Slurry rotary drum AI visual identification automatic scraping system

Abstract

The utility model provides an AI visual identification automatic scraping system for a slurry drum, and relates to the technical field of polycrystalline silicon. The device is applied to a rotary drum vacuum filter for filtering polycrystalline silicon slag slurry, and comprises an identification module for identifying and judging the drum surface of a rotary drum, a DCS (Distributed Control System) and an intelligent controller which are electrically connected, an industrial camera is electrically connected with the intelligent controller, and a control module executes scraping according to the judgment result of the identification module. Comprising a PLC and a driver which are electrically connected, the scraper is connected with the driver, the industrial camera is fixed to the upper portion of the rotary drum and shoots the complete drum surface of the rotary drum, and the PLC is electrically connected with the intelligent controller. According to the utility model, AI visual identification is adopted, and the scraper can be automatically fed and withdrawn according to the slag slurry condition of the drum surface.

Description

Technical Field

[0001] This utility model relates to the field of polycrystalline silicon technology, specifically to an AI visual recognition automatic scraping system for slurry drum. Background Technology

[0002] Polysilicon slurry is an unavoidable byproduct of polysilicon production. During the slurry treatment process, a rotary drum vacuum filter is needed for dewatering. The rotary drum vacuum filter uses negative pressure as the filtration driving force and is a continuous filter where the filter surface is on the surface of a cylindrical rotating drum.

[0003] In the process of treating slurry, the slurry on the drum surface of the rotary drum vacuum filter is scraped off by a scraper. Currently, the method is for the staff to continuously observe the slurry condition on the drum surface during production and manually control the scraper to smooth it out.

[0004] However, manual observation is prone to visual fatigue and is labor-intensive and costly. Furthermore, due to differences in experience and skills among different operators, judgment standards are inconsistent, and the standards for scraper entry and exit vary. There is also the problem of human error leading to excessive entry or inaccurate timing of scraper exit. Utility Model Content

[0005] The purpose of this invention is to develop an AI visual recognition automatic scraping system for slurry drums that uses AI visual recognition to automatically advance and retract the scraper based on the slurry condition on the drum surface.

[0006] This utility model is achieved through the following technical solution:

[0007] An AI-powered visual recognition automatic scraping system for slurry rotary drum filters is provided. This system is applied to a rotary drum vacuum filter used for filtering polycrystalline silicon slurry, and includes:

[0008] The identification module identifies and judges the drum surface, including:

[0009] Electrically connected DCS and intelligent controllers;

[0010] Industrial camera, electrically connected to intelligent controller;

[0011] The control module performs scraping based on the judgment result of the recognition module, including:

[0012] Electrically connected PLC and driver;

[0013] Scraper, connected to the driver;

[0014] The industrial camera is fixed to the upper part of the drum and captures images of the entire drum surface. The PLC is electrically connected to the intelligent controller.

[0015] Optionally, the identification module may further include a light source electrically connected to a DCS, intelligent controller, or independent power supply, wherein the light source is located on the upper part of the drum.

[0016] Optionally, the light source is a surface light source or a ring light source.

[0017] Optionally, the identification module further includes a display electrically connected to the intelligent controller, the display showing the images captured by the industrial camera in real time.

[0018] Optionally, a baffle is provided at the lens of the industrial camera, and the lens of the industrial camera is located within the area covered by the rotation trajectory of the baffle.

[0019] Optionally, an I / O card is electrically connected between the PLC and the driver.

[0020] Optionally, the actuator is a linear slide, a cylinder, or an electric actuator.

[0021] The beneficial effects of this utility model are:

[0022] This invention uses AI visual recognition to automatically feed and retract the blades on the drum surface, eliminating the need for manual observation, avoiding visual fatigue, reducing labor intensity and costs, and preventing judgment differences caused by varying experience and skills among different operators, thus improving scraping effect and efficiency. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1 This is a system architecture diagram of the present invention. Detailed Implementation

[0025] In the following description, only certain exemplary embodiments are briefly described. As those skilled in the art will recognize, the described embodiments can be modified in various ways without departing from the spirit or scope of the invention. Therefore, the drawings and description are considered to be exemplary in nature and not restrictive.

[0026] The embodiments of this utility model will now be described in detail with reference to the accompanying drawings.

[0027] like Figure 1As shown, this utility model discloses an AI visual recognition automatic scraping system for slurry drum. This automatic scraping system is applied to a drum vacuum filter for filtering polycrystalline silicon slurry. It includes an identification module and a control module. The identification module reads the image of the drum surface and, based on the slurry condition on the drum surface in the image, the control module scrapes the drum surface.

