Workpiece tracing and sorting system and method based on RFID and visual inspection

Abstract

The invention discloses a workpiece tracing and sorting system and method based on RFID and visual inspection, and the system comprises (1) a composite sensing and positioning unit which comprises an RFID tracing module which is used for collecting the identity information of a workpiece and a processing path; the visual detection module is used for acquiring visual feature quality data of the workpiece; the conveying line positioning unit is used for realizing accurate positioning of the tool plate; the intelligent sorting execution system (2) comprises an industrial robot unit used for executing the sorting action; the stereoscopic warehouse module is provided with a raw material warehouse subarea and a finished product warehouse subarea and is used for storing raw materials and finished products in a subarea manner; (3) the control center adopts a PLC (Programmable Logic Controller) to realize the following cooperative control: triggering the positioning unit to act and synchronizing RFID (Radio Frequency Identification Device) data and visual data; and a sorting instruction is generated based on the fusion data and the storage location state, and the robot track is optimized. According to the invention, high-precision and high-efficiency tracing and sorting of the whole life cycle of workpieces in multi-class mixed flow production are realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of industrial automation, in particular to a workpiece full life cycle tracing and intelligent sorting system and method fusing RFID tag identification, machine vision detection and conveying line accurate positioning, which is suitable for information tracing and automatic sorting operation in multi-category workpiece mixed flow production scene. BACKGROUND

[0002] In the industrial automation production line, the tracing and sorting of workpieces face the following technical bottlenecks:

[0003] The traditional tracing system relies on single RFID or vision technology, which has problems such as incomplete information collection (such as only recording position without associating quality data), large positioning error (more than ±2mm) leading to detection failure, etc.

[0004] The positioning accuracy of the conveying line in the sorting process is insufficient (such as relying only on blocking air cylinder single positioning), which is easy to cause robot grabbing deviation, leading to workpiece collision or sorting error;

[0005] Multi-source data (RFID information, vision detection results, positioning state) lack coordination mechanism, which cannot realize the three-dimensional tracing of workpiece "position-quality-process", especially in mixed flow production, the tracing accuracy is less than 95%;

[0006] The traditional sorting strategy is fixed, which is difficult to adapt to the dynamic sorting demand of multi-variety workpieces. SUMMARY

[0007] The purpose of the present application is to provide a workpiece tracing and sorting system and method based on RFID and vision detection, which realizes high-precision and high-efficiency tracing and sorting of workpieces in multi-category mixed flow production through special-shaped jacking mechanism accurate positioning, multi-source data collaborative modeling and dynamic sorting strategy, and solves the technical bottlenecks of large positioning error, single tracing dimension and poor sorting adaptability of traditional single technical solution.

[0008] In order to achieve the above purpose, the present application adopts the following technical scheme:

[0009] A workpiece tracing and sorting system based on RFID and vision detection, comprising:

[0010] (1) a composite perception and positioning unit, comprising:

[0011] An RFID tracing module for collecting workpiece identity information and processing path;

[0012] A vision detection module for obtaining workpiece visual feature quality data;

[0013] A conveying line positioning unit for realizing accurate positioning of the tool plate;

[0014] The conveying line positioning unit comprises:

[0015] A tooling plate detection sensor and a workpiece detection sensor for triggering the positioning process;

[0016] A blocking positioning mechanism for preliminary stopping of the tooling plate;

[0017] A special-shaped jacking mechanism with a guide slope for correcting the horizontal position of the tooling plate;

[0018] (2) An intelligent sorting execution system, comprising:

[0019] An industrial robot unit for performing sorting actions;

[0020] A stereoscopic warehouse module provided with raw material warehouse partitions and finished product warehouse partitions for storing raw materials and finished products in partitions;

[0021] (3) A control center using PLC to realize the following collaborative control:

[0022] Triggering the action of the positioning unit and synchronizing RFID data and visual data;

[0023] Generating sorting instructions and optimizing robot trajectories based on fused data and bin status.

[0024] Further, the bin of the stereoscopic warehouse module is provided with a workpiece presence state detection sensor.

[0025] Further, the special-shaped jacking mechanism comprises:

[0026] A jacking cylinder;

[0027] Two jacking pins with special-shaped structures at the ends, wherein the first jacking pin has a circular structure at the end, and the second jacking pin has a trapezoidal structure at the end; the trapezoidal structure corrects the horizontal position of the tooling plate through a guide slope.

[0028] Further, in the control center,

[0029] After the positioning unit completes position correction, synchronize RFID identity and path information, and visual detection quality feature data through a time stamp;

[0030] Establish a three-dimensional data model of "workpiece ID-positioning coordinate-quality state", wherein:

[0031] The mapping of workpiece ID and positioning coordinate satisfies: positioning accuracy ±0.5mm corresponds to a visual field center deviation ≤1 pixel.

