Flotation ash content detection device and detection method

Abstract

The embodiment of the invention provides a flotation ash content detection device and method, and the device comprises a box body which comprises a control chamber and a detection chamber; the control host is mounted in a control room; the observation pool is mounted in the detection chamber and is connected with a feeding pipe group; the light supplementing lamp is installed in the detection chamber, is in signal connection with the control host and comprises a top light source and a lateral light source, the top light source is located above the observation pool and faces the observation pool, and the lateral light source is located on the side portion of the observation pool and faces the observation pool; the camera is mounted in the detection chamber and faces the observation pool, and the camera is in signal connection with the control host; and the illumination sensor is mounted in the detection chamber and is in signal connection with the control host. A black box environment (namely a detection room) is defined by the box body, so that the influence of ambient light on a shot picture is isolated; meanwhile, a light supplementing lamp is arranged in the detection chamber to provide a constant light source environment, the definition and accuracy of images collected by the camera can be ensured, and therefore the accuracy of ash content detection is improved.

Description

Technical Field

[0001] The present application relates to the technical field of mining processing equipment, and in particular to a flotation ash detection device and a detection method. Background Art

[0002] Coal slime flotation is one of the important means of sorting clean coal. The effect of flotation directly affects the quality and production efficiency of clean coal. However, the evaluation of traditional flotation effect mainly relies on the experience of on-site operators, with a low degree of automation, making it difficult to conduct real-time control of the flotation process. Visual inspection technology is used for tailings inspection in some production lines, but the current time detection machines are usually affected by the lighting environment, and shadows, reflections, and other problems will appear during the image acquisition process, which in turn affects the accuracy of the tailings ash detection device. Summary of the invention

[0003] The purpose of the embodiments of the present application is to provide a flotation ash detection device and a detection method, which can solve the problem of low detection accuracy caused by the influence of the lighting environment in the existing peat flotation tailings visual detection machine.

[0004] The present application provides a flotation ash detection device, comprising:

[0005] The box body includes a control room and a detection room;

[0006] A control host, installed in the control room;

[0007] An observation pool is installed in the detection chamber and is connected to a feed pipe group;

[0008] A fill light, installed in the detection room and connected to the control host signal, comprising a top light source and a side light source, wherein the top light source is located above the observation pool and faces the observation pool, and the side light source is located at the side of the observation pool and faces the observation pool;

[0009] A camera, installed in the detection chamber and facing the observation pool, the camera being connected to the control host by signal;

[0010] The light sensor is installed in the detection room and is connected to the control host by signal.

[0011] In the above implementation process, when the detection work is carried out, the tailings solution to be detected is input into the observation pool through the feed pipe group. After turning on the fill light to illuminate, the light sensor monitors the light intensity in the detection room in real time, and transmits the data to the control host. The control host adjusts the brightness of the fill light according to the light intensity data to ensure that the tailings solution in the observation pool can obtain a stable and clear image when the camera is shooting. The camera shoots the image of the tailings solution in the observation pool and transmits the image data to the control host. The control host processes and analyzes the received image data, detects the ash content in the tailings solution through image recognition technology, and then the control host outputs the analysis results to the display screen or other output devices for the operator to view and record. Based on the flotation ash detection device of this embodiment, a black box environment (i.e., a detection room) is formed by using a box body to isolate the influence of ambient light on the photographed photos; at the same time, a fill light is set in the detection room to provide a constant light source environment, which can ensure the clarity and accuracy of the image collected by the camera, thereby improving the accuracy of ash detection. The entire detection process is controlled by the control host, without manual intervention, realizing automatic detection and improving production efficiency. The combination of light sensors and fill lights can monitor and adjust the light intensity in the test room in real time, ensuring that the light environment for testing can always remain at a constant value, improving the stability of the testing device. In addition, the control host and testing components are installed in the box, which has a compact structure and is easy to operate and maintain.

[0012] In some embodiments, the observation pool includes a buffer pool portion and an observation pool portion which are connected at the bottom, the feeding pipe group is connected to the buffer pool portion, and the fill light and the camera are respectively arranged toward the observation pool portion.

[0013] In the above implementation process, the bottom of the buffer tank section and the observation tank section are connected, allowing the tailings solution to first enter the buffer tank section and then naturally flow into the observation tank section. The main function of the buffer tank section is to receive the tailings solution from the feed pipe group and play a role of buffering and preliminary precipitation, which helps to reduce image blur or instability caused by direct impact or rapid flow, thereby improving the quality of subsequent image acquisition. The observation tank section is the core area of ​​image acquisition, and the fill light and camera are set towards here.

