A mineral processing production quality monitoring system and method based on multiple data sources

Abstract

The present invention relates to a mineral processing production quality monitoring system and method based on multiple data sources, including a raw material monitoring unit, a crushing monitoring unit, a ball mill monitoring unit, a flotation monitoring unit, a tailings monitoring unit and a concentrate monitoring unit; the quality of the mineral processing production process is monitored in a multi-data collaborative manner; in the mineral processing process, each process is evaluated and monitored for quality, the next process is predicted based on the score of the previous process, and a corresponding expected value is set. When the next process is not within the expected value range, an early warning is issued, and people are reminded to check the factors that affect the process, mechanical parameters, environmental technology, etc. The present invention specifically selects the decision parameters of each process, and obtains the decision parameters with an acquisition device. It can remotely monitor the production process of mineral processing, perform quality monitoring on each process of mineral processing, and intuitively display the production status of each process with data, so that people can understand the working conditions of each process.

Description

Technical Field

[0001] The present invention relates to the technical field of mineral processing, and in particular to a mineral processing production quality monitoring system and method based on multiple data sources. Background Art

[0002] The purpose of mineral processing is to separate useful minerals from gangue minerals in preparation for the next step of mineral processing. The main processes in the entire mineral processing process can be divided into crushing, screening, grinding, grading and selection.

[0003] The mineral processing line consists of a jaw crusher, ball mill, classifier, magnetic separator, flotation machine, thickener, and dryer. Together with a feeder, elevator, and conveyor, it forms a complete mineral processing line. This line boasts high efficiency, low energy consumption, high throughput, and economical efficiency.

[0004] The production process of the mineral processing line is as follows: the mined ore is initially crushed in a jaw crusher. After being crushed to a suitable fineness, it is evenly fed via an elevator and feeder to a ball mill, where the ore is crushed and ground. The fine ore milled in the ball mill enters the next process: classification. A spiral classifier cleans and classifies the ore mixture, leveraging the principle that solid particles with different specific gravities settle at different rates in liquid. The ore is then fed into a flotation cell, where different reagents are added based on the characteristics of the minerals to separate the desired minerals from other substances. After the desired mineral is separated, it undergoes initial concentration in a concentrator due to its high water content, followed by drying in a dryer to obtain the dry mineral.

[0005] According to this production process, the quality of each process is often manually inspected. This inspection method is inefficient and has poor accuracy, which cannot meet people's needs. As a result, the quality of mineral processing production is difficult to control and abnormalities in each process cannot be discovered in time. The quality monitoring of each process is inconvenient, untimely and has poor accuracy. Based on this, it is necessary to study a mineral processing production quality monitoring system and method based on multiple data sources. Summary of the Invention

[0006] In view of this, the purpose of the present invention is to provide a mineral processing production quality monitoring system and method based on multiple data sources, which effectively solves the problems of inconvenient, untimely and poor accuracy in the quality monitoring of each process in the existing mineral processing process.

[0007] To achieve the above objectives, the technical solution adopted by the present invention is: a mineral processing production quality monitoring system based on multiple data sources, including a raw material monitoring unit, a crushing monitoring unit, a ball milling monitoring unit, a flotation monitoring unit, a tailings monitoring unit, and a concentrate monitoring unit; the raw material monitoring unit is used to input the raw material type and call a quality monitoring strategy based on the raw material type, and the quality monitoring strategy includes a crushing monitoring strategy, a ball milling monitoring strategy, a flotation monitoring strategy, a concentrate monitoring strategy, and a tailings monitoring strategy;

[0008] The crushing monitoring unit uses images to collect the graded ore after crushing, obtains the actual crushing status through image analysis, inputs the actual crushing status into the crushing monitoring strategy, and obtains the crushing score;

[0009] The ball mill monitoring unit uses images to collect ball milled ore after ball milling, obtains the actual ball milling status through image analysis, inputs the actual ball milling status into the ball milling monitoring strategy, and obtains the ball milling score;

[0010] The flotation monitoring unit includes sensor monitoring and image monitoring; the sensor monitoring is used to collect air flow, liquid level, temperature and pH value; the image monitoring is used to obtain ore status and foam status; the data obtained by sensor monitoring and image monitoring are input into the flotation monitoring strategy to obtain the flotation score;

[0011] The concentrate monitoring unit includes sorting quantity monitoring and sorting quality monitoring. The sorting quantity monitoring uses a weighing sensor to obtain the total amount of sorted concentrate. The sorting quality monitoring uniformly collects the concentrate and analyzes the specific components of the concentrate. The content of the specific components is used as the target quality content of the concentrate. The total amount of concentrate and the target quality content are input into the concentrate monitoring strategy to obtain the concentrate score.

