Automobile production line safety monitoring system based on industrial internet

Abstract

The invention relates to the technical field of safety monitoring, and discloses an industrial internet-based automobile production line safety monitoring system, which is characterized in that a target frame is divided into N identification areas at fixed intervals from the head to the tail according to N groups of left and right symmetrical mechanical arms, and the nth group of mechanical arms is associated with the nth identification area; first labels are arranged on the nth group of mechanical arms and the nth identification area; arranging a second label on the Nth group of mechanical arms; obtaining feature data of each welding spot in the first identification area to the (N-1) th identification area, and dividing the corresponding welding spots into key welding spots and non-key welding spots based on the feature data; acquiring a feature image of the target frame at the corresponding matching moment based on the Nth group of mechanical arms, and sending the feature image to the kth group of mechanical arms for preferential identification; and in combination with a genetic algorithm, first welding is carried out on the key welding spots, second welding is carried out on the remaining welding spots, and a welding execution scheme is obtained.

Description

Technical Field

[0001] The present invention relates to the field of safety monitoring technology, and more specifically, to an automobile production line safety monitoring system based on industrial Internet. Background Art

[0002] In the current automatic welding production line, the frame welding is completed by multiple welding robotic arms deployed symmetrically on the left and right. To ensure the frame welding is firm, multiple robotic arms need to participate in the welding at the same time, which may cause the risk of mutual interference between robotic arms. To this end, each robotic arm is equipped with an obstacle avoidance system and an identification system. The former is used to detect and avoid collisions with other robotic arms, and the latter is responsible for identifying welding points on the frame. However, the existing technology has the following problems: 1. Safety issues: Each robot arm independently identifies the welding point, resulting in the welding time depending on the difficulty of identifying the welding point. Welding points that are easy to identify will be welded first, while welding points that are more difficult to identify (usually key welding points) will be delayed, resulting in poor welding stability of the frame.

[0003] 2. Efficiency issue: The welding time of key welding points is relatively long. In addition, the robot arm that has priority for welding needs to avoid obstacles and wait, which greatly prolongs the overall welding time. Summary of the invention

[0004] The present invention provides an automobile production line safety monitoring system based on the industrial Internet to solve the technical problems raised in the background technology.

[0005] The present invention provides an automobile production line safety monitoring system based on the industrial Internet, comprising: The area division module is used to divide the target vehicle frame into N identification areas from the front to the rear at fixed intervals according to N groups of bilaterally symmetrical mechanical arms, and associate the nth group of mechanical arms with the nth identification area; wherein 1≤n≤N; The tag matching module includes a first tag and a second tag; the first tag is arranged on the nth group of mechanical arms and the nth identification area; the second tag is arranged on the Nth group of mechanical arms; A solder joint classification module, used for obtaining characteristic data of each solder joint in the 1st to N-1th identification areas, and classifying the corresponding solder joints into key solder joints and non-key solder joints based on the characteristic data; A data monitoring module, for responding to the matching of the second label of the Nth group of mechanical arms with the first label of the kth identification area, obtaining a feature image of the target vehicle frame at the corresponding matching time based on the Nth group of mechanical arms, and sending the feature image to the kth group of mechanical arms for priority identification; wherein 1≤k≤N-1; The welding execution module is used to respond to the matching of the first tag of the nth group of robot arms with the first tag of the nth identification area, and in combination with the genetic algorithm, perform the first welding on the key welding points and perform the second welding on the remaining welding points to obtain a welding execution plan.

[0006] Furthermore, both the first tag and the second tag are RFID tags; wherein, when the first tag matches the first tag, the welding task of the corresponding robotic arm is executed; and when the second tag matches the first tag, a data link between the Nth robotic arm and the corresponding robotic arm is established.