[0014] In some embodiments, the detection chamber is also equipped with an aggregate funnel located at the bottom of the observation pool, and the aggregate funnel is connected to a discharge pipe; and the downward projection of the observation pool falls completely into the aggregate funnel, so that the tailings solution overflowing from the edge of the observation pool can fall into the aggregate funnel.

[0015] In the above implementation process, the downward projection of the observation pool is completely within the range of the aggregate funnel, which means that the tailings solution overflowing from the edge of the observation pool can directly fall into the aggregate funnel and be discharged and recovered through the discharge pipe connected to the aggregate funnel. Therefore, the design of the aggregate funnel ensures that the tailings solution overflowing from the edge of the observation pool will not flow to other parts of the detection room, but will be collected by the aggregate funnel and uniformly discharged and recovered. By collecting and discharging the tailings solution in time, the aggregate funnel helps to keep the observation pool clean and stable, avoiding possible contamination or damage.

[0016] In some embodiments, the top of the aggregate hopper is provided with a mesh arranged around the circumference of the observation pool portion.

[0017] In the above implementation process, when the tailings solution overflows from the edge of the observation pool, the solution will first contact the sieve. Since the mesh size of the sieve is moderate, it can effectively intercept solid particles with larger diameters in the solution, thereby preventing large particles from being fixed and falling into the discharge pipe, causing blockage.

[0018] In some embodiments, a scraping mechanism is further provided in the detection chamber, and the scraping mechanism includes a connected driver and a scraper, and the scraper extends into the buffer pool portion. The scraper can be driven by the driver to move in the buffer pool portion to scrape off the foam on the surface of the tailings solution in the buffer pool portion.

[0019] In the above implementation process, during the flotation process, under the action of the liquid medicine, a layer of foam may form on the surface of the tailings solution, which will interfere with the image acquisition of the camera, thereby affecting the accuracy of ash detection. The scraping mechanism drives the scraper to move in the buffer tank through the driver, and the scraper contacts the tailings solution and scrapes off the foam. By scraping off the foam on the surface of the tailings solution, the scraping mechanism reduces the interference of the foam on the camera image acquisition, thereby improving the accuracy of ash detection.

[0020] In some embodiments, a buffer pool cover is provided on the top side of the buffer pool portion, and a avoidance hole corresponding to the moving track of the scraper is provided on the buffer pool cover, and the scraper passes through the buffer pool cover through the avoidance hole.

[0021] In the above implementation process, the buffer tank cover is installed on the top side of the buffer tank part to close and protect the buffer tank part, which can prevent external impurities from entering the buffer tank part and reduce the evaporation and splashing of the tailings solution. The avoidance hole is cleverly designed on the buffer tank cover and corresponds to the moving track of the scraper, so that the scraper can move along the avoidance hole without being hindered.

[0022] In some embodiments, a device holder is provided in the detection chamber, the top end of the device holder is fixed to the top of the detection chamber, and the top light source, the camera, the light sensor and the bubble scraping mechanism are respectively installed on the device holder.

[0023] In the above implementation process, the device bracket provides stable support and fixation for key components, reducing the risk of displacement or damage caused by vibration or external force, which improves the stability and reliability of the entire detection system. By rationally arranging and installing key components, the device bracket helps to simplify the structure of the equipment, improve the compactness and aesthetics of the equipment, and at the same time, it also makes operation and maintenance more convenient and efficient. Since the device bracket can ensure that each key component maintains the correct position and posture, it helps to improve the accuracy and reliability of detection.

[0024] In some embodiments, the feed pipe group includes a tee, a feed connection pipe and a water inlet connection pipe, the tee includes a first interface, a second interface and a third interface, the first interface is connected to the feed connection pipe, the second interface is connected to the water inlet connection pipe, and the third interface is connected to the observation pool. The water inlet connection pipe is provided, and before the detection, the water inlet connection pipe can be opened to pass clean water to rinse the observation pool, and then the tailings solution is passed.

[0025] In the above implementation process, the above embodiment can rinse and clean the observation pool before detection by introducing a water inlet connecting pipe, which helps to reduce the impact of residues and impurities on the detection results, thereby improving the accuracy and reliability of the detection.

[0026] In some embodiments, an electric water valve is provided in the water inlet connecting pipe.

[0027] In the above implementation process, the electric water valve is connected to the control host through a signal line, and the control host can send a control signal to the electric water valve to control its opening and closing state. Through the signal connection between the electric water valve and the control host, the automatic control of the opening and closing of the water inlet connection pipe is realized, which improves the convenience and accuracy of the operation and reduces the need for manual intervention.

[0028] In some embodiments, the side walls and / or top walls and / or bottom walls of the control room are provided with a first cooling fan; the side walls and / or top walls and / or bottom walls of the detection room are provided with a second cooling fan.

