Labor organization optimization setting method based on deep learning

Abstract

The invention provides a labor organization optimization setting method based on deep learning, and the method comprises the steps: carrying out the unified input of information according to the names, ages and skilled items of employees, formulating an exclusive information nameplate, distributing the information nameplate to each employee, and carrying out the employee classification model training and prediction through employing a Resnet algorithm as a classification deep learning algorithm; the method comprises the steps that information data are imported into a VGG classification network and a ResNet classification network respectively for classification model training, and the deep learning classification network is optimized through the change result of classification accuracy and classification loss degree along with the number of times of training in combination with the accuracy of a training set and a test set. After the information on the individual exclusive information nameplate is identified, the information nameplate is moved to the station column of the optimal optimization scheme by the system, the employees can know the work content of the employees and team personnel on the same day according to the displayed station column, and the scheme of labor organization optimization is embodied more specifically and is not influenced by tasks and the number of personnel every day.

Description

Technical Field

[0001] The present invention relates to the technical field of labor organization optimization, and specifically to a labor organization optimization setting method based on deep learning. Background Art

[0002] Labor organization is a way to establish an effective labor production system based on the needs of the enterprise and the principles of division of labor and collaboration, by properly managing the relationships between labor groups, between workers, and between workers and their tools and objects. Optimizing labor mix, simply put, means placing the right people in the right positions, enabling everyone to work together efficiently. A reasonable labor mix can avoid duplication and confusion in work. This optimization allows for a clear division of labor, with one person responsible for a portion while others handle areas that require their skills. This will significantly speed up project progress. When everyone is focused on their areas of expertise, the probability of error is greatly reduced. Optimizing labor mix can reduce unnecessary manpower input. More precise staffing naturally reduces costs. For employees, working in the right positions allows them to leverage their strengths, making it easier to achieve results, creating a greater sense of fulfillment, and boosting their work enthusiasm.

[0003] The existing technology for allocating and optimizing labor organization personnel is relatively general, because the daily workload, work content, and staff may be different, which will affect the effect of optimized allocation. It is not convenient to intuitively allocate and display the best labor organization optimization allocation plan for daily work. Summary of the Invention

[0004] In view of the shortcomings of the existing technology, the purpose of the present invention is to provide a labor organization optimization setting method based on deep learning to solve the problems raised in the above background technology. The present invention has a novel structure. After identifying the information on the personal information nameplate, the system moves the information nameplate to the workstation column of the best optimization plan. Employees can know their own work content and team members for the day based on the displayed workstation column, which makes the labor organization optimization plan more concrete and is not affected by daily tasks and the number of personnel.

[0005] In order to achieve the above-mentioned purpose, the present invention is implemented through the following technical solution: a labor organization optimization setting method based on deep learning, the optimization setting method comprising the following steps:

[0006] (1) Enter information uniformly based on the employee's name, age, and expertise, and develop a unique nameplate for each employee;

[0007] (2) Use Resnet algorithm as the classification deep learning algorithm to train and predict employee classification models; deep learning classification network, import information data into VGG and ResNet classification networks respectively for classification model training, and select the deep learning classification network by combining the accuracy of the training set and the test set according to the results of the change of classification accuracy and classification loss with the number of training times. The optimal model is embedded in the sorting system for employee information to achieve non-destructive detection and intelligent grading of employee labor distribution;

[0008] (3) The established learning classification model is input into the recognition component system. Employees who come to work every day put their nameplates into the recognition component. The recognition component uses the employee's age and expertise to make the best allocation plan using the classification model.

[0009] (4) The result obtained by the identification component is fed back to the control system of the upgrade component. The lifting component moves the corresponding workstation column to a position parallel to the conveyor table and stores the information nameplate sent by the identification component into the corresponding workstation column. Employees can see the work items of the day and the cooperating employees in the workstation column;

[0010] (5) According to the daily workload and the number of employees on duty that day, reasonable workstation columns and lifting components are selected for installation, and after identification and optimization classification by the identification component, they are finally displayed on the workstation column.

[0011] Furthermore, according to the learning classification model storage described in step (2), the model is persistently stored on the local hard disk on the training host, and then the model file is imported into the control center host. The persistent storage uses Python's pickle operation to serialize the preferred model, and the pickle operation serializes the model trained by deep learning, and saves this serialized format to a file, which is then deserialized by the sugar content grading module of the control center to achieve online classification prediction.

[0012] Furthermore, according to the identification information described in step (3), the identification component includes a conveyor table, and the information nameplates with the personnel information are placed on the conveyor table in sequence. They first enter the identification station under the belt conveyor of the conveyor table, and the name, age, and expertise on the information nameplate are identified by the identification lens in the identification chassis.