[0012] The tailings monitoring unit includes tailings quantity monitoring and tailings loss monitoring. The tailings quantity monitoring is used to monitor the discharge flow of tailings, and the tailings loss monitoring is used to detect the target mass content contained in the tailings. The tailings discharge quantity and target value content are input into the tailings monitoring strategy to obtain a tailings score.

[0013] Furthermore, the type of the raw material is obtained by any one of chemical analysis, spectral analysis, electron microscopy and X-ray diffraction.

[0014] Furthermore, the crushing state includes the particle size distribution and particle size coefficient after crushing; the particle size distribution indicates the proportion of particles of different sizes in the material after processing; and the particle size coefficient indicates the degree of fineness of the material.

[0015] Furthermore, the ore state after flotation includes mud height, mud quality, slurry flow and tailings flow; the foam state includes foam size, foam density and foam color.

[0016] Furthermore, the quality monitoring strategy provides a multi-task sharing strategy for the crushing monitoring strategy, ball milling monitoring strategy, flotation monitoring strategy, concentrate monitoring strategy and tailings monitoring strategy. The multi-task sharing strategy constructs a feature sharing network. Each subtask network includes a task input module and a task execution module. The input features of the subtask are sent to the corresponding task input module, and executed in the task execution module to obtain the execution result of the subtask. The value of the loss function is calculated based on the subtask network label and the execution result; the loss function is used to measure the gap between the model's predicted results and the actual results, and to optimize and update the subtasks; then the execution results are used as the input of the next subtask, and the entire serial subtask process is optimized and updated.

[0017] Furthermore, the crushing score, ball milling score, flotation score, concentrate score and tailings score are input into a one-dimensional array, and a quality score line graph is drawn in sequence. The slope from one process to the next is obtained based on the line graph, and the production quality between processes is judged based on the slope. When the slope exceeds the preset slope range, an alarm is issued.

[0018] Furthermore, a corresponding raw material score is provided for the raw material type. Starting from the raw material score, the crushing score, ball milling score, flotation score, concentrate score and tailings score are sequentially connected by broken lines. The raw material score and the slopes obtained in sequence are used as the basis for quality evaluation of each process in the mineral processing process.

[0019] A monitoring method for a mineral processing production quality monitoring system based on multiple data sources includes the following steps:

[0020] Step 1: Determine the raw material type based on the raw material information.

[0021] determining a raw material score for the raw material based on the raw material type obtained by any one of chemical analysis, spectral analysis, electron microscopy, and X-ray diffraction;

[0022] Step 2: Obtain quality reference data

[0023] Input the raw material type into the pre-trained neural network model and generate the corresponding crushing parameters, ball milling parameters, flotation parameters, concentrate parameters and tailings parameters;

[0024] Step 3: Broken Evaluation

[0025] The crushing and grading process is monitored by cameras deployed on site. The image information obtained is used to obtain the particle size distribution and particle size coefficient after crushing. The crushing score is calculated based on the particle size distribution and particle size coefficient, and then the crushing process slope is obtained. The crushing process score is evaluated based on this slope. When the score exceeds the set range, a crushing process abnormality is issued and personnel are notified to check.

[0026] Step 4: Ball milling evaluation

[0027] The on-site cameras monitor the autogenous grinding of the ore. The obtained image information is used to obtain the grinding fineness and grinding distribution of the ore after autogenous grinding. The ball milling score is calculated based on the grinding fineness and grinding distribution, and then the ball milling process slope is obtained. The ball milling process score is evaluated based on the slope. When the score exceeds the set range, the ball milling process is abnormal and personnel are notified to check.