[0029] In the above implementation process, when the equipment is working, a certain amount of heat will be generated in the control room and the detection room. If this heat is not discharged in time, it will cause the temperature of the equipment to rise, thereby affecting its performance and stability. By setting up a cooling fan, effective air convection can be formed to discharge heat from the room. Through effective heat dissipation, the temperature in the control room and the detection room can be kept within a suitable range, thereby improving the stability and reliability of the equipment, reducing damage or failure of the equipment due to excessive temperature, and thus extending the service life of the equipment.

[0030] In some embodiments, a control panel is provided on the side wall of the control room, and a display surface of the control panel is exposed on the outer surface of the box body.

[0031] In the above implementation process, since the display surface of the control panel is exposed on the outer surface of the cabinet, the operator can directly monitor and operate the equipment without opening the cabinet door or performing additional operations, which greatly improves the convenience and efficiency of operation.

[0032] In some embodiments, a PID controller is installed in the control room, and the PID controller is connected to the fill light signal.

[0033] In the above implementation process, the PID controller is connected to the fill light through a signal line. The PID controller receives feedback signals from the system or other sensors, and calculates the adjustment amount according to the preset control algorithm. According to the output signal of the PID controller, the fill light can adjust its luminous intensity, color temperature or light type and other parameters to meet the needs of the detection system; at the same time, the PID controller will receive feedback signals again according to the actual effect of the fill light after adjustment, and make a new round of adjustments. This process is repeated until the output of the fill light reaches or approaches the preset reference value. The PID controller has high precision and stability, and can achieve precise control of the fill light, which helps to ensure that the detection system can obtain stable light output under various conditions.

[0034] On the other hand, the embodiment of the present application also provides a detection method based on the above-mentioned flotation ash detection device, comprising the steps of:

[0035] Open the feed valve of the feed pipe group to allow the tailings solution to flow into the observation pool;

[0036] Turn on the fill light and adjust the light intensity of the test environment to a constant value;

[0037] After the tailings solution in the observation pool is stabilized, the camera is started to capture multiple groups of tailings solution images;

[0038] The edge box in the control host obtains the captured image and performs grayscale calculation and average calculation;

[0039] The dispatch center in the control host obtains the grayscale average value and performs subsequent processing.

[0040] In the above implementation process, by taking multiple groups of images and calculating the grayscale average, the influence of image noise and random errors on the detection results can be significantly reduced, thereby improving the accuracy of detection.

[0041] In some embodiments, a deep neural network model is loaded into the edge box, and the training process of the deep neural network model includes the steps of:

[0042] S1. Configure the tailings solution to set the grayscale range;

[0043] S2. The camera takes a picture of the tailings solution;

[0044] S3. Establishing a corresponding relationship between the image and the gray value;

[0045] S4. Transform the image left-right, up-down, and add noise to increase the number of images, increase the robustness of the algorithm, and reduce the overfitting of the model;

[0046] S5. After image data normalization;

[0047] S6. Import a preset neural network structure;

[0048] S7. Evaluate the accuracy, absolute error, and relative error indicators of the calculation results;

[0049] S8. Determine whether the training model meets the requirements. If so, end the model training. Otherwise, modify the structure of the neural network and execute S6.

[0050] In the above implementation process, through data enhancement operations, the model can learn more image features and has strong robustness to noise and transformation in the image. Through the sophisticated training process and evaluation indicators, the model's prediction accuracy can be ensured to be high. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments of the present application will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0052] Figure 1 This is a schematic diagram of the front structure of the flotation ash detection device described in an embodiment of the present application;

[0053] Figure 2This is a schematic diagram of the back structure of the flotation ash detection device described in an embodiment of the present application;

[0054] Figure 3 This is one of the internal structure schematic diagrams of the flotation ash detection device described in the embodiment of the present application;

[0055] Figure 4 This is the second schematic diagram of the internal structure of the flotation ash detection device according to the embodiment of the present application;

[0056] Figure 5 A schematic diagram of components inside the detection chamber according to an embodiment of the present application;

[0057] Figure 6 for Figure 5 A schematic diagram of the structure from another perspective of the structure shown;

[0058] Figure 7 for Figure 5 A structural schematic diagram of the structure shown in another perspective;

[0059] Figure 8 This is a schematic diagram of the structure of the observation pool and the aggregate hopper combined in the embodiment of the present application;

[0060] Fig. 9 for Figure 8 An exploded schematic diagram of the structure shown;

[0061] Fig.10 for Figure 8 a cross-sectional view of the structure shown;

[0062] Fig.11 This is an exploded schematic diagram of the observation pool described in the embodiment of the present application;

[0063] Fig.12 This is a flow chart of the ash content detection method described in the embodiment of the present application;

[0064] Fig.13 This is a flowchart of the training method of the deep neural network model described in an embodiment of the present application.

