CN116213087A

CN116213087A - Intelligent metal detection method applied to mineral separation

Info

Publication number: CN116213087A
Application number: CN202211464758.1A
Authority: CN
Inventors: 凌云海; 王烨江; 赵建东
Original assignee: Rishangsheng New Building Material Design And Research Institute Co ltd
Current assignee: Rishangsheng New Building Material Design And Research Institute Co ltd
Priority date: 2022-11-22
Filing date: 2022-11-22
Publication date: 2023-06-06

Abstract

The invention provides an intelligent metal detection method applied to mineral separation, which comprises the following steps: shooting a large number of pictures of normal ore conveying through a camera with the front end erected on belt conveying equipment, manually adding pictures of common metal articles, and obtaining picture information; classifying and labeling the picture information, acquiring characteristic information and inputting the characteristic information into an intelligent analysis algorithm model; when the belt conveying equipment conveys ores, the image information acquired in real time is input into a trained intelligent analysis algorithm model, and when the ores are considered to contain metal objects, the system generates an alarm signal, the belt conveying equipment is stopped, and a detection picture and a video are popped up. The beneficial effects of the invention are as follows: the method realizes the detection of the metal object by means of the image, thereby replacing a metal detector, avoiding the iron passing accident of the crusher, ensuring the safety of equipment and post operators, guaranteeing the yield, improving the economic benefit of enterprises and solving a weakness problem in the aggregate production industry.

Description

Intelligent metal detection method applied to mineral separation

Technical Field

The invention relates to the field of recognition of mineral separation metal impurities, in particular to an intelligent metal detection method applied to mineral separation.

Background

At present, aggregate industry is concentrated from a scattered trend and is subject to normal construction exploitation from a non-treatment random mining trend. The production capacity of the existing new aggregate production line is higher and higher, and the automation degree is higher and higher. However, in the raw material iron removal link in the aggregate production line, when various metal blocks are mixed in the raw material, the conventional metal detector equipment is used for detection. Thus, not only is the false alarm rate high, but also some alloy blocks are difficult to detect. This may cause damage to plant equipment, an increase in the workload of plant personnel, a decrease in the yield of products, etc., and even a safety accident of metal block splashing. How to effectively and rapidly detect metal objects has become one of the bottleneck problems of the automation process of the aggregate production line.

Existing aggregate production lines all use metal detectors (metal detectors) to protect the crusher and prevent over-iron (large pieces of metal from falling into the crusher, possibly causing failure of the crusher). A metal detector is a widely used detector. The accuracy and reliability of metal detectors depends on the stability of the electromagnetic transmitter frequency, typically using operating frequencies from 80to 800 khz. The lower the working frequency is, the better the detection performance of iron is; the higher the operating frequency, the better the detection performance for high carbon steel. The sensitivity of the detector decreases with increasing detection range, and the magnitude of the sensing signal depends on the metal particle size and conductivity properties. In the ore raw materials of the aggregate production line, the main objects detected by the metal detector are iron wear-resistant parts worn or dropped when the ore is coarsely crushed, iron bolts, iron weldments dropped when overhauled and the like, and most of the foreign matters are iron products. Therefore, the frequency of the metal detector tends to be lowered, which reduces the sensitivity to non-ferromagnetic species (stainless steel species). In actual production, there are few shovel teeth of mine diggers, and the material of the shovel teeth is mainly high manganese steel. At this time, the metal detector often can not normally detect and report to the police, just take place on the rich sun production line of certain company that the forming relieved tooth is undetected, has got into the breaker, and the breaker can't normally discharge, and this forming relieved tooth is extruded by the crushing chamber and launched, directly pricks into the concrete column next to the breaker, and the bad point causes casualties accident.

In addition, the metal detector also has a certain limit on the conveying speed of the detected objects due to the pulsation of the current and the current filtering. If the transport speed exceeds a reasonable range, the sensitivity of the detector is reduced. In aggregate production, belt speeds of belt conveyors are increasingly faster, and some belt speeds are already close to 4 meters/second. The sensitivity of the metal detector is also reduced.

In summary, the metal detector is not suitable for detecting iron impurities in raw materials in automatic aggregate production.

Disclosure of Invention

The invention aims to overcome the defects of the prior art, and provides an intelligent metal detection method applied to mineral separation, which realizes intelligent metal detection, avoids iron passing accidents of a crusher, protects equipment and eliminates one production safety hidden trouble.

The aim of the invention is achieved by the following technical scheme. An intelligent metal detection method applied to mineral separation comprises the following steps:

(1) Shooting a large number of pictures of normal ore conveying through a camera with the front end erected on belt conveying equipment, manually adding pictures of common metal articles, and obtaining picture information;

(2) Classifying and marking the picture information, acquiring characteristic information and inputting the characteristic information into an intelligent analysis algorithm model of the rear-end camera AI system to train the intelligent analysis algorithm model;

(3) When the belt conveying equipment conveys ores, the image information acquired in real time is input into a trained intelligent analysis algorithm model, when the ores are considered to contain metal objects, the system generates an alarm signal, the belt conveying equipment is stopped, a detection picture and a video are popped up, and the central control operator confirms in time;

(4) If false alarm is considered, the alarm signal is canceled, otherwise, the process of iron removal through the electromagnetic iron remover is entered.