[0032] Further, the blocking positioning mechanism comprises a blocking cylinder and a stop block, and the stroke of the blocking cylinder is 50mm.

[0033] A workpiece tracing sorting method based on the system, comprising:

[0034] (a) a cooperative positioning step:

[0035] The workpiece plate is stopped by triggering the blocking mechanism through the sensor;

[0036] The special-shaped jacking mechanism is driven to insert the pin hole to correct the position;

[0037] (b) a data fusion step:

[0038] RFID identity data and visual quality data are synchronously collected;

[0039] (c) an intelligent sorting step:

[0040] The PLC fuses double-source data to generate a sorting instruction;

[0041] The industrial robot performs sorting based on the state of the storage location.

[0042] Further, in the cooperative positioning step, the special-shaped jacking pin is inserted into the pin hole to correct the horizontal position of the workpiece plate through the guide slope, and the correction time is ≤100 ms, and the positioning accuracy is ±0.2 mm.

[0043] Further, the intelligent sorting step includes a dynamic storage location allocation strategy: when the target storage location is occupied, adjacent empty storage locations of the same color are preferentially allocated, and the response time is ≤200 ms.

[0044] Further, in the intelligent sorting step, the sorting instruction is executed:

[0045] Qualified workpieces are classified into the warehouse according to color characteristics;

[0046] Unqualified workpieces are transferred to the waste area and the type is recorded.

[0047] Further, the path planning error of the industrial robot is ≤2%.

[0048] Beneficial effects: Compared with the prior art, the present application has the following technical effects:

[0049] 1. Double-source data cooperative positioning: RFID provides workpiece identity and path information, and the positioning unit ensures accurate workpiece position during visual detection, so that the tracing accuracy is improved to 99.5% (traditional single technical solution is 92%).

[0050] 2. Special-shaped pin secondary positioning: the guide correction function when the trapezoidal jacking pin is inserted, compared with the traditional circular pin positioning, the visual detection field center deviation is reduced from ±1.5 mm to ±0.5 mm, and the robot grasping success rate is improved from 95% to 99.8%.

[0051] 3. Dynamic sorting strategy: support rule customization and bin intelligent allocation, sorting efficiency in mixed flow production scene is improved by 30% compared with fixed strategy, and different industrial sorting needs can be simulated in practical teaching.

[0052] 4. Fault self-diagnosis: by comparing RFID read-write state, positioning cylinder pressure value and visual detection result, real-time detection of abnormal workstations (such as lifting pin not in place, color recognition timeout), automatic triggering of alarm and recording of fault points (positioning timestamp + fault type). BRIEF DESCRIPTION OF DRAWINGS

[0053] Figure 1 is a system overall architecture diagram;

[0054] Figure 2 is a collaborative work flow diagram of the conveying line positioning unit and the RFID reader;

[0055] Figure 3 is a multi-source data fusion algorithm logic diagram;

[0056] Figure 4 is an adaptive sorting strategy decision tree. DETAILED DESCRIPTION

[0057] The application will be further explained in conjunction with the accompanying drawings.

[0058] As shown in Figure 1 , a workpiece tracing and sorting system based on RFID and visual detection of the application comprises:

[0059] (1) a composite perception and positioning unit, comprising:

[0060] an RFID tracing module for collecting workpiece identity information and processing path;

[0061] In some embodiments, each tooling plate integrates a high-frequency RFID tag (storage capacity ≥ 2KB) to record workpiece ID, material attributes, processing path and other information; RFID readers (reading distance ≥ 30cm) are deployed at key workstations such as in-out bin, visual detection station, etc., to interact with the master control PLC in real time through Modbus protocol.

[0062] a visual detection module for obtaining workpiece visual feature quality data;

[0063] In some embodiments, a 5 million pixel industrial camera (GigE interface) is used with a ring light source to detect workpiece color, size and other features

[0064] a conveying line positioning unit for realizing precise positioning of tooling plates; the conveying line positioning unit is integrated in the detection station.

[0065] In some embodiments, the conveying line positioning unit comprises:

[0066] 1 tooling plate detection photoelectric sensor (response time ≤10 ms) and 1 workpiece detection photoelectric sensor;

[0067] 1 blocking cylinder (stroke 50 mm) and a block, realizing preliminary positioning of the tooling plate;

[0068] 1 jacking cylinder and 2 jacking pins, wherein the first jacking pin has a circular structure at the end, and the second jacking pin has a trapezoidal structure at the end; the secondary positioning (accuracy ±0.2 mm) is realized by the cooperation of the pin holes.