[0007] Furthermore, the corresponding solder joints are divided into key solder joints and non-key solder joints, including: The characteristic data includes: the connection function L of the welding point and the safety function S of the welding point; a number of structural units are manually defined for the target frame, and the structural units include but are not limited to: front and rear longitudinal beams, side beams and cross beams; if the welding point is used to connect any number of structural units, the connection function of the welding point is the number of connected structural units; a multi-angle force analysis is performed on the welded target frame; if the stress value of the welding point at any stress angle is greater than the preset stress threshold, the safety function of the welding point is the difference between the stress value and the preset stress threshold; The critical score is calculated for the solder joints using the following formula: ; in, Indicates the critical score of the solder joint, represents the first weight, represents the second weight; like , the corresponding solder joint is classified as a critical solder joint, otherwise it is classified as a non-critical solder joint.

[0008] Furthermore, the feature image is sent to the kth group of robotic arms for priority recognition, including: The feature images acquired by the left and right robotic arms in the Nth group of robotic arms are sent to the left and right robotic arms in the kth group of robotic arms respectively; The left robotic arm and the right robotic arm in the kth group of robotic arms recognize the feature image based on a preset recognition model to obtain the positions of several welding points in the kth identification area at the moment when the first label of the kth group of robotic arms matches the first label of the kth identification area.

[0009] Further, performing a first welding on the key welding point includes: Obtain the key solder joints on the left and right sides of the kth recognition area, as well as the score of each key solder joint; Based on the scores of the key welds, the key welds on the left and right sides of the k-th identification area are sorted from large to small to obtain a first left welding sort and a first right welding sort; And performing the first welding on the key welding point in the kth identified area based on the first left welding sequence and the first right welding sequence.

[0010] Further, performing a second welding on the remaining welding points includes: Obtain non-critical welding points on the left and right sides of the nth identification area; Initialize and generate a second welding sequence in R n-th identification areas that meet the constraint conditions; wherein the second welding sequence includes: a second left welding sequence and a second right welding sequence; The constraints are as follows: each solder joint is soldered only once; The second welding is performed on the non-critical welding points in the nth identified area based on the R second welding rankings.

[0011] Furthermore, a welding execution plan is obtained, including: Step 71, based on the first welding and R second welding of the N sub-areas, respectively establish R groups of motion interference simulation, each group of motion interference simulation includes the first welding and the rth second welding; 1≤r≤R; Step 72, running the R group motion interference simulation based on the digital twin, including: Set the recognition time of the feature image corresponding to the Nth group of robot arms to U; At the matching time T1 of the first tag of the kth group of robot arms and the first tag of the kth identification area, the first left welding sequence and the first right welding sequence are synchronously executed in the kth identification area; after the first left welding sequence and the first right welding sequence are completed, the second left welding sequence and the second right welding sequence are continued to be executed; At the matching time T1 of the first tag of the Nth group of robot arms and the first tag of the kth identification area, after the identification time U, the second left welding sequence and the second right welding sequence are performed in the Nth identification area; Among them, the moving paths of the robot arm from the starting point to the welding point, from the welding point to the welding point, and from the welding point to the end point are all Euclidean distance paths, and the moving speed of each robot arm is V; The spatial constraints include: if there is a jth robot on the moving path of the ith robot, avoid it; the avoidance time is Tr; i≠j, 1≤i≤N, 1≤j≤N; Time constraints include: obtaining the welding time required for each welding point, and the welding time required for the robot arm to stay at any welding point; Step 73, based on the space constraint and the time constraint, respectively obtain the completion time T2 of the motion interference simulation of the R groups of motions, and the welding time corresponding to each group of motion interference simulation is T2-T1; Step 74, taking the welding time T2-T1 as the fitness value of the R second welding; Step 76, sorting the R second welds from small to large based on the fitness values ​​to obtain a fitness sorting; retaining a preset number of second welds from the front to the back of the fitness sorting, and performing random crossover and mutation on the second weld sortings corresponding to the remaining second welds to obtain updated R second welds; Step 77, repeating steps 72 to 76 for a preset number of times to obtain a second welding with a ranking of 1 in the fitness ranking; Step 78, perform the first welding and the second welding to obtain a welding execution plan.