[0028] Step 5: Flotation Evaluation

[0029] Sensors and cameras are used to monitor the flotation site and obtain the status of the ore after flotation, including mud height, mud quality, slurry flow, tailings flow, foam size, foam density, and foam color. This is used to evaluate the flotation score, and then the flotation process slope is obtained. The flotation process score is evaluated based on this slope. When the score exceeds the set range, a flotation process abnormality is reported and personnel are notified to check.

[0030] Step 6: Concentrate Evaluation

[0031] Use weighing sensors and concentrate component analyzers to obtain the concentrate sorting amount and specific component ratio, which are used to evaluate the concentrate score, and then obtain the concentrate process slope. The concentrate process score is evaluated based on this slope. When the score exceeds the set range, the concentrate process is abnormal and personnel are notified to check.

[0032] Step 7: Tailings Evaluation

[0033] The flow meter and tailings composition analyzer are used to obtain the overall flow rate and target component content of the tailings, which are then used to evaluate the tailings score, and then the tailings process slope is obtained. The tailings process score is evaluated based on the slope; when the score exceeds the set range, a tailings process abnormality is detected and personnel are notified to check.

[0034] The beneficial effects of the above technical solution are: the present invention uses the raw material monitoring unit, crushing monitoring unit, ball mill monitoring unit, flotation monitoring unit, tailings monitoring unit and concentrate monitoring unit as data sources, and monitors the quality of the mineral processing production process in a multi-data collaborative manner; in the mineral processing process, each process is evaluated and monitored for quality, and the next process is predicted based on the score of the previous process, and the corresponding expected value is set. When the next process is not within the expected value range, an early warning is issued, and people are reminded to check the factors that affect the process, such as process, mechanical parameters, environmental technology, etc., to ensure the smooth and efficient progress of mineral processing, and to quickly detect abnormalities and remind staff.

[0035] In the present invention, multiple data sources participate in quality monitoring at the same time, the quality of the ore is displayed with a numerical score, and each process in the mineral processing process is divided into multiple partitions according to the main function. The processing results of each partition are displayed with a numerical score, which can more intuitively display the processing quality of each process, making it easier for people to monitor the various processes of mineral processing. At the same time, based on the changes in each process, the abnormal points in the process processing are judged, which makes it easier for people to discover abnormalities; and after the abnormality occurs, the subsequent process is judged by the slope, and continuous abnormalities will not occur.

[0036] In the process of each process, the main determining factors of the production process are selected, and the score of each process is determined by data association. The determining factors of each process are mainly collected by equipment, which can reflect the mineral processing situation of each process in real time. The measurement speed is fast, the data is easy to obtain, and it is convenient to monitor the mineral processing process.

[0037] Therefore, the present invention targets the process flow of mineral processing, specifically selects the decision parameters of each process, and obtains the decision parameters with acquisition equipment. It can remotely monitor the production process of mineral processing, perform quality monitoring on each process of mineral processing, and intuitively display the production status of each process with data, so that people can understand the working conditions of each process, quickly discover anomalies, and notify relevant personnel to carry out maintenance and related process adjustments, thereby greatly improving the quality of mineral processing. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 is a system flow chart of the present invention;

[0039] Figure 2 Schematic diagram of each strategy of the present invention;

[0040] Figure 3 The flowchart of the monitoring method of the mineral processing production quality monitoring system based on multiple data sources. DETAILED DESCRIPTION

[0041] The present invention is further described in detail below with reference to the accompanying drawings and specific embodiments:

[0042] Example 1. This embodiment aims to provide a mineral processing production quality monitoring system and method based on multiple data sources, which is mainly used for quality monitoring of mineral processing. In view of the problems of inconvenience, untimeliness and poor accuracy in quality monitoring of each process in the mineral processing process of the existing technology, this embodiment provides a mineral processing production quality monitoring system based on multiple data sources for the mineral processing process.