[0069] (2) Intelligent sorting execution system, comprising:

[0070] Industrial robot unit for performing sorting actions;

[0071] In some embodiments, a degree of freedom industrial robot is used, equipped with a two-fingered gripper, and based on the visual detection results (color classification accuracy ≥99.9%) and RFID information to perform sorting actions:

[0072] Qualified workpieces are classified by color and stored in the warehouse (e.g. red → No. 1-3 warehouse, blue → No. 4-6 warehouse);

[0073] Unqualified workpieces are transferred to the waste area, and the unqualified types (color deviation, size out-of-tolerance, etc.) are recorded by PLC.

[0074] The stereoscopic warehouse module is provided with raw material warehouse partitions and finished product warehouse partitions for raw material and finished product partition storage;

[0075] In some embodiments, the stereoscopic warehouse module is a four-layer aluminum profile shelf (12 warehouse positions), with 6 raw material warehouses on the left and 6 finished product warehouses on the right, and each warehouse position is provided with a sensor to detect the presence of workpieces.

[0076] (3) Control center, using PLC to realize the following collaborative control:

[0077] Triggering the positioning unit action and synchronizing the RFID data and visual data;

[0078] Generating sorting instructions based on the fusion data and warehouse position state and optimizing the robot trajectory.

[0079] In some embodiments, the control center comprises a control system and a human-machine interface, wherein:

[0080] The control system uses PLC, integrates the "positioning-detection-sorting" collaborative control algorithm, and realizes the following functions:

[0081] Positioning trigger: when the RFID reader detects that the tooling plate reaches the station, the light sensor detection is triggered synchronously with the action of the conveying line positioning unit;

[0082] Data fusion: synchronize RFID information (workpiece ID, path) and visual detection results (color, size) through time stamp, and establish a three-dimensional data model of "workpiece ID-positioning coordinates-quality state";

[0083] Sorting decision: generate sorting instructions according to double-source data, combine with the state of the stereoscopic warehouse location (real-time acquisition through sensors), and optimize the robot motion trajectory.

[0084] As Figure 2 The working flow chart of the conveying line positioning unit and the RFID reader is shown, which includes the timing relationship of positioning trigger, data acquisition, and detection execution.

[0085] The human-computer interaction interface adopts a touch screen, supports sorting rule customization (color priority / size priority), warehouse location strategy switching (fixed warehouse location / nearest storage), and historical data query (traceability time span ≥ 30 days).

[0086] Based on the workpiece tracing and sorting system based on RFID and visual detection, the present application proposes a multi-source data driven accurate positioning and tracing method, which includes the following steps:

[0087] (1) Conveying line positioning step:

[0088] The tooling plate runs to the positioning unit → the tooling plate detection photoelectric sensor is triggered → the blocking cylinder extends the blocking block for preliminary positioning; the workpiece detection photoelectric sensor confirms that there is a workpiece → the lifting cylinder drives the special-shaped lifting pin (trapezoidal end) to insert into the pin hole, and the horizontal position is corrected through the guide slope to complete the secondary positioning;

[0089] The RFID reader reads the workpiece information, and the visual detection camera collects images (photographing is completed within 200ms after positioning is completed) synchronously.

[0090] (2) Workpiece tracing step:

[0091] Establish the mapping relationship between the workpiece ID and the positioning coordinates (the positioning accuracy ±0.5mm corresponds to the deviation of the center of the visual detection field ≤1 pixel);

[0092] Record the arrival time and detection results of the workpiece at each positioning station through the RFID tag, and form a "position-time-quality" tracing chain;

[0093] Support reverse tracing of abnormal workpieces: if it is found that the size of a workpiece is out of tolerance, the positioning state (such as whether the lifting pin is inserted in place) and the historical processing path of the workpiece at the visual detection station can be queried through the RFID.

[0094] As Figure 3 shown, the fusion process of RFID information, visual detection results and positioning state is shown.

[0095] For the above multi-source data driven accurate positioning and tracing method, the application provides an adaptive sorting strategy optimization method, as Figure 4 shown, the allocation logic of the storage location under different rules is shown, and the details are as follows:

[0096] (1) Dynamic rule configuration:

[0097] Pre-set color classification rules (such as red / yellow / blue corresponding to different storage location partitions), size grading rules (qualified / overdue);

[0098] Support importing custom rules through touch screen (such as "workpiece A is stored in No. 5 storage location first"), and the rule analysis time is ≤100ms.

[0099] (2) Intelligent allocation of storage location:

[0100] Normal mode: allocate fixed storage location according to color rules;

[0101] Dynamic mode: when the target storage location is full, the PLC triggers the "same color adjacent storage location" strategy, queries the adjacent storage location state through RFID, and selects the nearest empty storage location (response time ≤200ms).