[0065] in,

[0066] 1. Box body; 111. First cooling fan; 121. Second cooling fan; 13. Device bracket; 2. Fill light; 21. Top light source; 22. Side light source; 3. Camera; 4. Light sensor; 5. Observation pool; 51. Buffer pool part; 52. Observation pool part; 53. Buffer pool cover; 531. Avoidance hole; 54. Feed pipe group; 541. Feed connecting pipe; 542. Water inlet connecting pipe; 543. Tee; 544. Electric water valve; 6. Collecting funnel; 61. Discharge pipe; 62. Net; 7. Scraping mechanism; 71. Driver; 72. Scraper; 8. Electrical mounting frame; 9. Control panel. DETAILED DESCRIPTION

[0067] The technical solutions in the embodiments of the present application will be described below in conjunction with the drawings in the embodiments of the present application.

[0068] It should be noted that similar reference numerals and letters represent similar items in the following drawings, so once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings. At the same time, in the description of this application, the terms "first", "second", etc. are only used to distinguish the description and cannot be understood as indicating or implying relative importance.

[0069] Reference Figure 1-Figure 11 , the embodiment of the present application provides a flotation ash detection device, comprising:

[0070] Box 1, including a control room and a detection room;

[0071] A control host, installed in the control room;

[0072] An observation pool 5 is installed in the detection chamber and is connected to a feed pipe group 54;

[0073] A fill light 2 is installed in the detection room and connected to the control host signal, and includes a top light source 21 and a side light source 22, wherein the top light source 21 is located above the observation pool 5 and faces the observation pool 5, and the side light source 22 is located at the side of the observation pool 5 and faces the observation pool 5;

[0074] A camera 3 is installed in the detection room and faces the observation pool 5, and the camera 3 is connected to the control host by signal;

[0075] The light sensor 4 is installed in the detection room and is connected to the control host by signal.

[0076] Among them, in the specific structure of the box body 1, it includes a frame and surrounding wall panels installed on the walls of each direction of the frame. In order to facilitate personnel operation, some of the surrounding wall panels can also be set as door panels that can be opened and closed. A control room and a detection room separated from each other are set in the box body 1. According to requirements, the control room and the detection room can be arranged up and down or left and right. The control room is used to install electronic equipment such as the control host to ensure the safe operation of the equipment and facilitate personnel operation. The detection room is used to install detection components such as the observation pool 5, the fill light 2, the camera 3 and the light sensor 4, providing a relatively closed and stable detection environment.

[0077] The control host is installed in the control room and is responsible for data processing and issuing control instructions for the entire detection device.

[0078] The observation pool 5 is installed in the detection room and is used to contain the tailings solution to be detected. It is connected to a feed pipe group 54 to facilitate the input of the tailings solution.

[0079] The fill light 2 includes a top light source 21 and a side light source 22, which are respectively located above and on the side of the observation pool 5, and are connected to the control host signal. The brightness of the light source can be adjusted according to the command of the control host to ensure the clarity and accuracy of image acquisition.

[0080] The camera 3 is installed in the detection room, facing the observation pool 5, and is used to capture images of the tailings solution. It is connected to the control host signal and transmits the captured images to the control host for processing.

[0081] The light sensor 4 is installed in the detection room for real-time monitoring of the light intensity in the detection room, and is connected to the control host signal to transmit the light intensity data to the control host so that the control host can adjust the brightness and angle of the fill light 2 according to the light intensity.

[0082] When performing the detection work, the tailings solution to be detected is input into the observation pool 5 through the feed pipe group 54. After turning on the fill light 2 to illuminate, the light sensor 4 monitors the light intensity in the detection room in real time and transmits the data to the control host. The control host adjusts the brightness of the fill light 2 according to the light intensity data to ensure that the tailings solution in the observation pool 5 can obtain a stable and clear image when the camera 3 takes pictures. The camera 3 takes an image of the tailings solution in the observation pool 5 and transmits the image data to the control host. The control host processes and analyzes the received image data, detects the ash content in the tailings solution through image recognition technology, and then outputs the analysis results to the display screen or other output devices for the operator to view and record.

[0083] Based on the flotation ash detection device of this embodiment, a black box environment (i.e., a detection room) is formed by using a box body 1 to isolate the influence of ambient light on the taken photos; at the same time, a fill light 2 is set in the detection room to provide a constant light source environment, which can ensure the clarity and accuracy of the image collected by the camera 3, thereby improving the accuracy of ash detection. The entire detection process is controlled by the control host without manual intervention, realizing automatic detection and improving production efficiency. The coordinated use of the light sensor 4 and the fill light 2 can monitor and adjust the light intensity in the detection room in real time, ensuring that the light environment for detection can always be maintained at a constant value, thereby improving the stability of the detection device. In addition, the control host and the detection components are both installed in the box body 1, with a compact structure and easy operation and maintenance.

