INTELLIGENT FLOTATION DOSING SYSTEM AND METHOD BASED ON DETECTION OF FLOTATION TAILING SLURRY
TECHNICAL FIELD The present disclosure relates to the technical field of flotation dosing, and in particular relates to an intelligent flotation dosing system and method based on detection of a flotation tailing slurry.
BACKGROUND The flotation technology is the most economical and effective method for separating fine coal slime, and is also an important method for deep coal separation. The flotation technology plays a vital role in the coal slime water treatment of a coal preparation plant and the implementation of closed-loop circulation of coal slime water. The accurate control of dosing amounts during the flotation process is an important means to ensure an ash content in flotation clean coals and a recovery rate of flotation clean coals and reduce the consumption of agents. If the dosing amounts during a flotation process are too large, the selectivity will be poor, which will increase the ash content in clean coal while increasing the consumption of agents and cannot guarantee a quality of clean coal. If the dosing amounts during the flotation process are too small, the selectivity will be too high, which will reduce the yield of clean coal. The current flotation systems of coal preparation plants have a low intelligent level, and the flotation dosing is mainly determined manually. Thus, a flotation effect depends on the production experience and the careful management degree of a flotation operator. According to sensory results such as a color of the slurry observed by eyes and whether there is a loss of coarse mineral particles in a flotation tailing determined by hand touch, the flotation dosing is determined manually, which cannot allow the accurate quantification, is lagging and subjective, and results in unstable production situations. Because a flotation dosing device is generally arranged at a high position, there is large physical exertion due to frequent adjustment. In addition, a pungent odor caused by the volatilization of agents will affect the health of a flotation operator, and thus a flotation working environment needs to be improved. The traditional automatic flotation dosing systems mostly adopt open-loop control. In these systems, flotation agents are fed according to fixed values determined by a flotation operator, and there is a lack of a feedback link of automatically adjusting flotation dosing amounts according to working conditions. With the increasingly-strict requirements for a quality of flotation on the market and the continuous improvement of an intelligent level of an industrial process, the intelligent coal slime flotation has attracted more and more attention. One of the key links for intelligent control of a flotation process is to allow the real-time detection of product indexes during a flotation process. Currently, an ash content in a flotation slurry is mainly predicted by an image method and a direct detection method. In the traditional image method, an ash content in a concentrate is predicted through the analysis of a foam and grayscale of a concentrate. Because the ash content in a concentrate fluctuates in a relatively-small range and there is a high requirement for a prediction accuracy, it is difficult to allow the accurate prediction of the ash content, and the deviation of prediction of the ash content in a concentrate will also expand an impact on a recovery rate. In the direct detection method, the ash content in a concentrate slurry is currently detected by a slurry ash analyzer in most cases, and there is a lack of detection of a tailing slurry, reducing the control of the recovery rate. In addition, the slurry ash analyzer involves a high investment cost, and a measurement result of the slurry ash analyzer is lagging to some degree, such that a real-time monitoring effect cannot be allowed and the dosing amounts cannot be adjusted in time, making a flotation efficiency reduced.
SUMMARY In view of the above analysis, an embodiment of the present disclosure is intended to provide an intelligent flotation dosing system and method based on detection of a flotation tailing slurry, to solve the problems such as lagging adjustment and inaccurate agent amounts in the existing manual adjustment for flotation dosing. In a first aspect, the present disclosure provides an intelligent flotation dosing system based on detection of a flotation tailing slurry, including a flotation information acquisition unit, an ash content intelligent prediction unit, and a distributed control dosing unit, where the flotation information acquisition unit is configured to acquire information about a flow rate, a concentration, a coarse particle content, and an image of a slurry, the ash content intelligent prediction unit is configured to process the information about the flow rate, the concentration, the coarse particle content, and the image of the slurry to obtain an ash content of the slurry, and the distributed control dosing unit is configured to adjust a dosing amount according to the ash content of the slurry and the flow rate, the concentration, and the coarse particle content of the slurry. Further, the system further includes a slurry conditioner and a flotation device, where the slurry conditioner is configured to pretreat the slurry to obtain a pretreated slurry and the flotation device is configured to allow flotation for the pretreated slurry. Further, the slurry conditioner includes a barrel body, an upper end of the barrel body is provided with a water-inlet pipe and a material-inlet pipe, and a lower end of the barrel body is provided with a material-outlet pipe.