[0043] like Figure 1The present invention shows a mineral processing production quality monitoring system based on multiple data sources, including a raw material monitoring unit, a crushing monitoring unit, a ball mill monitoring unit, a flotation monitoring unit, a tailings monitoring unit and a concentrate monitoring unit. This embodiment provides corresponding monitoring units for each process of the mineral processing, and the monitoring units are interconnected, and multiple data are used to jointly monitor the mineral processing process to ensure the quality of the mineral processing.

[0044] The raw material monitoring unit is used to input the raw material type, which is obtained through any of chemical analysis, spectral analysis, electron microscopy and X-ray diffraction. The quality monitoring strategy is called according to the raw material type. The quality monitoring strategy includes crushing monitoring strategy, ball milling monitoring strategy, flotation monitoring strategy, concentrate monitoring strategy and tailings monitoring strategy; the quality monitoring strategy provides a multi-task sharing strategy for the crushing monitoring strategy, ball milling monitoring strategy, flotation monitoring strategy, concentrate monitoring strategy and tailings monitoring strategy. The multi-task sharing strategy constructs a feature sharing network. Each subtask network includes a task input module and a task execution module. The input features of the subtask are sent to the corresponding task input module, and the task is executed in the task execution module to obtain the execution result of the subtask. The value of the loss function is calculated based on the subtask network label and the execution result. The loss function is used to measure the gap between the model's predicted results and the actual results, and to optimize and update the subtasks; then the execution result is used as the input of the next subtask, and the entire serial subtask process is optimized and updated.

[0045] Specifically, the crushing monitoring unit uses images to collect and grade ore after crushing, obtains the actual crushing status through image analysis, inputs the actual crushing status into the crushing monitoring strategy, and obtains the crushing score; the crushing status includes the particle size distribution and particle size coefficient after crushing; the particle size distribution indicates the proportion of particles of different sizes in the material after processing; the particle size coefficient indicates the degree of fineness of the material. In specific implementation, the particle size distribution is defined as D1, the particle size coefficient is D2, and the crushing score is Y, then Y=aD1+bD2, where a and b are correlation coefficients, which are used to determine the crushing score.

[0046] The ball mill monitoring unit uses images to capture ball-milled ore after ball milling, obtains the actual ball milling state through image analysis, inputs the actual ball milling state into the ball mill monitoring strategy, and obtains the ball milling score; monitors the ore grinding according to the camera arranged on site, uses the obtained image information to obtain the grinding fineness and grinding distribution of the ore after self-grinding, and calculates the ball milling score according to the grinding fineness and grinding distribution. The grinding fineness and grinding distribution of the ore after self-grinding are defined as K1 and K2 respectively, and the ball milling score is Q, Q=aK1+bK2, where a and b are correlation coefficients, and the ball milling score is determined by this.

[0047] In the existing technology, there is an image analysis type particle size distribution measuring instrument. This instrument is an existing technology. It only needs to input the conditions in the read image to automatically identify particles and quickly and accurately measure the particle size distribution, shape coefficient, etc., and then obtain the size and distribution of the particles.

[0048] The flotation monitoring unit includes sensor monitoring and image monitoring; sensor monitoring is used to collect air flow, liquid level, temperature and pH value; image monitoring is used to obtain ore status and foam status; the data obtained by sensor monitoring and image monitoring are input into the flotation monitoring strategy to obtain a flotation score; the ore status after flotation includes mud height, mud quality, slurry flow and tailings flow; the foam status includes foam size, foam density and foam color; in this embodiment, the sensor is mainly used to monitor changes in flotation conditions, and image monitoring is used to monitor flotation changes of the ore, and the ore status after flotation includes mud height, mud quality, slurry flow and tailings flow, and the post-flotation score is obtained based on the mud height, mud quality, slurry flow and tailings flow.

[0049] Flotation froth quality: Observe the quality and stability of the flotation froth. Good froth generally indicates effective flotation separation. Tailings discharge: Check the concentration and residual amount of target minerals in the tailings discharge. Lower tailings concentration and residual amount generally indicate better flotation results. Ore recovery: Evaluate the flotation process's recovery rate of the target ore by comparing the target mineral content in the feed with the target mineral content in the flotation product. Ore grading: Observe the grading of the ore in the flotation product to ensure the desired flotation product grade.