[0013] Furthermore, according to the lifting assembly described in step (4), the lifting assembly includes two first winding seats and two second winding seats, transmission belts are installed on the outer sides of the two first winding seats and the two second winding seats, and the pulleys of the transmission belts are fixedly connected to their respective winding shafts, and driving motors are respectively fixed to the positions of the first winding seat and the second winding seat at the bottom of the top plate, and the driving motors are respectively fixed to the winding shafts of the first winding seat and the second winding seat, a vertical frame is fixed to the middle top of the limit seat, and two groups of first electric push rods are fixed to the top of the vertical frame, and the extended ends of the first electric push rods are fixedly connected to the top plate.

[0014] Furthermore, after the lifting component receives the feedback information of the identification component, the first reeling seat reels in the traction rope and the second reeling seat unwinds the traction rope. At this time, the traction rope will move toward one side of the first reeling seat, and the work station bar installed on the traction rope will also move synchronously. The first reeling seat unwinds the traction rope and the second reeling seat rewinds the traction rope. At this time, the traction rope will move toward one side of the second reeling seat, and the work station bar installed on the traction rope will also move synchronously. The two modes can realize the movement of the work station bar on the traction rope, and move the work station bars at different positions to the horizontal position of the identification component for receiving the information nameplate.

[0015] Furthermore, according to the receiving information nameplate described in step (4), a movable slot and a sliding frame are provided on the back of the workstation column, and the first electric push rod inside the sliding frame can pass through the movable slot and connect with the information nameplate on the conveying platform. The metal plate at the bottom of the information nameplate placed flat on the conveying platform is parallel to the information nameplate at this time and is stored in the storage slot. After the electromagnetic suction plate contacts the metal plate, power is turned on to complete the adsorption of the electromagnetic suction plate and the metal plate.

[0016] Furthermore, a second electric push rod is provided on the base for pushing the slide of the identification component to move. After the extended end of the second electric push rod passes through the moving slot and is connected to the information nameplate, the second electric push rod on the base pushes the slide to move outward, and the conveying platform gradually moves away from the work station fence.

[0017] Furthermore, when the identification component is pushed outward, a toothed plate is fixed to one side of the bottom of the slide, and the toothed plate moves together with the slide, passes through the storage groove, and meshes with the gear at the bottom of the second screw. The gear drives the second screw to rotate, and one end of the pressure plate cooperates with the second screw thread and the other end slides downward along the slide rod. After moving a certain distance, the pressure plate contacts the upper surface of the information nameplate. At this time, the bottom of the information nameplate has left the conveying platform, and the pressure plate continues to move downward to rotate the information nameplate downward along the rotating axis from horizontal to vertical, and then the second electric push rod is used to pull the information nameplate back into the placement slot, and the information nameplate is sent into the slot of the work station column by the drive of the first screw in the slide frame.

[0018] Furthermore, according to the installation of the workstation bar described in step (5), clamp frames are provided at both ends of the workstation bar, and information nameplates of different personnel are placed in different workstation bars in sequence according to the identification components. According to the number of different projects and the division of labor on that day, a corresponding number of workstation bars and traction ropes are selected for installation. First, the height of the front end of the traction rope, that is, the height of the top plate, needs to be adjusted so that all the workstation bars can be displayed from the front end. The newly added workstation bar is fixed to the traction rope by the clamp frames and bolts at both ends, and a certain distance is left between adjacent workstation bars to avoid movement interference.

[0019] Furthermore, when increasing and decreasing the number of workstation columns, when the driving motors of the first winding seat and the second winding seat both unwind the traction rope, the overall length of the traction rope released becomes longer. By pushing the top plate upward by the first electric push rod, the length of the front end of the traction rope can be increased to meet the needs of more workstation columns to be installed and displayed at the front end of the traction rope. Synchronously winding the traction rope can shorten the total length of the traction rope, and the first electric push rod drives the top plate to descend, reducing the number of workstation columns.

[0020] Beneficial effects of the present invention:

[0021] 1. In the present invention, when the driving motors of the first winding seat and the second winding seat both unwind the traction rope, the overall length of the unwound traction rope becomes longer. By pushing the top plate upward through the first electric push rod, the length of the front end of the traction rope can be increased to meet the needs of more workstations to be installed and displayed at the front end of the traction rope. Similarly, synchronously winding the traction rope can shorten the total length of the traction rope, and the first electric push rod drives the top plate to descend, reducing the number of workstations.