Furthermore, after the intelligent analysis algorithm model is loaded, real-time video analysis, channel polling analysis, channel capture analysis, platform issuing picture analysis and channel polling capture analysis are performed by combining a large amount of data.

Furthermore, when an alarm signal appears, relevant data information of the alarm signal is input into a database to be used as the basis of the next judgment.

The beneficial effects of the invention are as follows:

1. the invention realizes the detection of metal objects by means of images, thereby replacing a metal detector, avoiding the iron passing accident of a crusher, ensuring the safety of equipment and post operators, ensuring the yield, improving the economic benefit of enterprises and solving a weakness problem in the aggregate production industry.

2. Can be suitable for various raw material types, and is suitable for foreign matter detection alarm under static or moving state. The system has high automation degree and does not need manual intervention. And as the length of use increases, the amount of data processed increases, and the sensitivity increases. Reduces the workload of post workers and improves the production efficiency.

Drawings

FIG. 1 is a block schematic diagram of the present invention.

Detailed Description

The invention will be described in detail below with reference to the attached drawings:

as shown in fig. 1, an intelligent metal detection method applied to mineral separation comprises the following steps:

(2) Classifying and marking the picture information, acquiring characteristic information and inputting the characteristic information into an intelligent analysis algorithm model of the rear-end camera AI system to train the intelligent analysis algorithm model; and after the intelligent analysis algorithm model is loaded, carrying out real-time video analysis, channel polling analysis, channel capture analysis, platform issuing picture analysis and channel polling capture analysis by combining a large amount of data.

(4) If false alarm is considered, the alarm signal is canceled, otherwise, the process of iron removal through the electromagnetic iron remover is entered. When an alarm signal appears, relevant data information of the alarm signal is input into a database to be used as a basis for next judgment.

The innovation of the invention mainly adopts novel video detection equipment across boundaries, namely an iDS-6700NX/X-AI series sea health superbrain system. The AI series sea-health superbrain system adopts an embedded design, integrates a high-performance GPU module, integrates IPC access, storage, management and control, can load a deep learning algorithm, realizes accurate self-defined intelligent analysis, improves the monitoring video value and serves the big data era. By utilizing the intelligent analysis of the system, a large number of ore raw material pictures on the belt conveyor can be compared, so that metal objects mixed in the ore raw material pictures can be effectively distinguished, an alarm is given in time, and the next processing is carried out.

The processing principle is as follows: firstly, high-definition IPC (intelligent camera) with the front end erected on belt conveying equipment shoots a large number of pictures of normal ore conveying, and a superbrain system obtains primary memory. Then, common metal articles (iron wear-resistant parts, iron bolts, iron weldments which fall during maintenance and the like) are manually added, so that the superbrain system obtains the memory of the object needing to be alarmed. This process requires a certain time and accumulation of a certain amount of data. And then downloading an intelligent analysis algorithm model from an fluorite AI open platform or an industry integrated application platform, and carrying out real-time video analysis, channel polling analysis, channel capture analysis, platform issuing picture analysis and channel polling capture analysis by combining a large amount of data just mentioned after loading the model. So far, the whole system is basically erected and can be put into operation.

In the initial operation stage and the debugging stage, the intelligent system can set the system to perform alarm action only, and after the intelligent system considers that the raw materials contain metal foreign matters, the intelligent system automatically pops up a detection picture and a video, and a central control operator confirms the detection picture and the video in time. If false alarm is considered, the alarm signal is cancelled, otherwise, the next iron removal process is carried out, and the iron removal process generally adopts equipment such as an electromagnetic iron remover and the like. After the whole action is finished, the intelligent superbrain system can automatically enter the current event into the database of the intelligent superbrain system and serve as the basis of the number of next judgment. This is why the longer the run time, the higher the system detection accuracy is, as mentioned above.

The identification system actually utilizes the camera AI identification function which is developed continuously and iterated in recent years, builds a model through an AI algorithm, continuously optimizes and improves the model, and plays a powerful function in industrial production. The camera technology is utilized to serve a factory production line of modern intelligent manufacturing. The graphical operation interface is adopted, so that the requirement on the operators of factories is not high. The original camera system of the factory is utilized, the factory is slightly modified, and the cost is low.

It should be understood that equivalents and modifications to the technical scheme and the inventive concept of the present invention should fall within the scope of the claims appended hereto.

Claims

1. An intelligent metal detection method applied to mineral separation is characterized in that: the method comprises the following steps:

2. The intelligent metal detection method applied to mineral separation according to claim 1, wherein the intelligent metal detection method is characterized in that: and after the intelligent analysis algorithm model is loaded, carrying out real-time video analysis, channel polling analysis, channel capture analysis, platform issuing picture analysis and channel polling capture analysis by combining a large amount of data.

3. The intelligent metal detection method applied to mineral separation according to claim 1, wherein the intelligent metal detection method is characterized in that: when an alarm signal appears, relevant data information of the alarm signal is input into a database to be used as a basis for next judgment.