Further, the flotation information acquisition unit includes a flow meter and a concentration meter, and the flow meter and the concentration meter both are arranged downstream from the slurry conditioner. Further, the flow meter and the concentration meter are configured to acquire the flow rate and the concentration of the slurry flowing into the flotation device from the slurry conditioner, respectively. Further, the flotation information acquisition unit further includes a force sensor and an industrial camera, and the force sensor and the industrial camera both are arranged at a tailing outlet of the flotation device. Further, the force sensor is configured to acquire vibration data of a baffle impacted by the slurry, and the industrial camera is configured to acquire the image of the slurry. Further, the ash content intelligent prediction unit includes an image ash content prediction unit and a vibration mode recognition unit, where the image ash content prediction unit is configured to calculate a grayscale of the slurry according to the image of the slurry, and the vibration mode recognition unit is configured to calculate the coarse particle content of the slurry according to vibration data of a baffle impacted by the slurry that is acquired by the force sensor. Further, the distributed control dosing unit includes a centralized control center unit and a flotation control center unit, where the centralized control center unit is configured to send a dosing instruction to the flotation control center unit according to the ash content, the coarse particle content, an amount, and the concentration of the slurry, and the flotation control center unit is configured to parse the dosing instruction and send a dosing amount instruction to an automatic dosing chamber. In a second aspect, the present disclosure provides an intelligent flotation dosing method based on detection of a flotation tailing slurry, where the method is applied to the intelligent flotation dosing system based on the detection of the flotation tailing slurry described above and includes the following steps: Si: treating a flotation feed through the slurry conditioner to obtain a treated flotation feed, acquiring the information about the flow rate and the concentration by the flow meter and the concentration meter, respectively, separating the treated flotation feed by the flotation device to obtain a tailing, and allowing the tailing to enter an information acquisition zone; S2: feeding information acquired by the force sensor to the vibration mode recognition unit based on a mode recognition algorithm, and analyzing the coarse particle content in the tailing based on the mode recognition algorithm of support vector machine (SVM) regression by a model trained with previous data; and feeding information acquired by the industrial camera to the image ash content prediction unit based on a neural network, and predicting the ash content in the tailing in real time through analysis of an image grayscale distribution in combination with the coarse particle content; and S3: transmitting the information about the ash content, the coarse particle content, the amount, and the concentration of the slurry into the centralized control center unit, and sending a dosing instruction by a decision-making system of the centralized control center unit to the flotation control center unit according to the information about the ash content, the coarse particle content, the amount, and the concentration of the slurry, to adjust the ash content of the tailing and allow closed-loop control. Compared with the prior art, the present disclosure has at least one of the following advantages. (1) The present disclosure can allow the 24 h uninterrupted real-time monitoring of flotation parameters to adjust a dosing amount timely, which has strong timeliness. (2) In the present disclosure, a mode recognition algorithm is introduced and a force sensor is used to detect a coarse particle content in a tailing slurry and timely feed back whether there is a loss of coarse mineral particles in the tailing slurry, which reduces a cost while ensuring the reliability of data and can improve a flotation efficiency. (3) In the intelligent dosing method of the present disclosure, a quantitative dosing strategy based on principal component analysis (PCA) and back propagation (BP) neural network regression analysis databases and machine learning training is introduced, and a dosing amount is mechanically and accurately controlled, which can avoid the under-adjustment or over-adjustment situations caused by operation and determination errors of workers and improve a recovery rate of clean coal. (4) The intelligent dosing method provided by the present disclosure involves a simple flow, a small investment, a low operating cost, and a significant economic benefit, does not require an investment on large-scale devices, and is easily obtained through transformation on the basis of the original factory. (5) In the present disclosure, the prediction regression correction is conducted in combination with the daily coal mining data of a factory, and an algorithm is adjusted, so as to prevent the deviation of a prediction result caused by factors such as a fluctuation of a coal sample and a change of a coal type. The above technical solutions in the present disclosure can also be combined with each other to provide increased preferred combination solutions. Other features and advantages of the present disclosure will be described in the following description, and some of these will become apparent from the description or be understood by implementing the present disclosure. The objectives and other advantages of the present disclosure may be implemented or derived by those specifically indicated in the description and accompanying drawings.