[0050] During flotation, the recovery rate of the target mineral is calculated using the following formula: Recovery equals the mass of the target mineral in the feed divided by the mass of the target mineral in the flotation product, multiplied by 100%. Grade is calculated using the following formula: Grade equals the mass of the target mineral in the flotation product divided by the total mass of the flotation product, multiplied by 100%. The flotation index, a measure of flotation performance, is calculated by comparing the mass of the target mineral in the flotation product with the mass of the target mineral in the tailings. The above parameters are weighted and used to calculate the flotation score.

[0051] The concentrate monitoring unit includes sorting quantity monitoring and sorting quality monitoring. The sorting quantity monitoring uses a weighing sensor to obtain the total amount of sorted concentrate. The sorting quality monitoring evenly collects the concentrate and analyzes the specific components of the concentrate. The content of the specific components is used as the target quality content of the concentrate. The total amount of concentrate and the target quality content are input into the concentrate monitoring strategy to obtain the concentrate score.

[0052] The tailings monitoring unit includes tailings quantity monitoring and tailings loss monitoring. The tailings quantity monitoring is used to monitor the discharge flow of tailings, and the tailings loss monitoring is used to detect the target mass content contained in the tailings. The tailings discharge quantity and target value content are input into the tailings monitoring strategy to obtain the tailings score.

[0053] The crushing score, ball milling score, flotation score, concentrate score and tailings score are input into a one-dimensional array, and a quality score line graph is drawn in sequence. The slope from one process to the next is obtained based on the line graph, and the production quality between processes is judged based on the slope. When the slope exceeds the preset slope range, an alarm is issued.

[0054] Provide a corresponding raw material score for the raw material type. Starting from the raw material score, the crushing score, ball milling score, flotation score, concentrate score and tailings score are connected in sequence through broken lines. The raw material score and the slopes obtained in sequence are used as the basis for quality evaluation of each process in the mineral processing process.

[0055] The present invention uses a raw material monitoring unit, a crushing monitoring unit, a ball mill monitoring unit, a flotation monitoring unit, a tailings monitoring unit and a concentrate monitoring unit as data sources, and performs quality monitoring on the ore dressing production process in a multi-data collaborative manner; in the ore dressing process, quality evaluation and monitoring are performed on each process, and the next process is predicted based on the score of the previous process, and a corresponding expected value is set. When the next process is not within the expected value range, an early warning is issued, and people are reminded to check factors that affect the process, such as the process, mechanical parameters, and environmental processes, to ensure the smooth and efficient conduct of ore dressing, quickly discover abnormalities, and remind staff.

[0056] A monitoring method for a mineral processing production quality monitoring system based on multiple data sources includes the following steps:

[0057] Step 1: Determine the raw material type based on the raw material information

[0058] determining a raw material score for the raw material based on the raw material type obtained by any one of chemical analysis, spectral analysis, electron microscopy, and X-ray diffraction;

[0059] Step 2: Obtain quality reference data

[0060] Input the raw material type into the pre-trained neural network model and generate the corresponding crushing parameters, ball milling parameters, flotation parameters, concentrate parameters and tailings parameters;

[0061] Step 3: Broken Evaluation

[0062] The crushing and grading process is monitored by cameras deployed on site. The image information obtained is used to obtain the particle size distribution and particle size coefficient after crushing. The crushing score is calculated based on the particle size distribution and particle size coefficient, and then the crushing process slope is obtained. The crushing process score is evaluated based on this slope. When the score exceeds the set range, a crushing process abnormality is issued and personnel are notified to check.

[0063] Step 4: Ball milling evaluation

[0064] The on-site cameras monitor the autogenous grinding of the ore. The obtained image information is used to obtain the grinding fineness and grinding distribution of the ore after autogenous grinding. The ball milling score is calculated based on the grinding fineness and grinding distribution, and then the ball milling process slope is obtained. The ball milling process score is evaluated based on the slope. When the score exceeds the set range, the ball milling process is abnormal and personnel are notified to check.