[0102] Embodiment

[0103] When the workpiece on the tooling plate runs to the visual detection station along the conveying line:

[0104] 1. Positioning unit action: tooling plate bottom detection block triggers photoelectric sensor → blocking cylinder extends blocking block to stop tooling plate → workpiece detection sensor confirms that there is a workpiece → jack-up cylinder drives special-shaped jack-up pin to insert into pin hole, uses trapezoidal slope to correct tooling plate position, realizes ±0.5mm precision positioning.

[0105] 2. Data acquisition: RFID reader reads tooling plate tag information (such as workpiece ID: W001, material: aluminum alloy, processing path: raw material warehouse → conveying line) → visual camera takes pictures of workpiece, identifies color (such as yellow) and size (qualified) through visual algorithm.

[0106] 3. Sorting execution: the master control PLC judges the workpiece to be a qualified yellow workpiece according to the RFID information and visual results → queries the state of No. 4-6 storage location of the stereoscopic warehouse, if No. 4 storage location is empty → generates sorting instruction → robot plans the shortest path (error ≤2%) to grab the workpiece → stores it in No. 4 storage location → RFID reader records the storage time and storage location information.

[0107] The above merely describes the preferred embodiments of the present application, and it should be pointed out that, for those skilled in the art, several improvements and refinements can be made without departing from the principles of the present application, and these improvements and refinements should also be considered as falling within the protection scope of the present application.

Claims

1. A workpiece traceability sorting system based on RFID and vision detection, characterized in that: Comprise: (1) Composite perception and positioning unit, comprising: RFID traceability module for collecting workpiece identity information and processing path; Visual inspection module for obtaining workpiece visual feature quality data; Conveying line positioning unit for realizing precise positioning of tooling plate; The conveying line positioning unit comprises: Tooling plate detection sensor and workpiece detection sensor for triggering positioning process; Blocking positioning mechanism for preliminary stop of tooling plate; Special-shaped jacking mechanism with guide slope for correcting horizontal position of tooling plate; (2) Intelligent sorting execution system, comprising: Industrial robot unit for executing sorting action; Stereo warehouse module provided with raw material warehouse partition and finished product warehouse partition for raw material and finished product partition storage; (3) Control center, adopting PLC to realize the following collaborative control: Triggering positioning unit action and synchronizing RFID data and visual data; Generating sorting instruction based on fusion data and library position state and optimizing robot trajectory.

2. The system of claim 1, wherein: The library position of the stereo warehouse module is provided with a workpiece existence state detection sensor.

3. The system of claim 1, wherein: The special-shaped jacking mechanism comprises: A jacking cylinder; Two jacking pins with special-shaped structure at the ends, wherein the first jacking pin end is circular structure and the second jacking pin end is trapezoidal structure; the trapezoidal structure corrects the horizontal position of the tooling plate through the guide slope.

4. The system of claim 1, wherein: In the control center, After the positioning unit completes position correction, RFID identity and path information, visual detection quality feature data are synchronized through time stamp; A three-dimensional data model of "workpiece ID-positioning coordinate-quality state" is established, wherein: The mapping of workpiece ID and positioning coordinate satisfies: positioning accuracy ±0.5mm corresponds to visual field center deviation ≤1 pixel.

5. The system of claim 1, wherein: The blocking positioning mechanism comprises a blocking cylinder and a stop block, and the blocking cylinder stroke is 50mm.

6. A workpiece traceability sorting method based on the system of claim 1, characterized by: Comprise: (a) Collaborative positioning step: Stop the tooling plate by triggering the blocking mechanism through the sensor; Drive the special-shaped jacking mechanism to insert the pin hole to correct the position; (b) Data fusion step: Synchronize RFID identity data and visual quality data acquisition; (c) Intelligent sorting step: PLC fuses double-source data to generate sorting instruction; Industrial robot executes sorting based on library position state.

7. The method of claim 6, wherein: In the collaborative positioning step, the special-shaped jacking pin inserts the pin hole to correct the horizontal position of the tooling plate through the guide slope, the correction time is ≤100ms, and the positioning accuracy is ±0.2mm.

8. The method of claim 6, wherein: The intelligent sorting step contains dynamic library position allocation strategy: when the target library position is occupied, preferentially allocate adjacent empty library position of the same color, and the response time is ≤200ms.

9. The method of claim 6, wherein: In the intelligent sorting step, the sorting instruction is executed: Qualified workpieces are classified into the warehouse according to color characteristics; Unqualified workpieces are transferred to the waste area and the type is recorded.

10. The method of claim 6, wherein: the industrial robot path planning error is < 2%.

Images