BRIEF DESCRIPTION OF THE DRAWINGS The accompanying drawings are provided merely to illustrate the specific embodiments, rather than to limit the present disclosure. The same reference numerals represent the same components throughout the accompanying drawings. FIG. 1 is a schematic structural diagram of an intelligent flotation dosing system in a specific embodiment. FIG. 2 is a control flow chart of an intelligent flotation dosing system in a specific embodiment. Reference Numerals: 1: slurry conditioner; 11: barrel body; 12: water-inlet pipe; 13: material-inlet pipe; 14: material-outlet pipe; 2: flotation device; 31: flow meter; 32: concentration meter; 33: force sensor; 34: industrial camera; and 35: light source.
DETAILED DESCRIPTION OF THE EMBODIMENTS Preferred embodiments of the present disclosure will be specifically described below with reference to the accompanying drawings. The accompanying drawings constitute a part of the present disclosure, and are used together with the embodiments of the present disclosure to explain the principles of the present disclosure rather than limit a scope of the present disclosure. In the description of the embodiments of the present disclosure, it should be noted that, unless otherwise clearly specified, the term "connected to" should be understood in a broad sense. For example, a connection may be a fixed connection, a removable connection, an integral connection, a mechanical connection, an electrical connection, a direct connection, or an indirect connection through a medium. Those of ordinary skill in the art may understand a specific meaning of the above term in the present disclosure based on a specific situation. The terms "top", "bottom", "above", "lower", and "upper" used in the full text refer to relative positions of components of a device, such as relative positions of top and bottom substrates inside the device. It can be understood that the device is multifunctional, which is independent of an orientation of the device in a space. Embodiment 1 In a specific embodiment of the present disclosure, as shown in FIG. 1 to FIG. 2, an intelligent flotation dosing system based on detection of a flotation tailing slurry (referred to as "intelligent flotation dosing system" hereinafter) is disclosed, The intelligent flotation dosing system includes a flotation information acquisition unit, an ash content intelligent prediction unit, and a distributed control dosing unit. The flotation information acquisition unit is configured to acquire information about a flow rate, a concentration, a coarse particle content, and an image of a slurry. The ash content intelligent prediction unit is configured to process the information about the flow rate, the concentration, the coarse particle content, and the image of the slurry to obtain an ash content of the slurry. The distributed control dosing unit is configured to adjust a dosing amount according to the ash content of the slurry and a flow rate, a concentration, and a coarse particle content of the slurry. The intelligent flotation dosing system further includes a slurry conditioner 1 and a flotation device 2, where the slurry conditioner 1 is configured to pretreat the slurry to obtain a pretreated slurry and the flotation device 2 is configured to allow flotation for the pretreated slurry. The slurry conditioner 1 includes a barrel body 11, an upper end of the barrel body 11 is provided with a water-inlet pipe 12 and a material-inlet pipe 13, and a lower end of the barrel body 11 is provided with a material-outlet pipe 14. The flotation information acquisition unit includes a flow meter 31 and a concentration meter 32, and the flow meter 31 and the concentration meter 32 both are arranged downstream from the slurry conditioner 1 and are configured to acquire a flow rate and a concentration of the slurry flowing into the flotation device 2 from the slurry conditioner 1, respectively. The flotation information acquisition unit further includes a force sensor 33 and an industrial camera 34, and the force sensor 33 and the industrial camera 34 both are arranged at a tailing outlet of the flotation device 2. The force sensor 33 is configured to acquire vibration data of a baffle impacted by the slurry, and the industrial camera 34 is configured to acquire an image of the slurry. It should be noted that the industrial camera 34 is equipped with a light source 35, a mist eliminator, and a lens hood. The ash content intelligent prediction unit includes an image ash content prediction unit and a vibration mode recognition unit; and the image ash content prediction unit is configured to calculate a grayscale of the slurry according to the image of the slurry, and the vibration mode recognition unit is configured to calculate a coarse particle content of the slurry according to vibration data of a baffle impacted by the slurry that is acquired by the force sensor 33. The distributed control dosing unit includes a centralized control center unit and a flotation control center unit; the centralized control center unit is configured to send a dosing instruction to the flotation control center unit according to specific conditions such as an ash content, a coarse particle