[0022] 2. The present invention moves the workstation bars at different positions to the horizontal position of the identification component for receiving information nameplates. This part is that after the identification component completes the information nameplate scanning and makes the best allocation plan through the model established by deep learning, the lifting component adjusts the corresponding workstation bar to receive the information nameplate. The limit seat used to limit the bottom of the traction rope can make the traction rope U-shaped, and the workstation bar can pass through the groove of the limit seat. The clamping frame of the workstation bar can move through the groove without interference, and the workstation bar can be moved from the front end to the rear end of the traction rope.

[0023] 3. The present invention places the information nameplates of different personnel in different workstation columns in sequence according to the identification components. According to the number of different projects and the division of labor on that day, the corresponding number of workstation columns and traction ropes are selected for installation. First, the height of the front end of the traction rope, that is, the height of the top plate, needs to be adjusted so that all the workstation columns can be displayed from the front end. The newly added workstation columns are fixed to the traction rope through the clamping frames and bolts at both ends. A certain distance is left between adjacent workstation columns to avoid movement interference.

[0024] 4. In the present invention, after the electromagnetic suction plate contacts the metal plate, power is turned on to complete the adsorption of the electromagnetic suction plate and the metal plate. The pressure plate of the identification component can squeeze the ordinary information nameplate and then rotate it along the rotating axis to become a vertical shape. At this time, the metal plate is perpendicular to the information nameplate, and the electromagnetic suction plate still maintains adsorption on the metal plate. Then, the information nameplate is pulled back to the placement slot by retracting the extended end of the first electric push rod. The vertical information nameplate just fits inside the placement slot. Then, the first screw is driven to rotate by the motor, and the screw plate cooperates with the first screw thread to slide along the slide frame. The extended end of the second electric push rod moves along the moving slot, and the information nameplate is sent into the slot of the work station column. The outer part of the slot is hollowed out, so that the name information on the information nameplate can be seen. There is also a limit strip to prevent the information nameplate from falling.

[0025] 5. The present invention pushes the slide to move outward through the second electric push rod on the base, and the conveying platform gradually moves away from the work station fence. During this process, the gear plate moves together with the slide, passes through the storage groove, and engages with the gear at the bottom of the second screw. The second screw is driven by the gear to rotate, and one end of the pressure plate cooperates with the second screw thread and the other end slides downward along the slide rod. After moving a certain distance, the pressure plate contacts the upper surface of the information nameplate. At this time, the bottom of the information nameplate has been separated from the conveying platform, and the pressure plate continues to move downward to rotate the information nameplate downward along the rotating axis from horizontal to vertical, making it convenient for the second electric push rod to pull the information nameplate back into the placement slot.

[0026] 6. Compared with the existing technology, the present invention identifies the information on the personal information nameplate and moves the information nameplate to the workstation column of the best optimization plan by the system. Employees can know their own work content and team members for the day based on the displayed workstation column, which makes the labor organization optimization plan more concrete and is not affected by daily tasks and the number of people. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 This is a process flow chart of the labor organization optimization setting method based on deep learning of the present invention;

[0028] Figure 2 This is a schematic diagram of the overall structure of the labor organization optimization setting method based on deep learning of the present invention;

[0029] Figure 3 A schematic diagram of the top structure of the lifting assembly of the labor organization optimization setting method based on deep learning of the present invention;

[0030] Figure 4 A schematic diagram of the connection between the workstation bar and the lifting component of the labor organization optimization setting method based on deep learning of the present invention;

[0031] Figure 5 A schematic diagram of the internal structure of the limit seat of the labor organization optimization setting method based on deep learning of the present invention;

[0032] Figure 6 A schematic diagram of the position relationship between the workstation bar and the sliding frame in the labor organization optimization setting method based on deep learning of the present invention;

[0033] Figure 7 This is a schematic diagram of the back structure of the workstation bar of the labor organization optimization setting method based on deep learning of the present invention;

[0034] Figure 8 This is a schematic diagram of the structure of the identification component of the labor organization optimization setting method based on deep learning of the present invention;

[0035] Figure 9 This is a schematic diagram of the pressure plate installation structure of the labor organization optimization setting method based on deep learning of the present invention;

[0036] Figure 10 A schematic diagram of the connection between the gear and the tooth plate of the labor organization optimization setting method based on deep learning of the present invention;

[0037] Figure 11 This is a schematic diagram of the connection between the information nameplate and the second electric push rod of the labor organization optimization setting method based on deep learning of the present invention.

