects a high-frequency handover area, generating a corresponding rack position adjustment plan, transmitting the adjustment instruction to the adjacent tool rack, and controlling the adjacent tool rack to adjust;

[0012] Step S105, during the tool rack structure adjustment process, using computer vision methods to track the position change of the tool in real time, and by comparing the tool trajectory changes before and after the adjustment, calculating the improvement rate of the handover smoothness;

[0013] Step S106, using a reinforcement learning algorithm, taking the tool rack structure model as input, obtaining an optimal rack layout solution, and applying it to the rack adjustment process;

[0014] Step S107, classifying and coding the tools, and placing tools with complementary functions in adjacent positions first when adjusting the tool rack in linkage.

[0015] Preferably, in step S101, the RF signal strength during tool handover is monitored, the tool rack usage is analyzed, the tool handover frequency is obtained, the tool usage time data is classified using a neural network algorithm, and the regional tool turnover rate and density are calculated. When the tool rack capacity is saturated, the regional handover intensity is evaluated, the handover frequency data is analyzed, and the high-frequency handover tool rack area is determined.

[0016] Preferably, in step S102, the tool rack position coordinates are analyzed to calculate the regional intersection frequency and density distribution, the k-means clustering algorithm is used to identify the high-frequency intersection area, and the sub-region boundaries are divided, the sub-region area and saturation are calculated, the sub-region structure data is generated, and the spatial arrangement characteristics and structural model of the tool rack are obtained through spatial relationship calculation.

[0017] Preferably, in step S103, the tool position signal collected by the sensor is analyzed to extract the tool movement direction and dwell time, generate a tool handover frequency statistics table, combine the tool movement trajectory and position information, calculate the handover path heat value, obtain a handover heat distribution map, use fuzzy inference algorithm to identify the handover hot spot area, and generate a tool rack position spacing table. If the length of the handover path between adjacent tool racks exceeds a threshold, an adjustment plan is generated according to the tool rack position structure model.

[0018] Preferably, in step S104, tool rack position sensor data is collected to generate a communication connection table, signal transmission delay is calculated to generate communication instructions, adjacent tool rack position relationship data is extracted and processed by deep learning to obtain rack adjustment instructions, and synchronous calculation is performed to achieve coordinated adjustment of rack positions.

[0019] Preferably, in step S105, an industrial camera array is used to capture tool motion images, extract position coordinates and track motion trajectories, calculate the moving speed and path length before adjustment, and generate a handover fluency benchmark value. After adjustment, images are captured again, and feature recognition is performed using a convolutional neural network. The adjusted handover fluency is calculated and compared with the benchmark value in time series. Finally, the handover fluency improvement rate is calculated to evaluate the rack adjustment effect.

[0020] Preferably, in step S106, the tool rack structure model data is read, the handover fluency index is determined to set the reward function, the rack structure parameters are initialized, and the layout plan is iteratively optimized to generate an optimized layout sequence, and the spatial distribution calculation is performed to obtain the layout convergence index. If the convergence threshold is met, multiple rounds of iterations are performed to obtain the optimal rack parameters and update the tool rack structure model.

[0021] Preferably, in step S107, tool attribute data and usage records are collected to generate tool feature data, and a tool function coding table is generated through classification by a clustering algorithm, the complementarity of tool function combinations is calculated, and function combination data is generated. The tool combination relationship is analyzed using a graph neural network, complementary function combinations are identified, tool proximity data is generated, and the tool rack layout is updated accordingly.

[0022] The multifunctional active tool rack linkage method based on the Internet of Things described in the present application has the advantages of real-time monitoring of tool usage by installing radio frequency identification sensors on the tool rack, identifying high-frequency handover areas, using the K-means clustering algorithm to spatially divide the high-frequency areas, determining the position coordinates of the sub-areas and updating the rack structure model, combining the handover heat map and fuzzy reasoning to generate a rack optimization plan, using the Internet of Things technology to achieve communication between tool racks, performing synchronous adjustments, applying computer vision to track tool position changes, evaluating adjustment effects, and using a reinforcement learning algorithm to continuously optimize the rack layout, improve handover fluency, consider the multifunctional properties of the tools, and place functionally complementary tools in adjacent positions.

[0023] The present invention realizes dynamic optimization of tool rack positions through intelligent perception, data analysis and automatic adjustment, and effectively improves tool handover efficiency and operation fluency. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 This is the process of a multifunctional activity tool rack linkage method based on the Internet of Things described in this application Figure 1 ;

[0025] Figure 2 This is the process of a multifunctional activity tool rack linkage method based on the Internet of Things described in this application Figure 2 ;

[0026] Figure 3 This is the process of a multifunctional activity tool rack linkage method based on the Internet of Things described in this application Figure 3 . DETAILED DESCRIPTION

[0027] like Figure 1-Figure 3 As shown, the multifunctional activity tool rack linkage method based on the Internet of Things described in the present application includes the following steps:

[0028] Step S101, by installing a radio frequency identification sensor on a tool rack in a target area, real-time data on tool handover frequency is acquired to identify the number of times a tool on the tool rack in the target area is used. If the number of times a tool is used within a preset time period is higher than a preset number of times threshold, the corresponding tool rack is determined to be a high-frequency handover area;

[0029] Step S102, using a clustering algorithm to perform spatial division on the high-frequency handover area, determining the cluster center of the high-frequency handover area, determining the position coordinates of the sub-areas in the high-frequency handover area according to the cluster center, and obtaining the tool rack structure model according to the position coordinates of the sub-areas;

[0030] Step S103, counting the number and direction of tool handovers in the sub-areas of the high-frequency handover area within a preset time period, generating a handover heat map, and determining an adjustment plan for adjacent tool racks using a fuzzy reasoning method based on the heat distribution in the map and in combination with the tool rack structure model;

[0031] Step S104, establishing a communication connection between tool racks through the Internet of Things, and when a tool rack detects a high-frequency handover area, generating a corresponding rack position adjustment plan, transmitting the adjustment instruction to the adjacent tool rack, and controlling the adjacent tool rack to adjust;

[0032] Step S105, during the tool rack structure adjustment process, using computer vision methods to track the position change of the tool in real time, and by comparing the tool trajectory changes before and after the adjustment, calculating the improvement rate of the handover smoothness;

[0033] Step S106, using a reinforcement learning algorithm, taking the tool rack structure model as input, obtaining an optimal rack layout solution, and applying it to the rack adjustment process;

[0034] Step S107, classifying and coding the tools, and placing tools with complementary functions in adjacent positions first when adjusting the tool rack in linkage.