[0084] In some embodiments, reference Figure 10-11The observation pool 5 includes a buffer pool portion 51 and an observation pool portion 52 which are connected at the bottom, the feed pipe group 54 is connected to the buffer pool portion 51, and the fill light 2 and the camera 3 are respectively arranged toward the observation pool portion 52.

[0085] The bottoms of the buffer tank 51 and the observation tank 52 are connected, allowing the tailings solution to first enter the buffer tank 51 and then naturally flow into the observation tank 52. The main function of the buffer tank 51 is to receive the tailings solution from the feed pipe group 54 and play a role of buffering and preliminary precipitation, which helps to reduce image blur or instability caused by direct impact or rapid flow, thereby improving the quality of subsequent image acquisition. The observation tank 52 is the core area of ​​image acquisition, and the fill light 2 and the camera 3 are both set towards this area.

[0086] In some embodiments, the detection chamber is also equipped with an aggregate funnel 6 located at the bottom of the observation pool 5, and the aggregate funnel 6 is connected to a discharge pipe 61; and the downward projection of the observation pool portion 52 completely falls into the aggregate funnel 6, so that the tailings solution overflowing from the edge of the observation pool portion 52 can fall into the aggregate funnel 6.

[0087] The downward projection of the observation pool 52 completely falls within the range of the aggregate hopper 6 , which means that the tailings solution overflowing from the edge of the observation pool 52 can directly fall into the aggregate hopper 6 and be discharged and recovered through the discharge pipe 61 connected to the aggregate hopper 6 .

[0088] Therefore, the design of the aggregation funnel 6 ensures that the tailings solution overflowing from the edge of the observation pool portion 52 will not flow to other parts of the detection chamber, but will be collected by the aggregation funnel 6 and uniformly discharged for recycling. By timely collecting and discharging the tailings solution, the aggregation funnel 6 helps to keep the observation pool portion 52 clean and stable, avoiding possible contamination or damage.

[0089] In some embodiments, reference Fig. 9 The top of the aggregate hopper 6 is provided with a leakage net 62 arranged around the periphery of the observation pool part 52.

[0090] When tailings solution overflows from the edge of the observation pool 52, the solution will first contact the filter 62. Since the mesh size of the filter 62 is moderate, it can effectively intercept solid particles with larger diameters in the solution, thereby preventing large particles from being fixed and falling into the discharge pipe 61, causing blockage.

[0091] In some embodiments, reference Figure 5-Figure 10A scraping mechanism 7 is also provided in the detection chamber. The scraping mechanism 7 includes a connected driver 71 and a scraper 72. The scraper 72 extends into the buffer pool 51. The driver 71 can drive the scraper 72 to move in the buffer pool 51 to scrape off the foam on the surface of the tailings solution in the buffer pool 51.

[0092] Among them, the driver 71 can be electric, pneumatic or hydraulic, and the specific choice depends on the actual application scenario and cost considerations. Its main function is to provide a stable driving force so that the scraper 72 can move in the buffer pool 51. Optionally, the driver 71 is a driving structure composed of a motor and a screw rod, which can drive the hanging plate to move back and forth in a straight line in the buffer pool 51. The scraper 72 is usually made of corrosion-resistant and wear-resistant materials to ensure its stability and reliability in long-term use. Its shape and size are carefully designed to effectively scrape off the foam on the surface of the tailings solution. Optionally, the scraper 72 is in the shape of a thin plate, which is adapted to the vertical section of the buffer pool 51 and is provided with a plurality of water holes.

[0093] During the flotation process, under the action of the liquid medicine, a layer of foam may form on the surface of the tailings solution, which will interfere with the image acquisition of the camera 3, thereby affecting the accuracy of ash detection. The scraping mechanism 7 drives the scraper 72 to move in the buffer tank 51 through the driver 71, and the scraper 72 contacts the tailings solution and scrapes off the foam. By scraping off the foam on the surface of the tailings solution, the scraping mechanism 7 reduces the interference of the foam on the image acquisition of the camera 3, thereby improving the accuracy of ash detection.

[0094] In some embodiments, a buffer pool cover 53 is provided on the top side of the buffer pool portion 51 , and a avoidance hole 531 corresponding to the moving trajectory of the scraper 72 is provided on the buffer pool cover 53 , and the scraper 72 passes through the buffer pool cover 53 through the avoidance hole 531 .

