Electrode holder electrode cap grinding detection method and system based on AI visual diagnosis

Abstract

The invention provides an electrode holder electrode cap grinding detection method and system based on AI visual diagnosis, through a visual identification system integrated with a light source system, an industrial camera, AI visual processing software, a PLC controller and the like, a robot moves to a camera identification position after grinding of an electrode holder is completed, the robot sends an in-place signal, the visual system carries out photographing identification, and the detection result is obtained. And after identification is completed, an identification result signal is sent to the robot. If the electrode cap grinding is OK, the robot directly returns to the safe standby position and enters the next procedure, and if the electrode cap grinding is NG, the robot stops at the current position, an alarm signal is sent, and manual intervention is conducted. According to the method, 100% real-time feedback and alarm mistake proofing of robot electrode holder electrode cap coping are achieved, compared with a traditional mode, no welding parameter needs to be adjusted, the quality problem caused by parameter change is effectively avoided, meanwhile, welding quality caused by improper electrode cap coping is prevented, the starting rate of whole-line equipment is increased, and the repair cost is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of automotive welding, and in particular, to a method and system for grinding and detecting welding tong electrode caps based on AI vision diagnosis. Background Art

[0002] In the body welding production line of an automotive manufacturing plant, during the robot welding process, quality accidents such as false welding of the body solder joints often occur due to the loss or wear of the grinding blade, resulting in unqualified grinding of the robot welding tong electrode cap.

[0003] The existing solution only judges whether the electrode cap of the welding tong is ground qualified by detecting the resistance of the empty gun of the welding tong. The principle is that after the robot welding tong is ground, a small current solder joint parameter is set, and the real-time resistance detected is compared with the set resistance value. If it exceeds the upper and lower limits of the set value, it is regarded as unqualified grinding, and the welding machine then feeds back the unqualified signal to the robot for alarm reminder. The prerequisite for this solution is that the welding machine must have an adaptive function, and welding machines with a relatively long service life on site do not have this function. Moreover, the empty gun resistance detection result is greatly affected by the electrode cap material and the electrode arm state, and false alarms often occur, which will have a certain impact on the equipment operation rate. Summary of the Invention

[0004] Aiming at the defects in the prior art, the purpose of the present invention is to provide a method and system for grinding and detecting welding tong electrode caps based on AI vision diagnosis. Through the AI vision recognition system, real-time feedback and alarm anti-error of the robot welding tong electrode cap grinding are realized, without adjusting any welding parameters, effectively avoiding quality problems caused by parameter changes, preventing welding quality problems caused by incorrect or insufficient grinding of the electrode cap, improving the operation rate of the whole line of equipment, and reducing the rework cost.

[0005] To achieve the above technical effects, the present invention adopts the following technical solutions:

[0006] According to the first aspect of the present invention, there is provided a method for grinding and detecting welding tong electrode caps based on AI vision diagnosis, including the following steps:

[0007] S1. The robot completes the electrode cap grinding operation;

[0008] S2. The robot moves the ground welding tong electrode cap to a preset visual inspection station and sends a in-place signal to the PLC controller;

[0009] S3. The PLC controller responds to the in-place signal and sends a photographing instruction to the vision system;

[0010] S4. After receiving the photographing instruction, the vision system turns on the light source system and collects a high-definition image of the electrode cap surface through an industrial camera;

[0011] The S5. AI vision processing software identifies the electrode cap features in the image based on a pre-trained machine learning model and generates a grinding quality identification result;

[0012] S6. Feed the identification result back to the robot control system:

[0013] If the result is qualified (OK), the robot returns to the safe standby position;

[0014] If the result is unqualified (NG), the robot stops moving and triggers an alarm signal, and at the same time maintains the position of the welding tongs for manual inspection.

[0015] Optionally, the electrode cap features include at least one of end face flatness, chamfer state, crack and excessive wear.

[0016] Optionally, in the step S6: the identification result is transmitted to the robot control system through the PLC controller.

[0017] Optionally, in the step S6: the alarm signal is implemented through an audible and visual alarm and / or the HMI interface.

[0018] According to the second aspect of the present invention, there is provided a welding tong electrode cap grinding detection system for implementing the above method, including:

[0019] Robot system: including a robot body and a robot control system, used to perform electrode cap grinding operations and move to the detection station;

[0020] Vision system: including an industrial camera, a light source system and AI vision processing software, used to collect images and output identification results;

[0021] PLC controller: configured to: (a) receive the in-place signal of the robot system and send a photographing instruction to the vision system; (b) receive the identification result of the vision system and forward it to the robot system.