content, an amount, and a concentration of the slurry; and the flotation control center unit is configured to parse the dosing instruction and send a dosing amount instruction to an automatic dosing chamber, and the automatic dosing chamber is configured to adjust dosing amounts of a flotation system by controlling main and auxiliary solenoid valves to adjust an ash content of a tailing, thereby allowing closed-loop control. Specifically, the automatic dosing chamber is configured to adjust a dosing amount of the slurry conditioner 1 by controlling a main solenoid valve and adjust a dosing amount of the flotation device 2 by controlling an auxiliary solenoid valve. An ash content prediction and calculation module is provided to conduct prediction regression correction in combination with the daily coal mining data of a factory and adjust an algorithm, so as to prevent the deviation of a prediction result caused by factors such as a fluctuation of a coal sample and a change of a coal type. Embodiment 2 In another specific embodiment of the present disclosure, as shown in FIG. 1 to FIG. 2, an intelligent flotation dosing method based on detection of a flotation tailing slurry is disclosed, where the method adopts the intelligent flotation dosing system based on detection of a flotation tailing slurry in Embodiment 1 and includes the following steps. SI: A flotation feed is treated through the slurry conditioner 1 to obtain a treated flotation feed, the information about the flow rate and the concentration is acquired by the flow meter 31 and the concentration meter 32, respectively, the treated flotation feed is separated by the flotation device 2 to obtain a tailing, and the tailing is allowed to enter an information acquisition zone. In this embodiment, the slurry conditioner 1 includes a barrel body 11, the barrel body 11 is provided with a water-inlet pipe 12, a material-inlet pipe 13, and a material-outlet pipe 14, the water-inlet pipe 12 and the material-inlet pipe 13 are arranged at an upper end of the barrel body 11, and the material-outlet pipe 14 is arranged at a lower end of the barrel body 11. The flow meter 31 and the concentration meter 32 both are arranged downstream from the slurry conditioner 1 and are configured to acquire a flow rate and a concentration of the slurry flowing into the flotation device 2 from the slurry conditioner 1, respectively. S2: Information acquired by the force sensor 33 is fed to the vibration mode recognition unit based on a mode recognition algorithm, and the coarse particle content in the tailing is analyzed based on the mode recognition algorithm of SVM regression by a model trained with previous data. The force sensor 33 is arranged at a specific position of a tailing discharge port. The mode recognition algorithm based on SVM regression can detect a coarse particle content by imitating the touch of a flotation operator to exclude the influence of the coarse particle content on the prediction of an ash content and allow the real-time monitoring of a coarse mineral particle loss. Information acquired by the industrial camera 34 is fed to the image ash content prediction unit based on a neural network such as a YOLOV5 network, and the ash content of the tailing is predicted in real time through analysis of an image grayscale distribution in combination with the coarse particle content. The devices such as the industrial camera 34, a light source 35, and a mist eliminator are arranged at an upper end of a flotation tailing outlet to monitor an image of a flotation tailing slurry. An independent computing unit module is provided to calculate a corresponding ash content using a convolutional neural network (CNN) model trained with samples according to a characteristic value extracted from a grayscale of an image of the slurry captured by the industrial camera 34, parameters of the flow meter and the concentration meter arranged at a tailing discharge port, and information such as an intensity of a given light source and a coarse particle content. The system can regularly conduct ash content prediction regression correction. An ash content prediction and calculation unit is provided to correct a prediction result by adjusting a principal component relationship and network weights of operation parameters according to true ash content data of regular coal mining of a given plant, thereby preventing the deviation of a prediction result caused by factors such as a fluctuation of a coal sample and a change of a coal type. In this embodiment, the force sensor 33 and the industrial camera 34 both are arranged at a tailing outlet of the flotation device 2. The force sensor 33 is configured to acquire vibration data of a baffle impacted by the slurry, and the industrial camera 34 is configured to acquire an image of the slurry. The industrial camera 34 is equipped with a light source 35, a mist eliminator, and a lens hood. S3: The information about the ash content, the coarse particle content, the amount, and the concentration of the slurry is transmitted into the centralized control center unit, and a dosing instruction is sent by a decision-making system of the centralized control center unit to the flotation control