[0012] The beneficial effect of the present invention is that by integrating technologies such as industrial Internet, RFID tags, digital twin simulation and genetic algorithms, the refined partitioning and intelligent scheduling of the entire frame welding process is realized, effectively solving the problems of untimely processing of key welding points and chaotic welding sequence caused by independent operation of robotic arms in traditional production, thereby significantly improving production efficiency and ensuring the welding quality of the frame. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 It is a module diagram of an automobile production line safety monitoring system based on industrial Internet of the present invention; Figure 2 It is a schematic diagram of a welding line of the present invention. DETAILED DESCRIPTION

[0014] The subject matter described herein will now be discussed with reference to example embodiments. It should be understood that the discussion of these embodiments is only to enable those skilled in the art to better understand and implement the subject matter described herein, and the functions and arrangements of the elements discussed may be changed without departing from the scope of protection of the contents of this specification. Each example may omit, replace or add various processes or components as needed. In addition, the features described relative to some examples may also be combined in other examples.

[0015] like Figure 1-2 As shown, an automobile production line safety monitoring system based on the industrial Internet includes: The area division module is used to divide the target vehicle frame into N identification areas from the front to the rear at fixed intervals according to N groups of bilaterally symmetrical mechanical arms, and associate the nth group of mechanical arms with the nth identification area; wherein 1≤n≤N; The tag matching module includes a first tag and a second tag; the first tag is arranged on the nth group of mechanical arms and the nth identification area; the second tag is arranged on the Nth group of mechanical arms; A solder joint classification module, used for obtaining characteristic data of each solder joint in the 1st to N-1th identification areas, and classifying the corresponding solder joints into key solder joints and non-key solder joints based on the characteristic data; A data monitoring module, for responding to the matching of the second label of the Nth group of mechanical arms with the first label of the kth identification area, obtaining a feature image of the target vehicle frame at the corresponding matching time based on the Nth group of mechanical arms, and sending the feature image to the kth group of mechanical arms for priority identification; wherein 1≤k≤N-1; The welding execution module is used to respond to the matching of the first tag of the nth group of robot arms with the first tag of the nth identification area, and in combination with the genetic algorithm, perform the first welding on the key welding points and perform the second welding on the remaining welding points to obtain a welding execution plan.

[0016] In one embodiment of the present invention, the first tag and the second tag are both RFID tags; wherein, when the first tag matches the first tag, the welding task of the corresponding robotic arm is executed; when the second tag matches the first tag, a data link between the Nth robotic arm and the corresponding robotic arm is established.

[0017] Specifically, RFID wireless radio frequency identification technology realizes the automatic matching of the robot arm and the identification area, ensures the intelligent execution of the welding task, and realizes data sharing between the robot arms. Each robot arm and each identification area are installed with a first tag. When the first tag on the robot arm enters its corresponding identification area (that is, the first tag matches), the robot arm is activated to perform the welding task. Robot arm A enters identification area 1, and its first tag recognizes the first tag in the area. The system confirms that robot arm A should perform the welding task and activates the robot arm. When the second tag matches the first tag in a certain identification area, the Nth group of robot arms obtains the welding point data in the area and transmits it to the front robot arm. The front robot arm can obtain complete welding information before starting welding, which improves production efficiency. Robot arm N enters identification area 3, and its second tag recognizes the first tag in the area. Robot arm N collects the frame feature image in the area and sends it to the front robot arms 1 and 2. A data sharing mechanism is formed between the robot arms to reduce task waiting time and improve overall production efficiency.

[0018] In one embodiment of the present invention, the corresponding solder joints are divided into key solder joints and non-key solder joints, including: The characteristic data includes: the connection function L of the welding point and the safety function S of the welding point; a number of structural units are manually defined for the target frame, and the structural units include but are not limited to: front and rear longitudinal beams, side beams and cross beams; if the welding point is used to connect any number of structural units, the connection function of the welding point is the number of connected structural units; a multi-angle force analysis is performed on the welded target frame; if the stress value of the welding point at any stress angle is greater than the preset stress threshold, the safety function of the welding point is the difference between the stress value and the preset stress threshold; The critical score is calculated for the solder joints using the following formula: ; in, Indicates the critical score of the solder joint, represents the first weight, represents the second weight; like , the corresponding solder joint is classified as a critical solder joint, otherwise it is classified as a non-critical solder joint.