[0038] In the figure: 1. Base; 11. Top plate; 2. Lifting assembly; 21. Limiting seat; 22. Vertical frame; 23. First electric push rod; 24. First reeling seat; 25. Second reeling seat; 26. Traction rope; 27. Transmission belt; 28. Through slot; 29. ​​Sliding frame; 210. First screw; 211. Screw plate; 212. Second electric push rod; 213. Electromagnetic suction plate; 214. Third electric push rod; 3. Station bar; 31. Insert Slot; 32. Placement slot; 33. Clamping frame; 34. Moving slot; 35. Bolt; 4. Identification component; 41. Slide rail; 42. Slide seat; 43. Conveyor platform; 44. Identification chassis; 45. Mounting frame; 46. Second screw; 47. Slide rod; 48. Pressure plate; 49. Gear; 410. Tooth plate; 411. Storage slot; 5. Information nameplate; 51. Rotation slot; 52. Rotation shaft; 53. Metal plate; 6. Drive motor. DETAILED DESCRIPTION

[0039] In order to make the technical means, creative features, objectives and effects achieved by the present invention easier to understand, the present invention is further described below in conjunction with specific implementation methods.

[0040] See also Figures 1 to 11 The present invention provides a technical solution: a labor organization optimization setting method based on deep learning, the optimization setting method comprising the following steps:

[0041] (1) Enter information uniformly based on the employee's name, age, and expertise, and develop a unique nameplate for each employee;

[0042] (2) Use Resnet algorithm as the classification deep learning algorithm to train and predict employee classification models; deep learning classification network, import information data into VGG and ResNet classification networks respectively for classification model training, and select the deep learning classification network by combining the accuracy of the training set and the test set according to the results of the change of classification accuracy and classification loss with the number of training times. The optimal model is embedded in the sorting system for employee information to achieve non-destructive detection and intelligent grading of employee labor distribution;

[0043] (3) The established learning classification model is input into the recognition component system. Employees who come to work every day put their nameplates into the recognition component. The recognition component uses the employee's age and expertise to make the best allocation plan using the classification model.

[0044] (4) The result obtained by the identification component is fed back to the control system of the upgrade component. The lifting component moves the corresponding workstation column to a position parallel to the conveyor table and stores the information nameplate sent by the identification component into the corresponding workstation column. Employees can see the work items of the day and the cooperating employees in the workstation column;

[0045] (5) According to the daily workload and the number of employees on duty that day, reasonable workstation columns and lifting components are selected for installation, and after identification and optimization classification by the identification component, they are finally displayed on the workstation column.

[0046] The lifting device 2 is a kind of lifting device of the present invention, and it is a kind of lifting and lowering device of the present invention. The identification component 4 includes a slide 42, a conveying platform 43 is fixed on the top of the slide 42, and an information nameplate 5 is placed on the surface of the conveying platform 43, and the conveying direction of the information nameplate 5 corresponds to the placement slot 32. An identification chassis 44 is fixed on the middle top of the conveying platform 43. When using the device, deep learning is used to enter the employee information and then establish a big data model. When arranging daily work, employees who are on duty that day need to send their personal information nameplate 5 to the identification component 4 for identification, and the identification component 4 is used to allocate them according to the employee's information, and the optimal labor organization is allocated to different workstation columns 3. Through the display of the workstation column 3, employees can know the work to be performed that day and the employees who are working with them. The method for establishing the deep learning model refers to the technical method in patent CN117862056A.