[0035] like Figure 1-Figure 3 As shown, in step S101, by installing a radio frequency identification sensor on a tool rack in a target area, tool handover frequency data is acquired in real time to identify the number of times a tool on the tool rack in the target area is used. If the number of times a tool is used within a preset time period is higher than a preset number of times threshold, the corresponding tool rack is determined to be a high-frequency handover area;

[0036] Obtain the radio frequency signal strength collected by the regional sensor array at the tool handover time, and compare the tool rack parking duration and access duration according to the radio frequency signal strength to obtain tool handover frequency data;

[0037] For the tool handover frequency data, a neural network algorithm is used to classify and aggregate the tool parking duration and access duration to obtain the regional tool turnover rate and regional tool density;

[0038] If the capacity saturation of the regional tool rack reaches the saturation threshold, the regional handover intensity is evaluated based on the tool turnover rate and the tool density in the regional area, based on the tool parking duration and the tool access duration data;

[0039] Based on the analysis of tool handover frequency data, the peak period of regional handover is obtained to determine the high-frequency handover tool rack area.

[0040] Specifically, in step S101, a radio frequency identification network is built through a regional sensor array to obtain the radio frequency signal strength at the tool rack position at the time of tool handover, and the tool rack parking duration and access duration are compared according to the radio frequency signal strength, and the tool handover frequency data is recorded in the radio frequency data processor;

[0041] For the tool handover frequency data, a neural network algorithm is used to classify and aggregate the tool parking duration and access duration, and the regional tool turnover rate and regional tool density are obtained from the cluster data, and the time period interval handover frequency threshold is set in the radio frequency data processor;

[0042] If the regional tool rack capacity saturation reaches a preset threshold, the regional handover intensity is evaluated in the radio frequency data processor based on the tool parking duration and access duration data of the regional tool rack according to the regional tool turnover rate and regional tool density;

[0043] The regional handover intensity data is analyzed in time series by a dynamic programming algorithm, and the peak period of regional handover is calculated from the tool handover frequency data, and the high-frequency handover tool rack area is determined in the radio frequency data processor;

[0044] The RFID network in the tool rack area is composed of an array of multiple RFID sensors. Each sensor is responsible for detecting RFID signals within a fixed area, usually with a coverage radius of 3 meters. There are overlapping areas between sensors to ensure signal continuity. The RFID signal strength decays with distance in accordance with the inverse square relationship. When the RFID tag enters the sensor detection range, the signal strength value received by the sensor changes with distance. The movement status of the tool can be judged by the signal strength change characteristics, including installing 12 sensors in a certain area to form a 3×4 matrix layout, with a distance of 2.5 meters between adjacent sensors and a coverage area of ​​about 54 square meters. The tool handover frequency data received by the RFID data processor contains information such as tool number, access time, return time, signal strength, etc. The neural network algorithm classifies and aggregates these data to identify the tool usage mode;

[0045] In the machining area of ​​a workshop, the monitoring data of the screwdriver tool group showed that the average handover frequency during the day shift was 8 times per hour, the median tool parking time was 25 minutes, and the median tool access time was 15 minutes. Therefore, the tool turnover rate in this area was 0.32 times / hour, and the regional tool density was 0.15 per square meter;

[0046] The regional tool rack capacity saturation reflects the usage status of the tool rack. When the saturation exceeds the preset threshold of 0.85, it indicates that the tool rack load is high and it is necessary to pay special attention to the tool handover situation in this area.

[0047] In the tool rack area of ​​a certain assembly workshop, a week of data analysis revealed that the screwdriver tool set had the highest handover intensity between 9 a.m. and 11 a.m., with an average of 12 handovers per hour. The tool parking time was reduced to 15 minutes, and the tool density in the area increased to 0.25 per square meter.

[0048] Time series analysis shows that tool use presents periodic characteristics. The dynamic programming algorithm is used to analyze the regional handover intensity data for 30 consecutive days to identify the time distribution pattern of the high-frequency handover area. It is found from the data of the machining area that the tool handover peaks from 9:00 to 11:00 a.m. and from 2:00 to 4:00 p.m. every day. The tool handover frequency during these two periods is 2.5 times that of the usual time. The regional handover intensity reaches 0.48 times / square meter / hour, and the capacity saturation of the tool rack is maintained above 0.92. Based on this, it is determined that the tool rack area during this period belongs to the high-frequency handover area.

[0049] The RF data processor monitors the changes of these indicators in real time. When it detects that the tool handover frequency in a certain area is always higher than the set threshold within a period lasting more than 1 hour, it automatically marks the area as a high-frequency handover area.

[0050] At the quality inspection station at the end of the production line, due to the need for process connection, the frequency of use of inspection tools is relatively high. The tool handover interval is generally within 10 minutes, and the regional handover intensity reaches 0.65 times / square meter / hour. This high-intensity use state will continue throughout the entire work shift. Therefore, this area is identified as a stable high-frequency handover area;

[0051] The RF data processor continuously monitors tool handover data, builds a dynamic feature model of tool usage, and achieves accurate identification of high-frequency handover areas.

[0052] like Figure 1-Figure 3 As shown, in step S102, a clustering algorithm is used to spatially divide the high-frequency handover area, determine the cluster center of the high-frequency handover area, determine the position coordinates of the sub-area in the high-frequency handover area according to the cluster center, and obtain the tool rack structure model according to the position coordinates of the sub-area;

[0053] Acquire a spatial position coordinate sequence of the tool rack, calculate the region handover frequency from the spatial position coordinate sequence, and obtain density distribution data of the tool rack in a spatial data processor;

[0054] Using the k-means clustering algorithm to perform spatial clustering calculation based on the density distribution data of the tool rack, identifying the regional handover heat distribution characteristics from the density distribution data, and obtaining the cluster center point coordinates of the high-frequency handover area;

[0055] Perform sub-region boundary division according to the cluster center point coordinates and the spatial cluster distance, calculate the sub-region area size and the region saturation degree from the sub-region boundary data, and generate sub-region structure data in the spatial data processor;

[0056] The spatial relationship calculation is performed on the sub-region structure data by a spatial data processor, the spatial arrangement characteristics of the tool rack are obtained from the density distribution data of the tool rack, and the tool rack position structure model is obtained.