[0065] Step 5: Flotation Evaluation

[0066] Sensors and cameras are used to monitor the flotation site and obtain the status of the ore after flotation, including mud height, mud quality, slurry flow, tailings flow, foam size, foam density, and foam color. The flotation score is evaluated in sequence to obtain the flotation process slope, and the flotation process score is evaluated based on the slope. When the score exceeds the set range, a flotation process abnormality is reported and personnel are notified to check.

[0067] Step 6: Concentrate Evaluation

[0068] Use weighing sensors and concentrate component analyzers to obtain the concentrate sorting amount and specific component ratio, which are used to evaluate the concentrate score, and then obtain the concentrate process slope. The concentrate process score is evaluated based on this slope. When the score exceeds the set range, the concentrate process is abnormal and personnel are notified to check.

[0069] Step 7: Tailings Evaluation

[0070] The flow meter and tailings composition analyzer are used to obtain the overall flow rate and target component content of the tailings, which are then used to evaluate the tailings score, and then the tailings process slope is obtained. The tailings process score is evaluated based on the slope; when the score exceeds the set range, a tailings process abnormality is detected and personnel are notified to check.

[0071] The present invention targets the process flow of mineral processing, specifically selects the determining parameters of each process, and obtains the determining parameters with acquisition equipment. It can remotely monitor the production process of mineral processing, perform quality monitoring on each process of mineral processing, and visually display the production status of each process with data, so that people can understand the working conditions of each process, quickly discover anomalies, and notify relevant personnel to carry out maintenance and related process adjustments, thereby greatly improving the quality of mineral processing.

Claims

1. A mineral processing production quality monitoring system based on multiple data sources, characterized by: It includes a raw material monitoring unit, a crushing monitoring unit, a ball milling monitoring unit, a flotation monitoring unit, a tailings monitoring unit and a concentrate monitoring unit; the raw material monitoring unit is used to input the raw material type and call the quality monitoring strategy according to the raw material type. The quality monitoring strategy includes a crushing monitoring strategy, a ball milling monitoring strategy, a flotation monitoring strategy, a concentrate monitoring strategy and a tailings monitoring strategy; The crushing monitoring unit uses images to collect the graded ore after crushing, obtains the actual crushing status through image analysis, inputs the actual crushing status into the crushing monitoring strategy, and obtains the crushing score; The ball mill monitoring unit uses images to collect ball milled ore after ball milling, obtains the actual ball milling status through image analysis, inputs the actual ball milling status into the ball milling monitoring strategy, and obtains the ball milling score; The flotation monitoring unit includes image monitoring; the image monitoring is used to obtain the ore state and foam state; the data obtained by the image monitoring is input into the flotation monitoring strategy to obtain the flotation score; The concentrate monitoring unit includes sorting quantity monitoring and sorting quality monitoring. The sorting quantity monitoring uses a weighing sensor to obtain the total amount of sorted concentrate. The sorting quality monitoring is used to uniformly collect the concentrate and analyze the specific components of the concentrate. The content of the specific components is used as the target quality content of the concentrate. The total amount of concentrate and the target quality content are input into the concentrate monitoring strategy to obtain the concentrate score. The tailings monitoring unit includes tailings quantity monitoring and tailings loss monitoring. The tailings quantity monitoring is used to monitor the discharge flow of the tailings, and the tailings loss monitoring is used to detect the target quality content contained in the tailings, and the tailings discharge quantity and target value content are input into the tailings monitoring strategy to obtain a tailings score; the quality monitoring strategy provides a multi-task sharing strategy for the crushing monitoring strategy, the ball milling monitoring strategy, the flotation monitoring strategy, the concentrate monitoring strategy and the tailings monitoring strategy. The multi-task sharing strategy constructs a feature sharing network. Each subtask network includes a task input module and a task execution module. The input features of the subtask are sent to the corresponding task input module, and the task is executed in the task execution module to obtain the execution result of the subtask. The value of the loss function is calculated according to the subtask network label and the execution result; The loss function is used to measure the gap between the model's predictions and actual results, and to optimize and update the subtasks. The execution results are then used as input for the next subtask, and the entire chain of subtasks is optimized and updated. The crushing score, ball milling score, flotation score, concentrate score, and tailings score are input into a one-dimensional array, and a quality score line graph is drawn in sequence. The slope from one process to the next is obtained from the line graph, and the production quality between processes is judged based on the slope. When the slope exceeds the preset slope range, an alarm is issued; Provide a corresponding raw material score for the raw material type. Starting from the raw material score, the crushing score, ball milling score, flotation score, concentrate score and tailings score are connected in sequence through broken lines. The raw material score and the slopes obtained in sequence are used as the basis for quality evaluation of each process in the mineral processing process.