[0047] In this embodiment, the lifting assembly 2 also includes a transmission belt 27, and the outer sides of the two first winding seats 24 and the two second winding seats 25 are equipped with transmission belts 27, and the pulleys of the transmission belts 27 are fixedly connected to their respective winding shafts. The positions of the first winding seat 24 and the second winding seat 25 at the bottom of the top plate 11 are respectively fixed with drive motors 6, and the drive motors 6 are respectively fixedly connected to the winding shafts of the first winding seat 24 and the second winding seat 25. A vertical frame 22 is fixed to the middle top of the limit seat 21, and two groups of first electric push rods 23 are fixed to the top of the vertical frame 22. The extended ends of the first electric push rods 23 are fixedly connected to the top plate 11. The limit seat 21 A through slot 28 is provided on the inner surface of the top plate 11, and the traction rope 26 passes through the two ends of the through slot 28. The two first winding seats 24 at the bottom of the top plate 11 are synchronously driven by the drive motor 6 and the transmission belt 27. The two second winding seats 25 are also synchronously driven by the drive motor 6 and the transmission belt 27. The drive motor 6 of the first winding seat 24 and the second winding seat 25 can be controlled manually. When the drive motors 6 of the first winding seat 24 and the second winding seat 25 both unwind the traction rope 26, the overall length of the traction rope 26 released becomes longer. By pushing the top plate 11 upward by the first electric push rod 23, the length of the front end of the traction rope 26 can be increased to meet the installation requirements of more workstation columns 3. The traction rope 26 is shown at the front end. Similarly, the synchronous winding of the traction rope 26 can shorten the total length of the traction rope 26. The first electric push rod 23 drives the top plate 11 to descend, reducing the number of workstation fences 3. When the first winding seat 24 winds up the traction rope 26 and the second winding seat 25 unwinds the traction rope 26, the traction rope 26 will move toward one side of the first winding seat 24, and the workstation fence 3 installed on the traction rope 26 will also move synchronously. When the first winding seat 24 unwinds the traction rope 26 and the second winding seat 25 winds up the traction rope 26, the traction rope 26 will move toward one side of the second winding seat 25, and the workstation fence 3 installed on the traction rope 26 will also move synchronously. The mode can realize the movement of the work station column 3 on the traction rope 26, and move the work station column 3 at different positions to the horizontal position of the identification component 4 for receiving the information nameplate 5. This part is that after the identification component 4 completes the scanning of the information nameplate 5 and makes the best allocation plan through the model established by deep learning, the lifting component 2 adjusts the corresponding work station column 3 to receive the information nameplate 5. The limit seat 21 used to limit the bottom of the traction rope 26 can make the traction rope 26 U-shaped, and the work station column 3 can pass through the through groove 28 of the limit seat 21. The clamping frame 33 of the work station column 3 can move through the through groove 28 without interference, and the work station column 3 can be moved from the front end to the rear end of the traction rope 26.

[0048] In this embodiment, a movable groove 34 is provided on the back of the workstation fence 3, and a sliding frame 29 is fixed to the back of the workstation fence 3 at the bottom of the front end of the vertical frame 22 corresponding to the traction rope 26, and a first screw 210 is rotatably installed inside the sliding frame 29 through a bearing, and the surface of the first screw 210 is threadedly sleeved on the screw plate 211, and two second electric push rods 212 are fixed on the outer surface of the screw plate 211, and the extended end of the second electric push rod 212 is passed through the movable groove 34 and fixed with an electromagnetic suction plate 213, a rotating groove 51 is provided at the bottom of one end of the information nameplate 5, and a rotating shaft 52 is rotatably installed inside the rotating groove 51, and a metal plate 53 is fixed on the outer surface of the rotating shaft 52, and the electromagnetic suction plate 2 13 and the metal plate 53 are magnetically adsorbed on the outside of the station fence 3. Two semicircular clamping frames 33 are installed at both ends of the station fence 3 through a rotating shaft. The clamping frames 33 are clamped on the traction rope 26. The outer end of the clamping frame 33 is threadedly inserted with a bolt 35. The structure of each station fence 3 is the same. The position of the sliding frame 29 always corresponds to the station fence 3 at the bottom of the traction rope 26, ensuring that the extended end of the second electric push rod 212 and the electromagnetic suction plate 213 can pass through the moving groove 34 and connect with the information nameplate 5 on the conveying platform 43. The metal plate 53 at the bottom of the information nameplate 5 placed flat on the conveying platform 43 is parallel to the information nameplate 5 at this time and is stored in the storage groove 411. After the metal plate 53 is in contact, power is turned on to complete the adsorption of the electromagnetic suction plate 213 and the metal plate 53. The ordinary information nameplate 5 can be squeezed by the pressure plate 48 of the identification component 4 and rotated along the rotating shaft 52 to become vertical. At this time, the metal plate 53 is perpendicular to the information nameplate 5. The electromagnetic suction plate 213 still maintains the adsorption of the metal plate 53. Then, the information nameplate 5 is pulled back to the placement slot 32 by retracting the extended end of the first electric push rod 23. The vertical information nameplate 5 just fits inside the placement slot 32. Then, the first screw 210 is driven by the motor to rotate. The screw plate 211 and the first screw 210 are threaded together to slide along the slide frame 29. The extended end of the second electric push rod 212 moves along the moving slot 34. The information nameplate 5 It is sent into the slot 31 of the work station column 3. The outer part of the slot 31 is hollowed out, and the name information on the information nameplate 5 can be seen. There is also a limit strip to prevent the information nameplate 5 from falling. According to the identification component 4, the information nameplates 5 of different personnel are placed in different work station columns 3 in turn. According to the number of different projects and the division of labor on that day, the corresponding number of work station columns 3 and traction ropes 26 are selected for installation. First, the height of the front end of the traction rope 26, that is, the height of the top plate 11, needs to be adjusted so that all the work station columns 3 can be displayed from the front end. The newly added work station column 3 is fixed to the traction rope 26 by the clamping frames 33 and bolts 35 at both ends. A certain distance is left between adjacent work station columns 3 to avoid movement interference.