[0057] Specifically, in step S102, the spatial position coordinate sequence of all tool racks in the high-frequency handover area is obtained through the regional sensor network, the regional handover frequency is calculated from the tool rack position coordinate data, and the density distribution data of the tool rack is generated in the spatial data processor;

[0058] The k-means clustering algorithm is used to perform spatial clustering calculation on the density distribution data of the tool rack, and the regional handover heat distribution characteristics are identified from the density distribution data, and the cluster center point coordinates of the high-frequency handover area are obtained in the spatial data processor;

[0059] According to the cluster center point coordinates and the spatial cluster distance, the high-frequency intersection area is divided into sub-area boundaries, the sub-area area size and the area saturation degree are calculated from the sub-area boundary data, and the sub-area structure data is generated in the spatial data processor;

[0060] The spatial relationship of the sub-region structure data is calculated by the spatial data processor, the spatial arrangement characteristics of the tool rack are obtained from the density distribution data of the tool rack, and the tool rack position structure model is constructed in the spatial data processor;

[0061] The regional sensor network adopts a grid layout, with a sensor node installed every 5 meters in the workshop tool rack area, and each node has independent coordinate positioning function;

[0062] By measuring the signal strength of the radio frequency tag on the tool rack, the precise spatial position coordinates of the tool rack are determined;

[0063] An 8×6 sensor network was deployed in an assembly workshop, covering an area of ​​240 square meters, collecting the position data of 48 tool racks in real time, with a tool rack position coordinate accuracy of 0.3 meters;

[0064] The spatial data processor generates density distribution data of tool racks by calculating the handover frequency of tool racks, showing the number of tool racks per unit area and their usage frequency;

[0065] The k-means clustering algorithm is calculated based on the spatial distribution characteristics of tool racks, and the center of the area where tool handover activities are frequent is found in the density distribution data;

[0066] In practical applications, by analyzing the density distribution data of 300 sampling points in a production line area, four significant cluster centers were identified. The average handover frequency of these center points was 3.2 times that of the surrounding areas.

[0067] The handover frequency of the cluster center near the main assembly station is the highest, reaching 12 times per hour, and the tool rack density is 0.4 per square meter. In the process of dividing the sub-area boundaries, the cluster center point is used as the benchmark and the boundary range is determined in combination with the spatial cluster distance;

[0068] In the case of a processing workshop, the minimum distance between the four cluster centers is 8 meters, so the sub-region radius is set to 4 meters, forming four non-overlapping circular sub-regions;

[0069] Each sub-area has an area of ​​about 50 square meters and contains 8 to 12 tool racks. By calculating the ratio of the number of tool racks to the area of ​​the area, the regional saturation index is obtained. The saturation of the busiest sub-area reaches 0.85;

[0070] The spatial data processor performs spatial relationship analysis based on the sub-area structure data and calculates the relative positions and spacings between tool racks. In the example of the machining area, the average spacing of the tool racks is 1.5 meters, and the minimum spacing is not less than 1 meter, forming a grid-like distribution structure.

[0071] By analyzing the density distribution data of tool racks, it was found that the frequency of use of tool racks near the operating station was significantly higher than that of the side away from the station. Based on this, a gradient layout was formed in the tool rack structure model. The density of tool racks gradually decreased from the periphery of the station to the outside, and the density ratio was about 3:2:1.

[0072] In the final assembly area at the end of the production line, due to the close connection of processes and frequent use of tools, higher requirements are placed on the layout of tool racks;

[0073] Through the calculation of the spatial data processor, the 12 tool racks are divided into three sub-areas, each of which is fan-shaped and has an area of ​​about 25 square meters;

[0074] The tool racks in the sub-areas are arranged in an arc shape, with the spacing gradually increasing from 1 meter on the inside to 1.8 meters on the outside, which not only ensures a high density of tool racks but also avoids mutual interference when tools are handed over;

[0075] The regional saturation is maintained between 0.75 and 0.82, and the spatial layout characteristics of the tool rack structure model are highly consistent with the workstation operation lines.

[0076] like Figure 1-Figure 3 As shown, in step S103, the number and direction of tool handovers in the sub-areas of the high-frequency handover area within a preset time period are counted to generate a handover heat map, and according to the heat distribution in the map and in combination with the tool rack structure model, a fuzzy reasoning method is used to determine the adjustment plan for adjacent tool racks;

[0077] Receiving tool position signals collected by a regional sensor array, extracting tool movement direction and dwell time data from the tool position signals, and generating a tool handover frequency statistics table in a spatial data processor according to the movement direction and dwell time data;

[0078] The tool handover frequency statistics table and tool movement trajectory data are used in combination with the position information in the tool rack structure model to calculate the heat value of the handover path data, and a handover heat distribution map is obtained in the spatial data processor;

[0079] Executing a fuzzy inference algorithm according to the handover heat distribution map, performing a spatial correlation operation on the heat distribution data, identifying the handover hotspot area between tool racks, and obtaining a tool rack position spacing table in the spatial data processor;

[0080] If the tool rack position spacing table and the handover heat distribution map show that the handover path length of adjacent tool racks exceeds the handover path length threshold, position parameters are obtained from the tool rack structure model, and a tool rack adjustment plan is generated in the spatial data processor.

[0081] Specifically, in step S103, the tool position data of the sub-areas in the high-frequency handover area within a preset time period is collected by the regional sensor array, the tool movement direction and the dwell time are extracted from the sensor received signal, and a tool handover frequency statistics table is generated in the spatial data processor;

[0082] According to the tool handover frequency statistics table and tool movement trajectory data, combined with the position information in the tool rack structure model, the regional heat value is calculated from the handover path data within a preset time period, and a handover heat distribution map is generated in the spatial data processor;

[0083] The fuzzy inference algorithm is used to perform spatial correlation calculation on the handover heat distribution map, the handover hotspot area between tool racks is identified from the heat distribution data, and the tool rack position spacing table is generated in the spatial data processor;

[0084] According to the tool rack position spacing table and the handover heat distribution map, the handover path length between adjacent tool racks is quantitatively calculated, the position adjustment parameters are obtained from the tool rack position structure model, and the tool rack position adjustment plan is obtained in the spatial data processor;

[0085] The area sensor array obtains tool movement information by real-time monitoring of changes in RF signal strength. Each sensor covers a circular area with a radius of 3 meters. Adjacent sensors overlap by 0.5 meters to eliminate blind spots. In the high-frequency handover area of ​​an assembly workshop, a 5×4 sensor array is arranged, covering an area of ​​120 square meters.

[0086] The spatial data processor records the tool position data every 30 seconds. Statistics show that the tool moves back and forth between the workbench and the tool rack. The average stay time of a single tool on the workbench is 15 minutes, and the average stay time on the tool rack is 8 minutes.

[0087] The tool handover frequency statistics table records the use of tools in different time periods, including tool number, pick-up time, return time, movement trajectory and other information;

[0088] During the 8-hour morning shift, the electric wrench tool group in the machining area of ​​a workshop generated an average of 24 handover records per hour, with the highest handover frequency between 9 a.m. and 11 a.m., reaching 36 times per hour.

[0089] The tool rack structure model shows that there are 16 tool racks in the area, arranged in a 4×4 matrix layout, with a distance of 1.2 meters between adjacent tool racks;

[0090] The handover heat distribution map uses a heat map to present the spatial distribution characteristics of tool handover activities. The colors from blue to red represent the change of heat values ​​from low to high.