[0019] Specifically, by calculating the critical scores of welds, the welds are divided into: critical welds (must be welded first) and non-critical welds (can be welded later). The goal of this method is to determine which welds are critical to the structural strength of the frame, ensure priority welding, and improve the stability of the frame. Optimize the welding sequence to avoid low welding efficiency and insufficient weld strength due to random welding. The importance of welds is evaluated by two key parameters: Connection function (L): measures whether the weld is used to connect multiple key structural units (such as longitudinal beams and cross beams). Safety function (S): measures the stress of the weld and selects welds that are subjected to greater stress in multiple directions. By giving different weight parameters, the critical score of each weld is obtained.

[0020] In one embodiment of the present invention, sending the feature image to the kth group of mechanical arms for priority recognition includes: The feature images acquired by the left and right robotic arms in the Nth group of robotic arms are sent to the left and right robotic arms in the kth group of robotic arms respectively; The left robotic arm and the right robotic arm in the kth group of robotic arms recognize the feature image based on a preset recognition model to obtain the positions of several welding points in the kth identification area at the moment when the first label of the kth group of robotic arms matches the first label of the kth identification area.

[0021] This method realizes information sharing between robotic arms through the process of image acquisition → transmission → recognition → task execution, so as to optimize the welding process through pre-recognition. Specifically, the preset recognition models are all existing technologies. They can be multi-modal open source models, such as Chatgpt or Tongyi Qianwen.

[0022] The comparison between the present invention and the prior art is as follows: Traditional welding method: the robotic arm scans and identifies each welding piece.

[0023] The welding spot of the present invention is preferentially identified: it is pre-identified by the rear-end robot arm.

[0024] Traditional welding method: independent recognition by the robotic arm.

[0025] The invention discloses a welding spot priority recognition method: an image sharing mechanism.

[0026] Traditional welding method: scan first and then weld.

[0027] The present invention prioritizes the identification of welding points: data is acquired first, and welding is performed directly.

[0028] Traditional welding method: low (the robot arm scans one by one and takes time).

[0029] The welding spot priority recognition of the present invention is: high (reducing scanning time and improving welding efficiency).

[0030] In one embodiment of the present invention, performing a first welding on a critical welding point includes: Obtain the key solder joints on the left and right sides of the nth recognition area, and the score of each key solder joint; Based on the scores of the key welds, the key welds on the left and right sides of the nth identification area are sorted from large to small to obtain a first left welding sort and a first right welding sort; And performing the first welding on the key welding points in the nth identified area based on the first left welding sequence and the first right welding sequence.

[0031] Specifically, ensure the priority welding of key welds to improve the structural strength and production efficiency of frame welding. Prioritize welding of key welds: weld the welds that are most critical to structural strength first to ensure frame stability. Welding point scoring and sorting: use a scoring mechanism to determine the priority of welds to ensure that the most important welds are welded first. Synchronous sorting on the left and right sides: sort the welds on the left and right sides independently, optimize the welding sequence, and improve the coordination efficiency of the robot arm.

[0032] In one embodiment of the present invention, performing a second welding on the remaining welding points includes: Obtain non-critical welding points on the left and right sides of the nth identification area; Initialize and generate a second welding sequence in R n-th identification areas that meet the constraint conditions; wherein the second welding sequence includes: a second left welding sequence and a second right welding sequence; The constraints are as follows: each solder joint is soldered only once; The second welding is performed on the non-critical welding points in the nth identified area based on the R second welding rankings.

[0033] Specifically, in the actual production process, the number of non-critical welds is far greater than that of critical welds. If welding is performed in the traditional order, the welding path will be lengthy and the welding time will be prolonged, thereby reducing production efficiency. Therefore, the welding sequence generation method of the genetic algorithm is used to optimize the welding path of non-critical welds, thereby improving welding efficiency. This method obtains class coding by sorting the welding of non-critical welds that meet the constraints, thereby supporting the subsequent optimization of the genetic algorithm.