[0049] In this embodiment, the identification component 4 also includes a slide rail 41, the base 1 is located at the bottom of the slide 42 and is fixed with the slide rail 41, and the slide 42 slides along the upper end of the slide rail 41, the base 1 is fixed with a third electric push rod 214 at the position corresponding to the slide 42, and the extended end of the third electric push rod 214 is fixedly connected to the slide 42, the top of the outlet end of the conveying platform 43 is provided with a pressing plate 48, and one end of the pressing plate 48 is threadedly plugged with a second screw 46, and the other end of the pressing plate 48 is slidably plugged with a slide rod 47, and the outer side of the second screw 46 is rotatably installed with a mounting bracket 45, the bottom of the mounting bracket 45 and the slide rod 47 are fixedly connected to the base 1, and the bottom of the mounting bracket 45 is rotatably installed with a gear 4 through a bearing 9, and the gear 49 is fixedly connected to the second screw 46, one side of the gear 49 is meshed with a toothed plate 410, and the toothed plate 410 is fixedly connected to the bottom of the slide 42, and the base 1 is provided with a receiving groove 411 at the position corresponding to the toothed plate 410, and the toothed plate 410 is slidably inserted into the receiving groove 411, and the information nameplates 5 with the personnel information are placed on the conveying platform 43 in sequence, and first enter the recognition station under the belt conveyor of the conveying platform 43, and the name, age, and expertise on the information nameplate 5 are recognized by the recognition lens in the recognition chassis 44. According to the data model established by deep learning, after the information is input into the model, the optimal allocation option is obtained, and the rear-end lifting component 2 is controlled to move the corresponding station The fence 3 moves to be parallel to the exit end of the conveyor platform 43, and the information nameplate 5 continues to be conveyed by the belt of the conveyor platform 43. When entering the exit end of the conveyor platform 43, the extended end of the second electric push rod 212 passes through the moving groove 34 and is connected to the information nameplate 5. Then the second electric push rod 212 on the base 1 pushes the slide 42 to move outward, and the conveyor platform 43 gradually moves away from the station fence 3. During this process, the gear plate 410 moves with the slide 42, passes through the receiving groove 411, and meshes with the gear 49 at the bottom of the second screw 46. The gear 49 drives the second screw 46 to rotate, and one end of the pressure plate 48 is threaded with the second screw 46, and the other end slides downward along the slide rod 47 (the pressure plate 48 is initially a certain height away from the information nameplate 5). (degrees, so it does not directly squeeze the surface of the information nameplate 5) The pressing plate 48 moves a distance and contacts the upper surface of the information nameplate 5. At this time, the bottom of the information nameplate 5 has left the conveying platform 43, and the pressing plate 48 continues to move downward to rotate the information nameplate 5 downward along the rotating shaft 52, from horizontal to vertical, and then the second electric push rod 212 is used to pull the information nameplate 5 back to the placement groove 32. The information nameplate 5 is sent into the slot 31 of the work station column 3 by the drive of the first screw 210 in the slide frame 29. After different personnel are identified by age and expertise through the identification component 4, the system selects the optimal work station column 3 and allocation plan to optimize the allocation of workers.