[0091] In the case of a general assembly workshop, the tool rack area near the workstation showed obvious hot spot distribution, with the highest heat value reaching 0.85, while the heat value of the area far away from the workstation was only 0.2;

[0092] The fuzzy inference algorithm calculated the thermal gradients of adjacent areas and identified four obvious hotspot areas where the tool rack spacing was relatively small, with an average of 0.8 meters;

[0093] The tool rack position spacing table records the relative distance between each tool rack, and determines the optimization direction of the tool rack position by analyzing the length of the handover path;

[0094] In the example of the machining area, through the analysis of the handover heat data for 7 consecutive days, it was found that the handover frequency between the tool racks numbered A3 and A4 was the highest, 8 times per hour on average, but the distance between them reached 1.5 meters, which was significantly higher than the reasonable distance;

[0095] The spatial data processor calculates the position adjustment parameters based on the tool rack structure model and recommends moving the A4 tool rack 0.4 meters toward the A3 direction to shorten the distance between the two to 1.1 meters;

[0096] In the quality inspection workstation area, due to the frequent replacement of different types of inspection tools required by the process, tool handover activities show obvious group characteristics;

[0097] The handover heat distribution map shows that six tool racks are concentrated in a circular area with a diameter of 2 meters, forming a highly active area with a heat value of 0.92;

[0098] By analyzing the tool movement trajectory, it was found that adjacent tool racks were frequently used, with an average of 15 handover records per hour;

[0099] The spatial data processor calculates the optimal sector layout plan based on the heat distribution characteristics, evenly distributes the six tool racks in a 120-degree sector, and uniformly adjusts the distance between adjacent tool racks to 0.8 meters.

[0100] like Figure 1-Figure 3 As shown, in step S104, a communication connection between tool racks is established through the Internet of Things. When a tool rack detects a high-frequency handover area, a corresponding rack position adjustment plan is generated, and the adjustment instruction is transmitted to the adjacent tool rack to control the adjacent tool rack to adjust;

[0101] Collecting tool rack position sensor data, wherein the tool rack position sensor data is used to generate a tool rack communication connection table;

[0102] Calculate the signal transmission delay according to the network topology relationship in the tool rack communication connection table, and use the signal transmission delay to generate a rack communication instruction;

[0103] Extracting adjacent tool rack position relationship data from the rack position communication instructions, and obtaining rack position adjustment instructions after the adjacent tool rack position relationship data is processed by a deep learning algorithm;

[0104] The positions of adjacent tool racks are synchronously calculated according to the rack adjustment instructions, and the rack movement data obtained by the synchronous calculation is used for the coordinated adjustment of the racks.

[0105] Specifically, in step S104, a mesh communication network is established between tool racks through a wireless communication module, tool rack position data and a network topology are obtained from a tool rack position sensor, and a tool rack communication connection table is generated in a communication data processor;

[0106] According to the network topology relationship in the tool rack communication connection table, the signal transmission delay and connection strength between the tool racks are calculated, the tool rack handover status is obtained from the high-frequency handover area data, and the rack position communication instructions are generated in the communication data processor;

[0107]

[0108] , M represents the total number of tool racks, dik represents the physical distance between tool rack i and tool rack k, vk represents the signal transmission speed of tool rack k, and the average signal transmission delay of tool rack i is calculated by summing up the transmission delays of all adjacent tool racks and taking the average value;

[0109]

[0110] L represents the number of tool racks directly connected to tool rack i, βil represents the connection bandwidth between tool rack i and tool rack l, αl represents the signal processing capability of tool rack l, and the average connection strength of tool rack i is calculated by summing up the connection strengths of all directly connected tool racks and taking the average value;

[0111] The tool rack status in the high-frequency handover area is detected in real time through a deep learning algorithm, the position relationship data of adjacent tool racks are extracted from the rack communication instructions, and rack adjustment instructions are generated in the communication data processor;

[0112] According to the rack adjustment instruction and the tool rack handover status, the position relationship between adjacent tool racks is synchronously calculated, rack movement data is obtained from the adjustment execution parameters, and rack coordinated adjustment is performed in the communication data processor;

[0113] The mesh communication network between tool racks is established through wireless communication modules. Each tool rack is equipped with a communication node, the operating frequency is 2.4GHz, and the communication radius is 10 meters;

[0114] In an assembly workshop, 16 tool racks are arranged in a 4×4 grid layout, and the physical distance between adjacent tool racks is 2 meters;

[0115] The tool rack position sensor collects position data at a frequency of 0.1 seconds with an accuracy of 0.01 meters. The communication data processor generates a tool rack communication connection table based on this data, recording the connection status and signal strength between nodes;

[0116] The communication quality is calculated based on the physical distance between tool racks and the signal strength. In an example of a machining area, the average signal transmission delay between adjacent tool racks is 5 milliseconds, and the connection strength reaches -65dBm;

[0117] Data from the high-frequency handover area showed that the four tool racks near the operating station had the highest handover frequency, reaching an average of 32 times per hour. Based on this, the communication data processor generated the highest priority rack communication instructions;

[0118] The deep learning algorithm identifies the spatial distribution characteristics of tool handover activities by analyzing tool rack status data in real time. In the case of the final assembly workshop, the algorithm found that tool racks numbered A2, A3, B2, and B3 formed a high-frequency handover cluster. The number of handovers between these four tool racks accounted for 75% of the total number of handovers, and the average handover interval was 4 minutes.

[0119] Based on this feature, the communication data processor generates specific instructions for adjusting the tool rack position, including moving direction and distance information. During the coordinated adjustment of the rack position, the tool rack moves synchronously according to the adjustment instructions;

[0120] In the quality inspection area, six tool racks were originally arranged in a straight line with a spacing of 1.5 meters. Frequent handover activities made it inconvenient to take and put tools. The communication data processor calculated the optimal arc layout plan and instructed the tool racks to adjust their positions synchronously to form an arc with a radius of 2 meters. The distance between adjacent tool racks was shortened to 1.2 meters.

[0121] During the adjustment process, the moving speed of each tool holder is maintained at 0.1 m / s, and the position deviation is controlled within 0.02 m. In the transfer area at the end of the production line, due to the close connection of processes and frequent changes in tool use, higher requirements are placed on the position of the tool holder;

[0122] The communication data processor analyzed the handover data for one week and found that the requirements for tool rack layout varied in different periods;

[0123] In the morning process, the tool racks are arranged in a star shape, with eight tool racks evenly distributed around the central workstation; in the afternoon process, the layout needs to be adjusted to a strip shape, with the tool racks arranged in two parallel rows;

[0124] The rack adjustment instructions generated by the communication data processor can achieve smooth switching between the two layouts. The transition time is controlled within 45 seconds. During this period, each tool rack maintains communication connection and the position adjustment accuracy reaches 0.05 meters.