[0095] The buffer tank cover 53 is installed on the top side of the buffer tank portion 51 to seal and protect the buffer tank portion 51. It can prevent external impurities from entering the buffer tank portion 51 and reduce the evaporation and splashing of the tailings solution. The avoidance hole 531 is cleverly designed on the buffer tank cover 53 and corresponds to the moving track of the scraper 72, so that the scraper 72 can move along the avoidance hole 531 without being hindered.

[0096] In some embodiments, reference Figure 3-Figure 7 A device bracket 13 is provided in the detection chamber, the top of which is fixed to the top of the detection chamber, and the top light source 21, the camera 3, the light sensor 4 and the bubble scraping mechanism 7 are respectively installed on the device bracket 13.

[0097] The device support 13 is a solid structure, the top of which is fixed to the top of the detection chamber, which provides sufficient support and stability to carry and fix key components such as the top light source 21, the camera 3, the light sensor 4 and the scraping bubble mechanism 7. Key components such as the top light source 21, the camera 3, the light sensor 4 and the scraping bubble mechanism 7 are respectively installed on the device support 13 through appropriate mounting methods and connectors. These components are precisely positioned on the support to ensure that they can perform their functions correctly and will not interfere with each other.

[0098] In the above embodiments, the device bracket 13 provides stable support and fixation for key components, reducing the risk of displacement or damage caused by vibration or external force, which improves the stability and reliability of the entire detection system. By rationally arranging and installing key components, the device bracket 13 helps to simplify the structure of the device, improve the compactness and aesthetics of the device, and at the same time, it also makes operation and maintenance more convenient and efficient. Since the device bracket 13 can ensure that each key component maintains the correct position and posture, it helps to improve the accuracy and reliability of detection.

[0099] In some embodiments, reference Fig. 9 and Fig.10 The feed pipe group 54 includes a tee 543, a feed connection pipe 541 and a water inlet connection pipe 542. The tee 543 includes a first interface, a second interface and a third interface. The first interface is connected to the feed connection pipe 541, the second interface is connected to the water inlet connection pipe 542, and the third interface is connected to the observation pool 5. The water inlet connection pipe 542 is set. Before the detection, the water inlet connection pipe 542 can be opened to pass clean water to rinse the observation pool 5, and then the tailings solution is passed.

[0100] Before the detection begins, the valve of the water inlet connecting pipe 542 can be opened to allow clean water to flow from an external water source into the observation pool 5. The clean water enters the tee 543 through the water inlet connecting pipe 542 and flows into the observation pool 5 from the third interface of the tee 543. Under the action of the clean water, the residues and impurities in the observation pool 5 are washed away, preparing for the subsequent tailings solution detection. After the cleaning is completed, the valve of the water inlet connecting pipe 542 is closed, and the feed valve of the feed connecting pipe 541 is opened. The tailings solution enters the tee 543 from the outside through the feed connecting pipe 541 and flows into the observation pool 5 from the third interface of the tee 543.

[0101] By introducing the water inlet connecting pipe 542 in the above embodiment, the observation pool 5 can be rinsed and cleaned before detection, which helps to reduce the impact of residues and impurities on the detection results, thereby improving the accuracy and reliability of the detection.

[0102] In some embodiments, an electric water valve 544 is provided in the water inlet connecting pipe 542 .

[0103] The electric water valve 544 is connected to the control host through a signal line, and the control host can send a control signal to the electric water valve 544 to control its opening and closing state. Through the signal connection between the electric water valve 544 and the control host, the automatic control of the opening and closing of the water inlet connecting pipe 542 is realized, which improves the convenience and accuracy of the operation and reduces the need for manual intervention.

[0104] In some embodiments, the side wall and / or top wall and / or bottom wall of the control room are provided with a first cooling fan 111; the side wall and / or top wall and / or bottom wall of the detection room are provided with a second cooling fan 121.

[0105] When the equipment is working, a certain amount of heat will be generated in the control room and the detection room. If this heat is not discharged in time, it will cause the temperature of the equipment to rise, which will affect its performance and stability. By setting up a cooling fan, effective air convection can be formed to discharge heat from the room. Through effective heat dissipation, the temperature in the control room and the detection room can be kept within a suitable range, thereby improving the stability and reliability of the equipment, reducing damage or failure of the equipment due to excessive temperature, and thus extending the service life of the equipment.

[0106] Optionally, the first cooling fan 111 and the second cooling fan 121 are connected to the control host and can be automatically started or shut down according to the working status and temperature conditions of the equipment. When it is detected that the indoor temperature exceeds the set value, the control host will automatically start the corresponding cooling fan to lower the indoor temperature.

[0107] In some embodiments, the side wall of the control room is provided with an electrical mounting rack 8, which can be used to fix the electrical components inside so that the electrical components can be neatly arranged therein.