[0034] In one embodiment of the present invention, obtaining a welding execution plan includes: Step 71, based on the first welding and R second welding of the N sub-areas, respectively establish R groups of motion interference simulation, each group of motion interference simulation includes the first welding and the rth second welding; 1≤r≤R; Step 72, running the R group motion interference simulation based on the digital twin, including: Set the recognition time of the feature image corresponding to the Nth group of robot arms to U; At the matching time T1 of the first tag of the kth group of robot arms and the first tag of the kth identification area, the first left welding sequence and the first right welding sequence are synchronously executed in the kth identification area; after the first left welding sequence and the first right welding sequence are completed, the second left welding sequence and the second right welding sequence are continued to be executed; At the matching time T1 of the first tag of the Nth group of robot arms and the first tag of the kth identification area, after the identification time U, the second left welding sequence and the second right welding sequence are performed in the Nth identification area; Among them, the moving paths of the robot arm from the starting point to the welding point, from the welding point to the welding point, and from the welding point to the end point are all Euclidean distance paths, and the moving speed of each robot arm is V; The spatial constraints include: if there is a jth robot on the moving path of the ith robot, avoid it; the avoidance time is Tr; i≠j, 1≤i≤N, 1≤j≤N; Time constraints include: obtaining the welding time required for each welding point, and the welding time required for the robot arm to stay at any welding point; Step 73, based on the space constraint and the time constraint, respectively obtain the completion time T2 of the motion interference simulation of the R groups of motions, and the welding time corresponding to each group of motion interference simulation is T2-T1; Step 74, taking the welding time T2-T1 as the fitness value of the R second welding; Step 76, sorting the R second welds from small to large based on the fitness values ​​to obtain a fitness sorting; retaining a preset number of second welds from the front to the back of the fitness sorting, and performing random crossover and mutation on the second weld sortings corresponding to the remaining second welds to obtain updated R second welds; Step 77, repeating steps 72 to 76 for a preset number of times to obtain a second welding with a ranking of 1 in the fitness ranking; Step 78, perform the first welding and the second welding to obtain a welding execution plan.

[0035] Specifically, traditional welding methods usually rely on manual setting of the welding sequence, but this method easily leads to low welding efficiency and a lengthy robot arm operation path. The present invention uses digital twin simulation + genetic algorithm optimization to generate and optimize the welding execution plan, making the arrangement of welding tasks more intelligent and precise.

[0036] The implementation process of the present invention can be divided into the following steps: Establish R-group motion interference simulation: Before performing welding, the system first establishes R-group motion interference simulation. This simulation is based on the division of the frame welding area (N sub-areas), and simulates: the first welding of key welds (determined parts) and the second welding of non-key welds (parts to be optimized by genetic algorithms). Before actual welding, the total welding time under R-group motion interference simulation is evaluated to ensure that the optimized welding scheme can minimize the interference between the robot arms and improve welding efficiency.

[0037] Set the recognition time U: Since the Nth group of robotic arms still needs to recognize the frame welding points before performing welding (the 1st to N-1 groups of robotic arms recognize in advance), the time U from recognition to welding of the robotic arms is set in the simulation.

[0038] Synchronous execution of welding tasks: The robot arm performs the first welding (key welding point) at the matching time T1 of the 1st to N-1st recognition areas. After the first welding is completed, the second welding sequence (non-key welding point) is continued. However, for the Nth recognition area, it is necessary to wait for the recognition time U. Since there is no key welding point set in the Nth recognition area, the second welding is directly performed in the Nth recognition area.

[0039] Optimize welding path: When performing welding, the robot arm uses the Euclidean distance path (i.e. point-to-point straight line arrival) to ensure that the robot arm's movement path during welding is the shortest and avoid invalid movement Optimize welding time based on spatial and temporal constraints.