[0125] The collaborative adjustment method of the tool rack fully considers the dynamic characteristics of tool handover activities. The rack adjustment instructions contain two types of information: absolute coordinates and relative positions, ensuring that the tool rack can maintain the predetermined spatial relationship and adapt to local handover requirements.

[0126] In actual applications, a typical adjustment cycle includes three stages: detection, calculation, and execution, with a total time of no more than 90 seconds. During the adjustment process, the tool holder always remains in normal working condition.

[0127] like Figure 1-Figure 3 As shown, in step S105, during the tool rack structure adjustment process, the position change of the tool is tracked in real time using a computer vision method, and the handover smoothness improvement rate is calculated by comparing the tool trajectory changes before and after the adjustment, so as to evaluate the effect of the rack adjustment;

[0128] Obtain a tool motion image sequence acquired by the industrial camera array in the rack adjustment area, and obtain tool position coordinate data from the image data collector;

[0129] Performing target tracking processing on the tool position coordinate data to obtain tool motion trajectory data, storing the tool motion trajectory data in a trajectory processor, and obtaining the tool motion trajectory before the rack position is adjusted;

[0130] According to the tool motion trajectory and tool position coordinate data before the rack position adjustment, the moving speed and motion path length of the tool before the adjustment are calculated by the trajectory processor;

[0131] Extract the tool handover time interval from the target tracking data, and generate the handover smoothness benchmark value before adjustment through the trajectory processor;

[0132] Obtain the tool motion image sequence collected after the rack position is adjusted, and use convolutional neural network to perform feature recognition processing on it;

[0133] Obtaining the adjusted tool motion trajectory from the trajectory processor, and calculating the adjusted current value of the handover smoothness through the trajectory processor;

[0134] According to the handover fluency benchmark value and the current value, the tool motion trajectory data is subjected to time series comparison processing; the handover fluency improvement rate is calculated from the trajectory change data, and the rack position adjustment effect is determined by the trajectory processor.

[0135] Specifically, in step S105, the tool motion image sequence is collected in the rack adjustment area by the industrial camera array, the tool position coordinate data is obtained from the image data collector, the tool motion trajectory data is generated by the target tracking algorithm, and the tool motion trajectory before the rack adjustment is recorded in the trajectory processor;

[0136] According to the tool motion trajectory and tool holder position data, the moving speed and motion path length of the tool before adjustment are calculated in the trajectory processor, the tool handover time interval is extracted from the position tracking data, and the handover smoothness benchmark value before adjustment is generated in the trajectory processor, where:

[0137]

[0138] vi represents the moving speed of the tool in the i-th time period, Pi represents the position vector of the tool at the i-th time point, ti represents the time at the i-th time point, and the average moving speed of the tool in the time period is obtained by calculating the position difference between two adjacent time points and dividing it by the time difference;

[0139]

[0140] L represents the path length of the tool during the entire motion process, Pi represents the position vector of the tool at the i-th time point, and n represents the total number of time points. The total path length of the tool during the entire motion process is obtained by accumulating the position difference between two adjacent time points;

[0141] A convolutional neural network is used to perform feature recognition on the tool motion image sequence collected after the rack position is adjusted, the adjusted tool motion trajectory is obtained from the trajectory processor, and the current value of the adjusted handover smoothness is calculated in the trajectory processor;

[0142] According to the handover fluency benchmark value and the current value, a time series comparison is performed on the tool motion trajectory data, a handover fluency improvement rate is calculated from the trajectory change data, and a rack position adjustment effect is determined in the trajectory processor;

[0143] The industrial camera array consists of 8 high-speed cameras, which capture tool motion images at a frequency of 120 frames per second. Each camera covers an area of ​​40 square meters, and there is a 20% overlap area between adjacent cameras for image stitching;

[0144] The image data collector pre-processes the raw image, including noise elimination and distortion correction, and the positioning accuracy of the tool position coordinates reaches 0.5 cm;

[0145] The target tracking algorithm generates continuous tool motion trajectory data by identifying the characteristic marks on the tool surface, with a sampling interval of 25 milliseconds;

[0146] In the example of the machining area, the trajectory processor recorded the tool motion data before the rack adjustment, showing that the round-trip path of the tool between the workbench and the tool rack showed obvious arc characteristics, with an average path length of 4.8 meters;

[0147] There are obvious fluctuations in the moving speed of the tool, which decreases to 0.3 m / s when approaching the tool holder, and the handover time interval is 12 seconds on average;

[0148] These data form the handover fluency benchmark value, with a quantitative index of 0.65. The convolutional neural network uses a two-stream structure to process the tool motion image sequence. The spatial stream is responsible for identifying the instantaneous position of the tool, and the temporal stream tracks the tool's motion trend;

[0149] In the application case of the final assembly workshop, the tool movement trajectory after the rack adjustment showed more regular characteristics, the path length was shortened to 3.6 meters, the moving speed remained at a stable level of about 0.5 meters per second, and the handover time interval was reduced to 8 seconds;

[0150] The trajectory processor calculates the current value of the adjusted handover smoothness based on these parameters to be 0.82. The timing comparison analysis shows that the optimization effect of the tool motion trajectory is most significant in the quality inspection station area;

[0151] Before the adjustment, the paths of tools passing between different tool racks overlapped, with an average of 2.5 changes of direction per pass, and a fluctuation range of 6 seconds in the pass time.

[0152] After the rack position was adjusted, the tool motion trajectory was simplified to a straight line, the number of direction changes was reduced to 1.2 times, and the standard deviation of the handover time was reduced to 1.5 seconds;

[0153] The trajectory processor calculated that the handover fluency in this area has increased by 26%. In the transfer area at the end of the production line, due to the complex process connection, the efficiency of tool handover is required to be high;

[0154] Trajectory analysis data showed that the tool motion trajectory before adjustment showed obvious circuitous characteristics, with lateral displacement accounting for 45% of the total path length, and frequent pauses and accelerations during the handover process;

[0155] The rack position adjustment adopts a fan-shaped layout scheme, the tool movement trajectory is transformed into a radial shape, the lateral displacement ratio is reduced to 15%, and the movement speed curve tends to be smooth;

[0156] The calculation result of the handover smoothness improvement rate is 31%, of which the shortened handover time contributed 18 percentage points and the route optimization contributed 13 percentage points;

[0157] By comparing the evaluation data of different areas, it was found that the improvement of tool handover fluency is highly correlated with the rationality of the tool rack layout. In different workstations such as machining, assembly, and quality inspection, the optimized tool motion trajectories all show geometric characteristics that are adapted to the characteristics of the operation. The handover fluency improvement rate generally exceeds 20%, reflecting the adaptability and effectiveness of the rack adjustment plan.