[0108] In some embodiments, a control panel 9 is provided on the side wall of the control room, and a display surface of the control panel 9 is exposed on the outer surface of the box body 1 .

[0109] Since the display surface of the control panel 9 is exposed on the outer surface of the box body 1, the operator can directly monitor and operate the equipment without opening the door or performing additional operations, which greatly improves the convenience and efficiency of operation.

[0110] In some embodiments, a PID controller is installed in the control room, and the PID controller is connected to the fill light 2 signal.

[0111] A PID controller, or proportional-integral-derivative controller, is a common feedback loop component in industrial control applications.

[0112] The PID controller is connected to the fill light 2 through a signal line. The PID controller receives feedback signals from the system or other sensors, and calculates the adjustment amount according to the preset control algorithm. According to the output signal of the PID controller, the fill light 2 can adjust its luminous intensity, color temperature or light type and other parameters to meet the needs of the detection system; at the same time, the PID controller will receive feedback signals again according to the actual effect of the fill light 2 after adjustment, and make a new round of adjustments. This process is cyclical until the output of the fill light 2 reaches or approaches the preset reference value. The PID controller has high precision and stability, and can achieve precise control of the fill light 2, which helps to ensure that the detection system can obtain stable light output under various conditions.

[0113] On the other hand, refer to Fig.12 This embodiment also provides a detection method based on the above-mentioned flotation ash detection device, comprising the steps of:

[0114] Open the feed valve of the feed pipe group 54 to allow the tailings solution to flow into the observation pool 5;

[0115] Turn on fill light 2 and adjust the light intensity of the test environment to a constant value;

[0116] After the tailings solution in the observation pool 5 is stabilized, the camera 3 is started to capture multiple groups of tailings solution images;

[0117] The edge box in the control host obtains the captured image and performs grayscale calculation and average calculation;

[0118] The dispatch center in the control host obtains the grayscale average value and performs subsequent processing.

[0119] Among them, the captured image is transmitted to the edge box in the control host for processing. The edge box is responsible for obtaining image data and performing grayscale calculation. Grayscale calculation is to convert each pixel in the image into a grayscale value. These grayscale values ​​reflect the brightness information of each pixel in the image. In order to obtain more accurate grayscale values, the edge box will perform grayscale calculation on multiple groups of images and calculate the average value. This step helps to reduce the impact of image noise and random errors on grayscale values. The grayscale average value calculated by the edge box is transmitted to the dispatch center in the control host for subsequent processing. The dispatch center is the core part of the detection system. It is responsible for receiving, processing and storing data from various components. After receiving the grayscale average value, the dispatch center will perform further analysis and processing to obtain the ash content or other related parameters of the tailings solution.

[0120] Based on the detection method of this embodiment, by taking multiple groups of images and calculating the grayscale average value, the influence of image noise and random errors on the detection results can be significantly reduced, thereby improving the accuracy of detection.

[0121] In some embodiments, reference Fig.13 , a deep neural network model is loaded into the edge box, and the training process of the deep neural network model includes the steps of:

[0122] S1. Configure the tailings solution to set the grayscale range;

[0123] S2. Camera 3 takes a picture of the tailings solution;

[0124] S3. Establishing a corresponding relationship between the image and the gray value;

[0125] S4. Transform the image left-right, up-down, and add noise to increase the number of images, increase the robustness of the algorithm, and reduce the overfitting of the model;

[0126] S5. After normalization of image data;

[0127] S6. Import a preset neural network structure;

[0128] S7. Evaluate the accuracy, absolute error, and relative error indicators of the calculation results;

[0129] S8. Determine whether the training model meets the requirements. If so, end the model training. Otherwise, modify the structure of the neural network and execute S6.

[0130] Optionally, the grayscale range is set according to specific needs. The equipment in the embodiment of the present application is used to indirectly analyze the ash content of the tailings solution. When the ash concentration of the tailings solution is too high or too low, the image captured by the camera may be indistinguishable. Therefore, the grayscale range of the embodiment of the present application is 55%-75%.

[0131] Among them, a corresponding relationship is established between the image and the gray value. Optionally, the image data is used as the input of the neural network, and the gray value is the output of the neural network.

[0132] Among them, transforming the image left and right, up and down, and adding noise operations can increase the scope of neural network learning, allowing it to learn more images. After the neural network learns more pictures, it helps to increase the reliability of its algorithm.

[0133] Optionally, to determine whether the training model meets the requirements, the following criteria can be used: |calculated value - theoretical value| / theoretical value ≤ 1%.

[0134] In the above embodiments, through data enhancement operations, the model can learn more image features and has strong robustness to noise and transformation in the image. Through a sophisticated training process and evaluation indicators, the prediction accuracy of the model can be ensured to be high.