[0022] Optionally, the vertical distance between the central axis of the light source system and the grinding end face of the electrode cap is 50 mm.

[0023] Optionally, the industrial camera is vertically arranged and its central axis is accurately aligned with the central axis of the electrode cap.

[0024] Optionally, it further includes an alarm device, which is communicatively connected to the PLC controller and is used to trigger audible and visual alarms and / or HMI prompts in response to the NG signal.

[0025] Compared with the prior art, the present invention has the following beneficial effects:

[0026] Through a vision recognition system integrated with a light source system, an industrial camera, AI vision processing software, a PLC controller, etc., the present invention realizes 100% real-time feedback and alarm error prevention for the grinding of the electrode caps of robot welding tongs. Compared with the traditional method, no welding parameters need to be adjusted, effectively avoiding quality problems caused by parameter changes. At the same time, it prevents welding quality problems caused by incorrect or insufficient grinding of the electrode caps, improves the operating rate of the whole line of equipment, and reduces the rework cost. Description of the Drawings

[0027] By reading the following detailed description of the non-limiting embodiments with reference to the accompanying drawings, other features, objects, and advantages of the present invention will become more apparent:

[0028] Figure 1 It is a flowchart of the method for detecting the grinding of the electrode caps of the welding tongs based on AI vision diagnosis described in the first embodiment. Detailed Embodiments

[0029] To make the objectives, technical solutions, and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application. Apparently, the described embodiments are some, but not all, of the embodiments of the present application. Usually, the components of the embodiments of the present application described and illustrated in the accompanying drawings here can be arranged and designed in various different configurations.

[0030] Therefore, the following detailed description of the embodiments of the present application provided in the accompanying drawings is not intended to limit the scope of the present application claimed, but merely represents selected embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without making creative efforts fall within the scope of protection of the present application.

[0031] It should be noted that: similar reference numerals and letters denote similar items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings. In addition, all directional indications (such as up, down, left, right, front, back, bottom...) in the present application are only used to explain the relative positional relationship and movement conditions between components in a specific posture (as shown in the drawings). If the specific posture changes, the directional indications will also change accordingly. Further, the descriptions involving "first", "second", etc. in the application are only for descriptive purposes and cannot be understood as indicating or implying their relative importance or implicitly indicating the quantity of the indicated technical features.

[0032] The First Embodiment

[0033] As Figure 1As shown, this embodiment provides a welding clamp electrode cap grinding detection method based on AI visual diagnosis. Through the AI visual recognition system, 100% real-time feedback and alarm error prevention are achieved during robot welding clamp electrode cap grinding. No welding parameters need to be adjusted, effectively avoiding quality problems caused by parameter changes. At the same time, it prevents incorrect electrode cap repairs that may affect welding quality, improves the availability of the entire line equipment, and reduces rework costs. Specifically, it includes the following steps:

[0034] S1. Preparation and triggering: The robot completes the electrode cap grinding operation.

[0035] S2. Positioning and Imaging: The robot's welding clamp, carrying the ground electrode cap, autonomously moves to a pre-set visual inspection station. Upon reaching the designated visual inspection location, the robot's control system sends an "in-place" signal to the PLC controller, which in turn sends a trigger signal to the visual system to capture a picture.

[0036] S3. Visual inspection: After the visual system receives the "trigger photo" signal, the industrial camera performs image acquisition and takes a high-definition image of the current electrode cap.

[0037] S4.AI visual processing software processes and analyzes the captured images in real time. Using pre-trained models, it proactively identifies key features of the electrode cap (such as end face flatness, chamfer condition, cracks, excessive wear, etc.) and calculates whether the grinding quality is up to standard (OK) or not (NG).

[0038] S5. Result feedback: The AI vision processing software sends the recognition result ("OK" or "NG" signal) to the robot control system through the PLC controller.

[0039] S6. Decision and Execution: If the robot control system receives an "OK" signal, the robot automatically returns to its designated safe standby position (HOME) without performing any additional operations. The system then proceeds to the next work cycle. If the robot control system receives an "NG" signal, the robot immediately stops at its current position, maintaining the welding clamp position for easy inspection. Simultaneously, the PLC controller triggers an alarm (audio-visual alarm, HMI prompt, etc.) to notify the operator. The operator then examines the cause of the alarm and determines whether manual adjustments or re-grinding are necessary.