[0158] like Figure 1-Figure 3 As shown, in step S106, a reinforcement learning algorithm is used to improve the handover fluency as the optimization goal, and the tool rack structure model is used as input. By continuously trying and adjusting the tool rack structure, the optimal rack layout solution is obtained, and it is applied to the subsequent rack adjustment process;

[0159] Reading tool rack structure model data, obtaining a handover fluency index from the tool rack structure model data, and determining a reward function value for rack layout optimization according to the handover fluency index;

[0160] Initializing tool rack structure parameters according to the reward function value, optimizing and iterating rack layout schemes using the tool rack structure parameters, and obtaining rack adjustment data from the optimized and iterative calculations;

[0161] Generate an iterative optimization layout sequence through the rack position adjustment data, perform spatial distribution calculation on the iterative optimization layout sequence, and obtain a layout solution convergence index from the spatial distribution calculation;

[0162] If the convergence index of the layout scheme meets the preset convergence threshold, multiple rounds of iterative operations are performed on the tool rack structure according to the layout optimization parameters, optimal rack parameters are obtained from the multiple rounds of iterative operations, and the tool rack structure model is updated using the optimal rack parameters.

[0163] Specifically, in step S106, the tool rack structure model data is read through the deep reinforcement learning algorithm, the reward function value of the rack layout optimization is obtained from the handover fluency index, the rack space mapping sequence is established in the optimization data processor, and the tool rack structure parameters are initialized;

[0164] According to the tool rack structure parameters and the handover fluency index, the tool rack layout scheme is optimized and iteratively calculated, rack adjustment data is obtained from the change of the reward function value, and an iterative optimization layout sequence is generated in the optimization data processor;

[0165] The spatial distribution calculation of the iterative optimization layout sequence is performed through the optimization data processor, the layout scheme convergence index is extracted from the rack adjustment data, and the layout optimization parameters are recorded in the optimization data processor;

[0166] According to the layout optimization parameters and the layout scheme convergence index, multiple rounds of iterative calculations are performed on the tool rack structure, the optimal rack parameters are obtained from the optimized layout sequence, and the tool rack structure model is updated in the optimization data processor;

[0167] The deep reinforcement learning algorithm optimizes the layout plan by constructing the state space of the tool rack. The tool rack structure model includes multi-dimensional features such as position coordinates, spacing relationship, and handover frequency.

[0168] In the application case of an assembly workshop, the state space consists of the position parameters of 16 tool racks, each of which has two degrees of freedom, horizontal and vertical. The handover smoothness index is used as the core metric of the reward function, with an initial value of 0.65;

[0169] The optimized data processor discretizes the rack space and maps the continuous position parameters to a 20×20 grid to form 400 candidate position points;

[0170] During the optimization iteration process, the algorithm explores different layout solutions based on the Monte Carlo tree search strategy;

[0171] In the example of the machining area, each round of iteration involves 32 layout changes, and the adjustment range of the rack position is between 0.2 meters and 1.5 meters. The reward function calculates the score based on the adjusted handover fluency. When the fluency improves by more than 5%, a positive reward is given, and when it decreases by more than 3%, a negative penalty is imposed. After 150 rounds of iterations, the algorithm generates 8 different layout optimization sequences, each containing 48 layout states. The convergence of the layout scheme is evaluated through multiple indicators, including the changing trend of handover fluency, the uniformity of rack position distribution, and space utilization. In the quality inspection station area, the 8 tool racks are distributed in a matrix in the initial layout, and the space utilization rate is 62%.

[0172] The optimization data processor recorded the key parameters in the layout evolution process: the minimum spacing between adjacent tool racks was adjusted from 1.2 meters to 0.8 meters, the maximum spacing was reduced from 2.5 meters to 1.6 meters, and the average length of the handover path was reduced by 28%; the convergence index of the layout scheme reached a stable value of 0.92 at the 180th iteration; multiple rounds of iterative calculations showed that in the transfer area at the end of the production line, the optimal layout scheme showed obvious functional zoning characteristics; 12 tool racks were divided into 3 groups, each group of 4 tool racks adopted a fan-shaped layout, and a safety distance of 2 meters was maintained between groups; by comparing the performance of different iteration sequences, the optimization data processor found that the layout scheme always maintained the optimal state after 200 iterations, with the handover fluency reaching 0.88 and the space utilization rate increased to 78%;

[0173] In the complex environment of the final assembly workshop, the tool rack structure model needs to adapt to the changing operational requirements. The optimization algorithm explores the optimal layout in a 4×5 grid space. By dynamically adjusting the learning rate and exploration range, the optimization process maintains stability while having sufficient adaptability.

[0174] Experimental data show that when the algorithm processes the layout optimization of 20 tool racks, the convergence time is controlled within 300 seconds. The final layout solution improves the handover fluency from the initial 0.58 to 0.85, and remains stable during the 7-day verification period.

[0175] The update frequency of the tool rack structure model is set to once every 8 hours to balance the optimization effect and computing resource consumption.

[0176] like Figure 1-Figure 3 As shown, in step S107, the tools are classified and coded according to their multifunctional attributes, and when the tool rack is adjusted in linkage, tools with complementary functions are preferentially placed in adjacent positions;

[0177] The tool function attribute data is collected through the tool attribute identifier, the function combination information is obtained from the tool use record, and the tool feature data is obtained in the function data processor;

[0178] According to the tool feature data, a clustering algorithm is used to classify the tool function features, and a tool function coding table is generated in a function data processor;

[0179] According to the tool function coding table and the tool use frequency data, the tool function combination relationship is calculated to generate function combination data;

[0180] After receiving the function combination data, a correlation analysis is performed through a graph neural network, complementary function combinations are identified from tool combination relationships, tool proximity data is generated in a function data processor, and the tool rack layout is updated according to the tool proximity data.