2. The mineral processing production quality monitoring system based on multiple data sources according to claim 1 is characterized in that: The type of raw material is obtained by any of chemical analysis, spectroscopic analysis, electron microscopy and X-ray diffraction.

3. The mineral processing production quality monitoring system based on multiple data sources according to claim 1 is characterized in that: The crushing state includes the particle size distribution and particle size coefficient after crushing; the particle size distribution indicates the proportion of particles of different sizes in the material after processing; the particle size coefficient indicates the degree of fineness of the material.

4. The mineral processing production quality monitoring system based on multiple data sources according to claim 1 is characterized in that: The ore state includes mud height, mud quality, slurry flow and tailings flow; the foam state includes foam size, foam density and foam color.

5. The mineral processing production quality monitoring system based on multiple data sources according to claim 1 is characterized in that: The flotation monitoring unit also includes sensor monitoring; the sensor monitoring is used to collect air flow, liquid level, temperature and pH value; the data obtained by sensor monitoring and image monitoring are input into the flotation monitoring strategy to obtain the flotation score.

6. The monitoring method of the mineral processing production quality monitoring system based on multiple data sources according to claim 1 is characterized in that: The method includes the following steps: Step 1: Determine the raw material type based on the raw material information. determining a raw material score for the raw material based on the raw material type obtained by any one of chemical analysis, spectral analysis, electron microscopy, and X-ray diffraction; Step 2: Obtain quality reference data Input the raw material type into the pre-trained neural network model and generate the corresponding crushing parameters, ball milling parameters, flotation parameters, concentrate parameters and tailings parameters; Step 3: Broken Evaluation The crushing and grading process is monitored by cameras deployed on site. The image information obtained is used to obtain the particle size distribution and particle size coefficient after crushing. The crushing score is calculated based on the particle size distribution and particle size coefficient, and then the crushing process slope is obtained. The crushing process score is evaluated based on this slope. When the score exceeds the set range, a crushing process abnormality is issued and personnel are notified to check. Step 4: Ball milling evaluation The on-site cameras monitor the autogenous grinding of the ore. The obtained image information is used to obtain the grinding fineness and grinding distribution of the ore after autogenous grinding. The ball milling score is calculated based on the grinding fineness and grinding distribution, and then the ball milling process slope is obtained. The ball milling process score is evaluated based on the slope. When the score exceeds the set range, the ball milling process is abnormal and personnel are notified to check. Step 5: Flotation Evaluation Sensors and cameras are used to monitor the flotation site and obtain the status of the ore after flotation, including mud height, mud quality, slurry flow, tailings flow, foam size, foam density, and foam color. This is used to evaluate the flotation score, and then the flotation process slope is obtained. The flotation process score is evaluated based on this slope. When the score exceeds the set range, a flotation process abnormality is reported and personnel are notified to check. Step 6: Concentrate Evaluation Use weighing sensors and concentrate component analyzers to obtain the concentrate sorting amount and specific component ratio, which are used to evaluate the concentrate score, and then obtain the concentrate process slope. The concentrate process score is evaluated based on this slope. When the score exceeds the set range, the concentrate process is abnormal and personnel are notified to check. Step 7: Tailings Evaluation The flow meter and tailings composition analyzer are used to obtain the overall flow rate and target component content of the tailings, which are then used to evaluate the tailings score, and then the tailings process slope is obtained. The tailings process score is evaluated based on the slope; when the score exceeds the set range, a tailings process abnormality is detected and personnel are notified to check.

Images