[0050] When using the device, deep learning is used to enter the employee information and then establish a big data model. During the daily work arrangement, the driving motors 6 of the first winding seat 24 and the second winding seat 25 both unwind the traction rope 26. At this time, the overall length of the traction rope 26 released becomes longer, and the first electric push rod 23 pushes the top plate 11 to move upward, which can increase the length of the front end of the traction rope 26 to meet the needs of more workstation columns 3 installed and displayed at the front end of the traction rope 26. Similarly, synchronously winding the traction rope 26 can shorten the total length of the traction rope 26. The first electric push rod 23 drives the top plate 11 to descend, reducing the number of workstation columns 3. Employees who are on duty that day need to send their personal information nameplate 5 into the identification component 4 for identification, and first enter the identification station under the belt conveyor of the conveyor platform 43. , the name, age, and expertise on the information nameplate 5 are identified by the recognition lens in the recognition chassis 44. According to the data model established by deep learning, the information is input into the model to obtain the optimal allocation option. The system controls the first reel-up seat 24 to unwind the traction rope 26, and the second reel-up seat 25 to rewind the traction rope 26. At this time, the traction rope 26 will move toward one side of the second reel-up seat 25, and the work station column 3 installed on the traction rope 26 will also move synchronously. The two modes can realize the movement of the work station column 3 on the traction rope 26, and move the work station columns 3 at different positions to the horizontal position of the identification component 4 to receive the information nameplate 5. The position of the slide frame 29 always corresponds to the work station column 3 at the bottom of the traction rope 26, ensuring that the second electric push rod 212 is extended The electromagnetic suction plate 213 can pass through the movable slot 34 and be connected to the information nameplate 5 on the conveying platform 43. The metal plate 53 at the bottom of the information nameplate 5 placed flat on the conveying platform 43 is parallel to the information nameplate 5 at this time and is stored in the storage slot 411. After the electromagnetic suction plate 213 contacts the metal plate 53, power is turned on to complete the adsorption of the electromagnetic suction plate 213 and the metal plate 53. The second electric push rod 212 on the base 1 pushes the slide 42 to move outward, and the conveying platform 43 gradually moves away from the work station fence 3. During this process, the gear plate 410 moves with the slide 42, passes through the storage slot 411, and engages with the gear 49 at the bottom of the second screw 46. The gear 49 drives the second screw 46 to rotate, and one end of the pressure plate 48 is threaded with the second screw 46 and the other end moves downward along the slide rod 47. After sliding, the pressing plate 48 moves a distance and contacts the upper surface of the information nameplate 5. At this time, the bottom of the information nameplate 5 has left the conveying platform 43, and the pressing plate 48 continues to move downward, rotating the information nameplate 5 downward along the rotating shaft 52 from horizontal to vertical, and then the second electric push rod 212 pulls the information nameplate 5 back into the placement slot 32. The information nameplate 5 is sent into the slot 31 of the work station column 3 by driving the first screw 210 in the slide frame 29. The first screw 210 is driven by the motor to rotate, and the screw plate 211 and the first screw 210 are threadedly matched to slide along the slide frame 29. The extended end of the second electric push rod 212 moves along the moving slot 34, and the information nameplate 5 is sent into the slot 31 of the work station column 3. The outer part of the slot 31 is hollowed out.The name information on the information nameplate 5 can be seen, and there is also a limit strip to prevent the information nameplate 5 from falling. According to the identification component 4, the information nameplates 5 of different people are placed in different workstation columns 3 in sequence. Through the display of the workstation columns 3, employees can know the work they need to do that day and the employees who are traveling with them.

[0051] The basic principles, main features and advantages of the present invention are shown and described above. It is obvious to those skilled in the art that the present invention is not limited to the details of the above exemplary embodiments, and that the present invention can be implemented in other specific forms without departing from the spirit or basic features of the present invention.

[0052] In addition, it should be understood that although this specification is described in terms of implementation methods, not every implementation method contains only one independent technical solution. This narrative method of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment can also be appropriately combined to form other implementation methods that can be understood by those skilled in the art.

Claims

1. A labor organization optimization setting method based on deep learning, characterized by: The optimization setting method comprises the following steps: (1) Enter information uniformly based on the employee's name, age, and expertise, and develop a unique nameplate for each employee; (2) Use Resnet algorithm as the classification deep learning algorithm to train and predict employee classification models; deep learning classification network, import information data into VGG and ResNet classification networks respectively for classification model training, and select the deep learning classification network by combining the accuracy of the training set and the test set according to the results of the change of classification accuracy and classification loss with the number of training times. The optimal model is embedded in the sorting system for employee information to achieve non-destructive detection and intelligent grading of employee labor distribution; (3) The established learning classification model is input into the recognition component system. Employees who come to work every day put their nameplates into the recognition component. The recognition component uses the employee's age and expertise to make the best allocation plan using the classification model. (4) The result obtained by the identification component is fed back to the control system of the upgrade component. The lifting component moves the corresponding workstation column to a position parallel to the conveyor table and stores the information nameplate sent by the identification component into the corresponding workstation column. Employees can see the work items of the day and the cooperating employees in the workstation column; (5) According to the daily workload and the number of employees on duty that day, reasonable workstation columns and lifting components are selected for installation, and after identification and optimization classification by the identification component, they are finally displayed on the workstation column.

2. The labor organization optimization setting method based on deep learning according to claim 1 is characterized by: According to the learning classification model storage described in step (2), the model is persistently stored on the local hard disk on the training host, and then the model file is imported into the control center host. The persistent storage uses Python's pickle operation to serialize the preferred model. The pickle operation serializes the model trained by deep learning, saves this serialized format into a file, and is deserialized by the sugar content grading module of the control center to achieve online classification prediction.