[0181] Specifically, in step S107, tool function attribute data is collected through a tool attribute identifier, function combination information is extracted from tool usage records, tool function features are classified and calculated using a clustering algorithm, and a tool function coding table is generated in a function data processor;

[0182] According to the tool function coding table and the tool use frequency data, the complementarity of the tool function combination relationship is calculated, the function association degree is obtained from the multifunctional attribute data, and the function combination data is generated in the function data processor;

[0183]

[0184] Compl(A, B) represents the complementarity between tool A and tool B, K represents the total number of tool functions, fA(k) represents the frequency of use of tool A on function k, and fB(k) represents the frequency of use of tool B on function k;

[0185] This formula measures the degree of functional complementarity between two tools by calculating the ratio of the sum of the minimum usage frequency of the two tools on each function to the sum of the maximum usage frequency;

[0186] Through the graph neural network, the functional combination data is analyzed for correlation, the complementary functional combination is identified from the tool combination relationship data, and the tool proximity data is obtained in the functional data processor;

[0187] According to the tool proximity data and the tool rack position data, the tool rack linkage layout is optimized and sorted, the layout optimization parameters are extracted from the proximity calculation results, and the tool rack position layout is updated in the functional data processor;

[0188] The tool attribute identifier establishes a multi-dimensional functional attribute mapping by analyzing the physical characteristics and usage scenarios of the tool;

[0189] In an assembly workshop, the functional attributes of 48 tools were collected, including tool type, application scope, operation mode and other characteristics. The clustering algorithm divided these tools into 6 main functional categories, each category has 3 subcategories, forming a hierarchical functional coding system;

[0190] The functional data processor generates an 8-digit code for each tool, the first 2 digits represent the main functional category, the middle 3 digits represent the sub-functional category, and the last 3 digits are the tool serial number;

[0191] The tool function combination relationship was quantified by analyzing the frequency of joint use of tools. In the practice of the machining area, the tool use data was recorded for 30 consecutive days. It was found that the joint use frequency of screwdriver tools and wrench tools was the highest, accounting for 35% of the total use times;

[0192] The degree of functional association is expressed as a normalized value from 0 to 1. When the frequency of joint use of two types of tools exceeds 30%, the degree of association is greater than 0.8;

[0193] The functional data processor generates a 16×16 functional combination data matrix based on this, and the graph neural network identifies complementary functional combinations by constructing a tool functional association graph;

[0194] In the quality inspection workstation area, the network connects 24 tool nodes according to the functional association strength, and the edge weight is determined by the functional combination data;

[0195] The analysis results show that measurement tools and inspection tools form the strongest complementary relationship, with a correlation strength of 0.92. The network also discovered three typical functional complementary combinations, each containing 4 to 6 tools of different categories, with an average correlation strength of more than 0.75 between these tools.

[0196] Tool rack linkage layout optimization is spatially reorganized based on tool proximity data. In the application case of the general assembly workshop, 32 tool racks are rearranged according to the functional complementarity relationship to form 8 functional complementary groups, with 4 tool racks in each group distributed in a fan shape; the distance between adjacent groups is dynamically adjusted according to the degree of functional association, and the distance between groups with high association is shortened to 1.2 meters, while the distance between groups with low association is maintained at more than 2 meters;

[0197] The layout optimization parameters show that the new layout scheme reduces the average access distance between complementary tools by 45%. At the flexible workstation at the end of the production line, the tool function combination shows obvious timing characteristics. The functional data processor analyzes the time series of tool use and finds that there are unique functional complementarity requirements in different process stages. The assembly stage mainly uses fixed and adjustable tool combinations, while the testing stage mainly uses detection and calibration tool combinations. The layout optimization algorithm generates a double-layer annular tool rack layout based on this. The inner ring is used to place frequently used basic tools, and the outer ring is used to arrange special tools according to the process sequence. The radial spacing between the two rings is 0.8 meters, allowing the operator to easily switch between different functional groups.

[0198] In the description of the present application, it should be understood that the directions or positional relationships indicated by directional words such as "front, back, up, down, left, right", "lateral, vertical, horizontal" and "top, bottom" are usually based on the directions or positional relationships shown in the drawings, and are only for the convenience of describing the present application and simplifying the description. Unless otherwise specified, these directional words do not indicate or imply that the device or element referred to must have a specific direction or be constructed and operated in a specific direction, and therefore cannot be understood as limiting the scope of protection of the present application.

[0199] For those skilled in the art, various other corresponding changes and deformations can be made according to the technical solutions and concepts described above, and all of these changes and deformations should fall within the protection scope of the claims of this application.

Claims

1. A multifunctional active tool rack linkage method based on the Internet of Things, characterized in that: The following steps are involved: Step S101, by installing a radio frequency identification sensor on a tool rack in a target area, real-time data on tool handover frequency is acquired to identify the number of times a tool on the tool rack in the target area is used. If the number of times a tool is used within a preset time period is higher than a preset number of times threshold, the corresponding tool rack is determined to be a high-frequency handover area; Step S102, using a clustering algorithm to perform spatial division on the high-frequency handover area, determining the cluster center of the high-frequency handover area, determining the position coordinates of the sub-areas in the high-frequency handover area according to the cluster center, and obtaining the tool rack structure model according to the position coordinates of the sub-areas; Step S103, counting the number and direction of tool handovers in the sub-areas of the high-frequency handover area within a preset time period, generating a handover heat map, and determining an adjustment plan for adjacent tool racks using a fuzzy reasoning method based on the heat distribution in the map and in combination with the tool rack structure model; Step S104, establishing a communication connection between tool racks through the Internet of Things, and when a tool rack detects a high-frequency handover area, generating a corresponding rack position adjustment plan, transmitting the adjustment instruction to the adjacent tool rack, and controlling the adjacent tool rack to adjust; Step S105, during the tool rack structure adjustment process, using computer vision methods to track the position change of the tool in real time, and by comparing the tool trajectory changes before and after the adjustment, calculating the improvement rate of the handover smoothness; Step S106, using a reinforcement learning algorithm, taking the tool rack structure model as input, obtaining an optimal rack layout solution, and applying it to the rack adjustment process; Step S107, classifying and coding the tools, and placing tools with complementary functions in adjacent positions first when adjusting the tool rack in linkage.

2. According to the method of claim 1, a multifunctional movable tool rack linkage method based on the Internet of Things is characterized in that: In step S101, by installing a radio frequency identification sensor on a tool rack in a target area, tool handover frequency data is acquired in real time, and the number of times a tool on the tool rack in the target area is used is identified. If the number of times a tool is used within a preset time period is higher than a preset number of times threshold, the corresponding tool rack is determined to be a high-frequency handover area, including: Obtain the radio frequency signal strength collected by the regional sensor array at the tool handover time, and compare the tool rack parking duration and access duration according to the radio frequency signal strength to obtain tool handover frequency data; For the tool handover frequency data, a neural network algorithm is used to classify and aggregate the tool parking duration and access duration to obtain the regional tool turnover rate and regional tool density; If the capacity saturation of the regional tool rack reaches the saturation threshold, the regional handover intensity is evaluated based on the tool turnover rate and the tool density in the regional area, based on the tool parking duration and the tool access duration data; Based on the analysis of tool handover frequency data, the peak period of regional handover is obtained to determine the high-frequency handover tool rack area.