[0028] The identification module includes an electrically connected DCS (Distributed Control System) and intelligent controller, as well as a light source electrically connected to the DCS and an industrial camera and display electrically connected to the intelligent controller.

[0029] An industrial camera is fixed on top of the drum. The industrial camera can capture the entire drum surface. A baffle is installed at the lens of the industrial camera. The lens of the industrial camera is within the area covered by the rotation trajectory of the baffle. The rotation of the baffle can block the lens of the industrial camera to protect it. When the industrial camera is needed, the baffle can be rotated to remove the lens, allowing the industrial camera to capture the drum surface. The display can show the captured image of the industrial camera in real time.

[0030] The light source is located on the top of the drum and is a surface light source or ring light source specifically designed for industrial machine vision, ensuring that the diatomaceous earth surface of the drum is clearly and evenly illuminated. In addition to being electrically connected to the DCS, the light source can also be electrically connected to the intelligent controller. The light source can also be powered by a separate power supply and its brightness can be controlled by a separate switch.

[0031] The signal from the industrial camera is directly connected to the intelligent controller via a network cable. The intelligent controller performs real-time identification and judgment by continuously receiving video streams.

[0032] The control module includes a PLC (Programmable Logic Controller) electrically connected to the intelligent controller. The PLC is electrically connected to an I / O card, which in turn is electrically connected to a driver. The driver is connected to the scraper. The intelligent controller performs real-time identification and judgment based on the video stream received from the industrial camera, then sends signals to the PLC. The PLC then controls the driver via the I / O card. The driver moves the scraper towards the drum surface to advance the scraper, or moves it away from the drum surface to retract the scraper. The driver can be a linear actuator such as a linear slide, cylinder, or electric push rod.

[0033] The identification of the surface smoothness of diatomaceous earth on the drum surface requires a deep learning-based segmentation model. This model needs to be trained, with the training sample set collected from on-site photos. The data samples need to be manually labeled, and a good accuracy rate can be achieved when the dataset size reaches more than 500 images. For the identification of concave and convex parts on the plane, concave and convex parts are highlighted with different colors, and parameters such as the area, location, and proportion of the observation area of ​​the concave and convex parts can be identified.

[0034] This invention uses AI visual recognition to automatically feed and retract the blades on the drum surface, eliminating the need for manual observation, avoiding visual fatigue, reducing labor intensity and costs, and preventing judgment differences caused by varying experience and skills among different operators, thus improving scraping effect and efficiency.

[0035] The above embodiments are merely preferred embodiments of this utility model and are not intended to limit the technical solutions of this utility model. Any technical solution that can be implemented based on the above embodiments without creative effort should be considered to fall within the scope of protection of this utility model patent.

Claims

1. A slurry rotary drum AI visual recognition automatic scraping system, characterized in that, This automatic scraping system is applied to a rotary drum vacuum filter for filtering polycrystalline silicon slurry, and includes: The identification module identifies and judges the drum surface, including: Electrically connected DCS and intelligent controllers; Industrial camera, electrically connected to intelligent controller; The control module performs scraping based on the judgment result of the recognition module, including: Electrically connected PLC and driver; Scraper, connected to the driver; The industrial camera is fixed to the upper part of the drum and captures images of the entire drum surface. The PLC is electrically connected to the intelligent controller.

2. The slurry drum AI visual recognition automatic scraping system according to claim 1, characterized in that, The identification module also includes a light source electrically connected to a DCS, intelligent controller, or independent power supply, and the light source is located on the upper part of the drum.

3. The slurry drum AI visual recognition automatic scraping system according to claim 2, characterized in that, The light source is a surface light source or a ring light source.

4. The slurry drum AI visual recognition automatic scraping system according to claim 1, characterized in that, The identification module also includes a display electrically connected to the intelligent controller, which displays the images captured by the industrial camera in real time.

5. The slurry drum AI visual recognition automatic scraping system according to claim 1, characterized in that, The industrial camera has a baffle that rotates around its lens, and the lens of the industrial camera is located within the area covered by the rotation trajectory of the baffle.

6. The slurry drum AI visual recognition automatic scraping system according to claim 1, characterized in that, An I / O card is electrically connected between the PLC and the driver.

7. The slurry drum AI visual recognition automatic scraping system according to claim 1, characterized in that, The actuator is a linear slide, a cylinder, or an electric push rod.

Images