[0040] Spatial constraints: Set the moving paths of different robotic arms to ensure that when the jth robotic arm exists on the path of the i-th robotic arm, it can actively stop and wait for the jth robotic arm to leave the path of the i-th robotic arm before the i-th robotic arm continues to move. The avoidance time is set to Tr. Since each robotic arm has a corresponding identification area (welding area), the avoidance of the robotic arm only exists when the i-th robotic arm and the j-th robotic arm are both in the path, not during the welding process. Therefore, when the moving speed is the same, all V, the avoidance time is also the same, all Tr.

[0041] Time constraint: Set the time required for welding each welding point (the welding time is based on historical data. Since welding is performed by a robot arm, the welding time for each welding point is basically the same), and ensure that the robot arm stays at each welding point for the corresponding welding time. In this way, the real welding process is simulated in the motion interference simulation to obtain the real welding time.

[0042] To optimize the welding sequence, the system uses a genetic algorithm for iterative optimization: the welding time T2-T1 is used as the evaluation index to calculate the fitness values ​​of different welding sequences. The shorter the welding time, the higher the fitness value. The second welding scheme of group R is sorted by fitness, and the top N optimal schemes are selected. The remaining schemes are randomly cross-crossed and mutated to generate new welding schemes for the next round of optimization. After multiple iterations, the optimal welding scheme is obtained: after multiple rounds of optimization, the system finally obtains the optimal welding sequence, that is, the welding scheme ranked first in fitness. After completing the simulation and optimization, the system sends the optimal welding scheme to the production line to perform the final welding task. Since the welding sequence has been optimized, the welding time is shortest, the robot arm interference is minimal, and the welding quality is highest.

[0043] The above describes an embodiment of the present embodiment, but the present embodiment is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of the present embodiment, ordinary technicians in this field can also make many forms, all of which are within the protection of the present embodiment.

Claims

1. An automobile production line safety monitoring system based on industrial Internet, characterized in that: include: The area division module is used to divide the target vehicle frame into N identification areas from the front to the rear at fixed intervals according to N groups of bilaterally symmetrical mechanical arms, and associate the nth group of mechanical arms with the nth identification area; wherein 1≤n≤N; The tag matching module includes a first tag and a second tag; the first tag is arranged on the nth group of mechanical arms and the nth identification area; the second tag is arranged on the Nth group of mechanical arms; A solder joint classification module, used for obtaining characteristic data of each solder joint in the 1st to N-1th identification areas, and classifying the corresponding solder joints into key solder joints and non-key solder joints based on the characteristic data; A data monitoring module, for responding to the matching of the second label of the Nth group of mechanical arms with the first label of the kth identification area, obtaining a feature image of the target vehicle frame at the corresponding matching time based on the Nth group of mechanical arms, and sending the feature image to the kth group of mechanical arms for priority identification; wherein 1≤k≤N-1; The welding execution module is used to respond to the matching of the first tag of the nth group of robot arms with the first tag of the nth identification area, and in combination with the genetic algorithm, perform the first welding on the key welding points and perform the second welding on the remaining welding points to obtain a welding execution plan.

2. According to the industrial Internet-based automobile production line safety monitoring system of claim 1, it is characterized in that: The first tag and the second tag are both RFID tags; when the first tag matches the first tag, the welding task of the corresponding robotic arm is executed; when the second tag matches the first tag, a data link between the Nth robotic arm and the corresponding robotic arm is established.

3. According to the industrial Internet-based automobile production line safety monitoring system of claim 2, it is characterized in that: The corresponding solder joints are divided into critical solder joints and non-critical solder joints, including: The characteristic data includes: the connection function L of the welding point and the safety function S of the welding point; a number of structural units are manually defined for the target frame, and the structural units include but are not limited to: front and rear longitudinal beams, side beams and cross beams; if the welding point is used to connect any number of structural units, the connection function of the welding point is the number of connected structural units; a multi-angle force analysis is performed on the welded target frame; if the stress value of the welding point at any stress angle is greater than the preset stress threshold, the safety function of the welding point is the difference between the stress value and the preset stress threshold; The critical score is calculated for the solder joints using the following formula: ； in, Indicates the critical score of the solder joint, represents the first weight, represents the second weight; like , the corresponding solder joint is classified as a critical solder joint, otherwise it is classified as a non-critical solder joint.