3. According to the method for linking a multifunctional movable tool rack based on the Internet of Things in claim 1, it is characterized in that: In step S102, a clustering algorithm is used to spatially divide the high-frequency handover area, a cluster center of the high-frequency handover area is determined, the position coordinates of the sub-areas in the high-frequency handover area are determined according to the cluster center, and a tool rack structure model is obtained according to the position coordinates of the sub-areas, including: Acquire a spatial position coordinate sequence of the tool rack, calculate the region handover frequency from the spatial position coordinate sequence, and obtain density distribution data of the tool rack in a spatial data processor; Using the k-means clustering algorithm to perform spatial clustering calculation based on the density distribution data of the tool rack, identifying the regional handover heat distribution characteristics from the density distribution data, and obtaining the cluster center point coordinates of the high-frequency handover area; Perform sub-region boundary division according to the cluster center point coordinates and the spatial cluster distance, calculate the sub-region area size and the region saturation degree from the sub-region boundary data, and generate sub-region structure data in the spatial data processor; The spatial relationship calculation is performed on the sub-region structure data by a spatial data processor, the spatial arrangement characteristics of the tool rack are obtained from the density distribution data of the tool rack, and the tool rack position structure model is obtained.

4. According to the method for linking a multifunctional movable tool rack based on the Internet of Things as claimed in claim 1, it is characterized in that: In step S103, the number and direction of tool handovers in the sub-areas of the high-frequency handover area within a preset time period are counted to generate a handover heat map. According to the heat distribution in the map and in combination with the tool rack structure model, a fuzzy reasoning method is used to determine an adjustment plan for adjacent tool racks, including: Receiving tool position signals collected by a regional sensor array, extracting tool movement direction and dwell time data from the tool position signals, and generating a tool handover frequency statistics table in a spatial data processor according to the movement direction and dwell time data; The tool handover frequency statistics table and tool movement trajectory data are used in combination with the position information in the tool rack structure model to calculate the heat value of the handover path data, and a handover heat distribution map is obtained in the spatial data processor; Executing a fuzzy inference algorithm according to the handover heat distribution map, performing a spatial correlation operation on the heat distribution data, identifying the handover hotspot area between tool racks, and obtaining a tool rack position spacing table in the spatial data processor; If the tool rack position spacing table and the handover heat distribution map show that the handover path length of adjacent tool racks exceeds the handover path length threshold, position parameters are obtained from the tool rack structure model, and a tool rack adjustment plan is generated in the spatial data processor.

5. According to the method of claim 1, a multifunctional movable tool rack linkage method based on the Internet of Things is characterized in that: In step S104, a communication connection is established between tool racks through the Internet of Things. When a tool rack detects a high-frequency handover area, a corresponding rack position adjustment plan is generated, and an adjustment instruction is transmitted to an adjacent tool rack to control the adjacent tool rack to adjust, including: Collecting tool rack position sensor data, wherein the tool rack position sensor data is used to generate a tool rack communication connection table; Calculate the signal transmission delay according to the network topology relationship in the tool rack communication connection table, and use the signal transmission delay to generate a rack communication instruction; Extracting adjacent tool rack position relationship data from the rack position communication instructions, and obtaining rack position adjustment instructions after the adjacent tool rack position relationship data is processed by a deep learning algorithm; The positions of adjacent tool racks are synchronously calculated according to the rack adjustment instructions, and the rack movement data obtained by the synchronous calculation is used for the coordinated adjustment of the racks.

6. The method for linking a multifunctional movable tool rack based on the Internet of Things according to claim 1, characterized in that: In step S105, during the tool rack structure adjustment process, the position change of the tool is tracked in real time using a computer vision method, and the improvement rate of the handover smoothness is calculated by comparing the tool trajectory changes before and after the adjustment, including: Obtain a tool motion image sequence acquired by the industrial camera array in the rack adjustment area, and obtain tool position coordinate data from the image data collector; Performing target tracking processing on the tool position coordinate data to obtain tool motion trajectory data, storing the tool motion trajectory data in a trajectory processor, and obtaining the tool motion trajectory before the rack position is adjusted; According to the tool motion trajectory and tool position coordinate data before the rack position adjustment, the moving speed and motion path length of the tool before the adjustment are calculated by the trajectory processor; Extract the tool handover time interval from the target tracking data, and generate the handover smoothness benchmark value before adjustment through the trajectory processor; Obtain the tool motion image sequence collected after the rack position is adjusted, and use convolutional neural network to perform feature recognition processing on it; Obtaining the adjusted tool motion trajectory from the trajectory processor, and calculating the adjusted current value of the handover smoothness through the trajectory processor; Performing time series comparison processing on tool motion trajectory data according to the handover smoothness benchmark value and the current value; The handover smoothness improvement rate is calculated from the trajectory change data, and the rack adjustment effect is determined by the trajectory processor.

7. The method for linking a multifunctional movable tool rack based on the Internet of Things according to claim 1, characterized in that: In step S106, a reinforcement learning algorithm is used to take the tool rack structure model as input to obtain the optimal rack layout solution, and the solution is applied to the rack adjustment process, including: Reading tool rack structure model data, obtaining a handover fluency index from the tool rack structure model data, and determining a reward function value for rack layout optimization according to the handover fluency index; Initializing tool rack structure parameters according to the reward function value, optimizing rack layout schemes by using the tool rack structure parameters, and obtaining rack adjustment data from the optimized iterative calculations; generating an iterative optimization layout sequence by using the rack adjustment data, performing spatial distribution calculations on the iterative optimization layout sequence, and obtaining a layout scheme convergence index from the spatial distribution calculations; If the convergence index of the layout scheme meets the preset convergence threshold, multiple rounds of iterative operations are performed on the tool rack structure according to the layout optimization parameters, optimal rack parameters are obtained from the multiple rounds of iterative operations, and the tool rack structure model is updated using the optimal rack parameters.

8. The method for linking a multifunctional movable tool rack based on the Internet of Things according to claim 1, characterized in that: In step S107, the tools are classified and coded, and when the tool rack is adjusted in linkage, tools with complementary functions are preferentially placed in adjacent positions, including: The tool function attribute data is collected through the tool attribute identifier, the function combination information is obtained from the tool use record, and the tool feature data is obtained in the function data processor; According to the tool feature data, a clustering algorithm is used to perform classification operations on tool function features, and a tool function coding table is generated in a function data processor.

9. The method for linking a multifunctional movable tool rack based on the Internet of Things according to claim 8, characterized in that: The tool function coding table and the tool use frequency data are used to calculate the complementarity of the tool function combination relationship to generate function combination data.

10. The method for linking a multifunctional movable tool rack based on the Internet of Things according to claim 9, characterized in that: After receiving the function combination data, a correlation analysis is performed through a graph neural network, complementary function combinations are identified from tool combination relationships, tool proximity data is generated in a function data processor, and the tool rack layout is updated according to the tool proximity data.

Images