[0135] The above description is only an embodiment of the present application and is not intended to limit the scope of protection of the present application. For those skilled in the art, the present application may have various changes and variations. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application should be included in the scope of protection of the present application. It should be noted that similar reference numerals and letters represent similar items in the following drawings, so once an item is defined in one drawing, it does not need to be further defined and explained in the subsequent drawings.

[0136] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0137] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

Claims

1. A flotation ash detection device, characterized in that: include: The box body includes a control room and a detection room; A control host, installed in the control room; An observation pool is installed in the detection chamber and is connected to a feed pipe group; A fill light, installed in the detection room and connected to the control host signal, comprising a top light source and a side light source, wherein the top light source is located above the observation pool and faces the observation pool, and the side light source is located at the side of the observation pool and faces the observation pool; A camera, installed in the detection chamber and facing the observation pool, the camera being connected to the control host by signal; The light sensor is installed in the detection room and is connected to the control host by signal.

2. The flotation ash detection device according to claim 1, characterized in that: The observation pool comprises a buffer pool portion and an observation pool portion which are connected at the bottom, the feeding pipe group is connected to the buffer pool portion, and the fill light and the camera are respectively arranged toward the observation pool portion.

3. The flotation ash detection device according to claim 2, characterized in that: The detection chamber is also provided with an aggregate funnel at the bottom of the observation pool, the aggregate funnel being connected to a discharge pipe; and the downward projection of the observation pool falls completely into the aggregate funnel, so that the tailings solution overflowing from the edge of the observation pool can fall into the aggregate funnel.

4. The flotation ash detection device according to claim 3, characterized in that: The top of the aggregate hopper is provided with a leakage net arranged around the periphery of the observation pool.

5. The flotation ash detection device according to claim 2, characterized in that: The detection chamber is also provided with a scraping mechanism, which includes a connected driver and a scraper. The scraper extends into the buffer pool. The driver can drive the scraper to move in the buffer pool to scrape off the foam on the surface of the tailings solution in the buffer pool.

6. The flotation ash detection device according to claim 5, characterized in that: A buffer pool cover is provided on the top side of the buffer pool portion, and a position avoidance hole corresponding to the moving track of the scraper is provided on the buffer pool cover, and the scraper passes through the buffer pool cover through the position avoidance hole.

7. The flotation ash detection device according to claim 5, characterized in that: A device bracket is provided in the detection chamber, the top end of the device bracket is fixed to the top of the detection chamber, and the top light source, the camera, the light sensor and the bubble scraping mechanism are respectively installed on the device bracket.

8. The flotation ash detection device according to claim 1, characterized in that: The feed pipe group includes a tee, a feed connecting pipe and a water inlet connecting pipe, the tee includes a first interface, a second interface and a third interface, the first interface is connected to the feed connecting pipe, the second interface is connected to the water inlet connecting pipe, and the third interface is connected to the observation pool.

9. The flotation ash detection device according to claim 8, characterized in that: An electric water valve is arranged in the water inlet connecting pipe.

10. The flotation ash detection device according to claim 1, characterized in that: The side walls and / or top walls and / or bottom walls of the control room are provided with a first cooling fan; the side walls and / or top walls and / or bottom walls of the detection room are provided with a second cooling fan.

11. The flotation ash detection device according to claim 1, characterized in that: A control panel is disposed on the side wall of the control room, and a display surface of the control panel is exposed on the outer surface of the box body.

12. The flotation ash detection device according to claim 1, characterized in that: A PID controller is installed in the control room, and the PID controller is connected with the fill light signal.

13. A detection method for a flotation ash detection device as claimed in any one of claims 1 to 12, characterized in that: Includes steps: Open the feed valve of the feed pipe group to allow the tailings solution to flow into the observation pool; Turn on the fill light and adjust the light intensity of the test environment to a constant value; After the tailings solution in the observation pool is stabilized, the camera is started to capture multiple groups of tailings solution images; The edge box in the control host obtains the captured image and performs grayscale calculation and average calculation; The dispatch center in the control host obtains the grayscale average value and performs subsequent processing.

14. The detection method according to claim 13, characterized in that: The edge box is loaded with a deep neural network model, and the training process of the deep neural network model includes the steps of: S1. Configure the tailings solution to set the grayscale range; S2. The camera takes a picture of the tailings solution; S3. Establishing a corresponding relationship between the image and the gray value; S4. Transform the image left-right, up-down, and add noise to increase the number of images, increase the robustness of the algorithm, and reduce the overfitting of the model; S5. After image data normalization; S6. Import a preset neural network structure; S7. Evaluate the accuracy, absolute error, and relative error indicators of the calculation results; S8. Determine whether the training model meets the requirements. If so, end the model training. Otherwise, modify the structure of the neural network and execute S6.

Images