3. The labor organization optimization setting method based on deep learning according to claim 1 is characterized by: According to the identification information described in step (3), the identification component includes a conveyor table, and the information nameplates with personal information are placed on the conveyor table in sequence. They first enter the identification station under the belt conveyor of the conveyor table, and the name, age, and expertise on the information nameplate are identified by the identification lens in the identification chassis.

4. The labor organization optimization setting method based on deep learning according to claim 1 is characterized by: According to the lifting assembly described in step (4), the lifting assembly includes two first winding seats and two second winding seats, transmission belts are installed on the outer sides of the two first winding seats and the two second winding seats, and the pulleys of the transmission belts are fixedly connected to their respective winding shafts, driving motors are fixed at positions corresponding to the first winding seat and the second winding seat at the bottom of the top plate, and the driving motors are fixedly connected to the winding shafts of the first winding seat and the second winding seat, respectively, a vertical frame is fixed to the middle top of the limit seat, and two groups of first electric push rods are fixed to the top of the vertical frame, and the extended ends of the first electric push rods are fixedly connected to the top plate.

5. The labor organization optimization setting method based on deep learning according to claim 4 is characterized in that: After the lifting component receives the feedback information of the identification component, the first reeling seat reels in the traction rope and the second reeling seat unwinds the traction rope. At this time, the traction rope will move toward one side of the first reeling seat, and the work station bar installed on the traction rope will also move synchronously. The first reeling seat unwinds the traction rope and the second reeling seat rewinds the traction rope. At this time, the traction rope will move toward one side of the second reeling seat, and the work station bar installed on the traction rope will also move synchronously. The two modes can realize the movement of the work station bar on the traction rope, and move the work station bars at different positions to the horizontal position of the identification component for receiving the information nameplate.

6. The labor organization optimization setting method based on deep learning according to claim 1 is characterized by: According to the work station column described in step (4), a movable groove and a sliding frame are provided on the back of the work station column. The first electric push rod inside the sliding frame can pass through the movable groove and connect with the information nameplate on the conveying platform. The metal plate at the bottom of the information nameplate placed flat on the conveying platform is parallel to the information nameplate and is stored in the storage groove. After the electromagnetic suction plate contacts the metal plate, power is turned on to complete the adsorption of the electromagnetic suction plate and the metal plate.

7. The labor organization optimization setting method based on deep learning according to claim 6 is characterized in that: A second electric push rod is provided on the base for pushing the slide of the identification component to move. After the extended end of the second electric push rod passes through the moving slot and is connected to the information nameplate, the second electric push rod on the base pushes the slide to move outward, and the conveying platform gradually moves away from the work station fence.

8. The labor organization optimization setting method based on deep learning according to claim 7 is characterized in that: When the identification component is pushed outward, a tooth plate is fixed to one side of the bottom of the slide, and the tooth plate moves with the slide, passes through the storage groove, and meshes with the gear at the bottom of the second screw. The gear drives the second screw to rotate, and one end of the pressure plate cooperates with the second screw thread and the other end slides downward along the slide rod. After moving a certain distance, the pressure plate contacts the upper surface of the information nameplate. At this time, the bottom of the information nameplate has left the conveying platform, and the pressure plate continues to move downward to rotate the information nameplate downward along the rotating axis from horizontal to vertical, and then the second electric push rod is used to pull the information nameplate back into the placement slot, and the information nameplate is sent into the slot of the work station column by the drive of the first screw in the slide frame.

9. The labor organization optimization setting method based on deep learning according to claim 5 is characterized by: According to the installation of the workstation bar described in step (5), clamp frames are set at both ends of the workstation bar, and the information nameplates of different personnel are placed in different workstation bars in turn according to the identification components. According to the number of different projects and the division of labor on that day, the corresponding number of workstation bars and traction ropes are selected for installation. First, the height of the front end of the traction rope, that is, the height of the top plate, needs to be adjusted so that all the workstation bars can be displayed from the front end. The newly added workstation bar is fixed to the traction rope by the clamp frames and bolts at both ends, and a certain distance is left between adjacent workstation bars to avoid movement interference.

10. The labor organization optimization setting method based on deep learning according to claim 9 is characterized in that: When increasing and decreasing the number of workstation columns, when the driving motors of the first winding seat and the second winding seat both unwind the traction rope, the overall length of the traction rope released becomes longer. By pushing the top plate upward by the first electric push rod, the length of the front end of the traction rope can be increased to meet the needs of more workstation columns to be installed and displayed at the front end of the traction rope. Synchronously rewinding the traction rope can shorten the total length of the traction rope. The first electric push rod drives the top plate to descend, reducing the number of workstation columns.

Images