4. According to the industrial Internet-based automobile production line safety monitoring system of claim 3, it is characterized in that: Send the feature image to the kth group of robotic arms for priority recognition, including: The feature images acquired by the left and right robotic arms in the Nth group of robotic arms are sent to the left and right robotic arms in the kth group of robotic arms respectively; The left robotic arm and the right robotic arm in the kth group of robotic arms recognize the feature image based on a preset recognition model to obtain the positions of several welding points in the kth identification area at the moment when the first label of the kth group of robotic arms matches the first label of the kth identification area.

5. According to the industrial Internet-based automobile production line safety monitoring system of claim 4, it is characterized in that: Perform the first weld on the critical weld points, including: Obtain the key solder joints on the left and right sides of the nth recognition area, and the score of each key solder joint; Based on the scores of the key welds, the key welds on the left and right sides of the nth identification area are sorted from large to small to obtain a first left welding sort and a first right welding sort; And performing the first welding on the key welding points in the nth identified area based on the first left welding sequence and the first right welding sequence.

6. The automotive production line safety monitoring system based on industrial Internet according to claim 5 is characterized in that: Perform a second weld on the remaining weld points, including: Obtain non-critical welding points on the left and right sides of the nth identification area; Initialize and generate a second welding sequence in R n-th identification areas that meet the constraint conditions; wherein the second welding sequence includes: a second left welding sequence and a second right welding sequence; The constraints are as follows: each solder joint is soldered only once; The second welding is performed on the non-critical welding points in the nth identified area based on the R second welding rankings.

7. The automotive production line safety monitoring system based on industrial Internet according to claim 6 is characterized in that: Get a welding execution plan, including: Step 71, based on the first welding and R second welding of the N sub-areas, respectively establish R groups of motion interference simulation, each group of motion interference simulation includes the first welding and the rth second welding; 1≤r≤R; Step 72, running the R group motion interference simulation based on the digital twin, including: Set the recognition time of the feature image corresponding to the Nth group of robot arms to U; At the matching time T1 of the first tag of the kth group of robot arms and the first tag of the kth identification area, the first left welding sequence and the first right welding sequence are synchronously executed in the kth identification area; after the first left welding sequence and the first right welding sequence are completed, the second left welding sequence and the second right welding sequence are continued to be executed; At the matching time T1 of the first tag of the Nth group of robot arms and the first tag of the kth identification area, after the identification time U, the second left welding sequence and the second right welding sequence are performed in the Nth identification area; Among them, the moving paths of the robot arm from the starting point to the welding point, from the welding point to the welding point, and from the welding point to the end point are all Euclidean distance paths, and the moving speed of each robot arm is V; The spatial constraints include: if there is a jth robot on the moving path of the ith robot, avoid it; the avoidance time is Tr; i≠j, 1≤i≤N, 1≤j≤N; Time constraints include: obtaining the welding time required for each welding point, and the welding time required for the robot arm to stay at any welding point; Step 73, based on the space constraint and the time constraint, respectively obtain the completion time T2 of the motion interference simulation of the R groups of motions, and the welding time corresponding to each group of motion interference simulation is T2-T1; Step 74, taking the welding time T2-T1 as the fitness value of the R second welding; Step 76, sorting the R second welds from small to large based on the fitness values ​​to obtain a fitness sorting; retaining a preset number of second welds from the front to the back of the fitness sorting, and performing random crossover and mutation on the second weld sortings corresponding to the remaining second welds to obtain updated R second welds; Step 77, repeating steps 72 to 76 for a preset number of times to obtain a second welding with a ranking of 1 in the fitness ranking; Step 78, perform the first welding and the second welding to obtain a welding execution plan.

Images