[0040] Second embodiment

[0041] This embodiment provides a welding clamp electrode cap grinding and detection system based on AI visual diagnosis, which is used to implement the welding clamp electrode cap grinding and detection method based on AI visual diagnosis described in the first embodiment. It specifically includes the following components: a robotic system, a visual system, and a PLC controller.

[0042] Among them, the robot system includes a robot body and a robot control system. The robot body is used to carry the welding tong electrode cap from the grinding position to the fixed detection position. The robot control system is connected to the PLC controller, and is used to send a "in place" signal to the PLC controller, or receive the recognition result of the grinding quality of the electrode cap sent by the PLC controller, and then control the robot to return to the safe standby position (grinding position) or the fixed detection position.

[0043] The vision system includes an imaging unit and AI vision processing software. The imaging unit includes an industrial camera and a light source system, and is used to obtain a high-definition image of the electrode cap surface. Among them, the light source system is designed according to the geometric characteristics and surface reflection characteristics of the electrode cap, and is arranged on the upper rack of the fixed detection position. The vertical distance between its central axis and the grinding end face of the electrode cap is 50 mm, and is used to provide a stable, uniform and high-contrast lighting environment. The industrial camera is vertically arranged above the fixed detection position, and its central axis is accurately aligned with the central axis of the electrode cap, and is used to accurately capture the image of the ground electrode cap.

[0044] The AI vision processing software is the core part of this system, and is equipped with an active learning image processing model based on machine learning, and is used to actively identify the key features of the electrode cap and calculate the recognition result of the grinding quality (OK / NG).

[0045] The PLC controller is used to receive the "in place" signal of the robot and send a photo-taking trigger instruction to the vision system. At the same time, it is used to receive the "OK" or "NG" signal of the recognition result and transmit this signal to the robot control system.

[0046] The system also includes an alarm device, which is communicatively connected to the PLC controller and is used to trigger an audible and visual alarm and / or HMI prompt in response to the NG signal.

[0047] The specific embodiments of the present invention have been described above. Through the above description, relevant staff can make various changes and modifications completely within the scope of not deviating from the technical idea of this invention.

Claims

1. A method for detecting and grinding the electrode cap of a welding tong based on AI vision diagnosis, characterized in that, It includes the following steps: S1. The robot completes the electrode cap grinding operation; S2. The robot carries the ground welding tong electrode cap and moves to the preset visual inspection station, and sends an in-place signal to the PLC controller; S3. The PLC controller responds to the in-place signal and sends a photographing instruction to the vision system; S4. After receiving the photographing instruction, the vision system turns on the light source system and acquires a high-definition image of the surface of the electrode cap through an industrial camera; S5. The AI vision processing software identifies the features of the electrode cap in the image based on a pre-trained machine learning model, and generates a grinding quality identification result; S6. The identification result is fed back to the robot control system: If the result is qualified (OK), the robot returns to the safe standby position; If the result is unqualified (NG), the robot stops moving and triggers an alarm signal, and at the same time maintains the position of the welding tong for manual inspection.

2. The method according to claim 1, characterized in that, The electrode cap features include at least one of end face flatness, chamfer state, crack and excessive wear.

3. The method according to claim 1, characterized in that, In step S6: The identification result is transmitted to the robot control system through the PLC controller.

4. The method according to claim 1, characterized in that In step S6: The alarm signal is implemented through an audible and visual alarm and / or an HMI interface.

5. A welding tongs electrode cap grinding and detection system for implementing the method according to any one of claims 1 to 4, characterized in that, It includes: Robot system: including a robot body and a robot control system, used to perform electrode cap grinding operations and move to the inspection station; Vision system: including an industrial camera, a light source system and AI vision processing software, used to acquire images and output identification results; PLC controller: configured to: (a) receive the in-place signal from the robot system and send a photographing instruction to the vision system; (b) receive the identification result from the vision system and forward it to the robot system.

6. The system according to claim 5, wherein The vertical distance between the central axis of the light source system and the ground end face of the electrode cap is 50 mm.

7. The system according to claim 5, characterized in that, The industrial camera is vertically arranged and its central axis is precisely aligned with the central axis of the electrode cap.

8. The system according to claim 5, wherein It also includes an alarm device, which is communicatively connected to the PLC controller and is used to trigger audible and visual alarms and / or HMI prompts in response to the NG signal.

Images