center unit according to the above specific conditions. The flotation control center unit is configured to parse the dosing instruction and send a dosing amount instruction to an automatic dosing chamber, and the automatic dosing chamber is configured to adjust dosing amounts of a flotation system by controlling main and auxiliary solenoid valves to adjust an ash content of a tailing, thereby allowing closed-loop control. The centralized control center unit is configured to control a flotation dosing link through a PCA algorithm according to predicted information such as an ash content, a coarse particle content, a slurry flow rate, and a slurry concentration in combination with database information. Specifically, the centralized control center unit is configured to acquire information such as a real-time image from the industrial camera and determine a flotation working condition according to predicted information such as an ash content, a coarse particle content, a slurry flow rate, and a slurry concentration. Through the decision-making system, the flotation dosing link is controlled using the PCA algorithm according to database information. The dosing information is sent to a dosing station, and the flotation dosing is accurately controlled through a mechanical dosing mode of a frequency converter + a mechanical diaphragm metering pump. When a change of a dosing amount exceeds a set warning threshold, the system gives an alarm that the change needs to be confirmed manually. Preferably, the frequency converter is Siemens G120C 0.75KWLO (0.55KWHO) class C, which is widely used for the frequency conversion of pumps and fans, supports bus control and analog input control, and can be conveniently controlled by a flotation industrial control system. Preferably, the mechanical diaphragm metering pump is Milton Roy GM0090PQ9MNN, where a pump head is made of a polyvinyl chloride (PVC) material and a diaphragm is made of a polytetrafluoroethylene (PTFE) material; and the mechanical diaphragm metering pump can accurately and effectively deliver a relatively-viscous flotation agent, and a flow rate of the mechanical diaphragm metering pump can be conveniently controlled by the frequency converter in combination with an inverter motor. Preferably, the flow meter of a dosing pipe is an NKGF-06F1Il/SLZ circular gear flow meter, which has characteristics such as high accuracy, small range, and organic corrosion resistance and is suitable for the real-time monitoring of a flotation dosing amount. Preferably, a programmable logic controller (PLC) is a Siemens S7 200 smart ST-20 transistor output, which can cooperate with an EAM03 analog input/output module to allow the real-time closed-loop control of a dosing amount by controlling the frequency converter according to information fed back by the flow meter. The intelligent flotation dosing system provided in this embodiment is configured to adjust a dosing amount according to data information of a flotation tailing slurry (a flow rate, a concentration, a coarse particle content, and an ash content of the slurry), and can allow the 24 h uninterrupted real-time monitoring of flotation parameters to adjust a dosing amount timely, which has strong timeliness. In the intelligent flotation dosing system provided in this embodiment, a mode recognition algorithm is introduced and a force sensor is used to detect a coarse particle content in a tailing slurry and timely feed back whether there is a loss of coarse mineral particles in a tailing slurry, which reduces a cost while ensuring the reliability of data and can improve a flotation efficiency. In the intelligent flotation dosing system provided in this embodiment, a quantitative dosing strategy based on PCA and BP neural network regression analysis databases and machine learning training is introduced, and a dosing amount is mechanically and accurately controlled, which can avoid the under-adjustment or over-adjustment situations caused by operation and determination errors of workers and improve a recovery rate of clean coal. With the intelligent flotation dosing system provided in this embodiment, an intelligent dosing process involves a simple flow, a small investment, a low operating cost, and a significant economic benefit, does not require an investment on large-scale devices, and is easily obtained through transformation on the basis of the original factory. The prediction regression correction is conducted in combination with the daily coal mining data of a factory, and an algorithm is adjusted, so as to prevent the deviation of a prediction result caused by factors such as a fluctuation of a coal sample and a change of a coal type. The present disclosure establishes a machine visual and tactile dosing decision-making mechanism based on an artificial neural network by imitating the situation that a flotation operator currently relies on visual and tactile senses for dosing. The present disclosure reduces a labor intensity of a flotation dosing worker, solves the problems such as lagging adjustment and inaccurate agent amounts caused by manual adjustment, can control a dosing amount for a flotation machine in a timely and long-term manner, and can reduce a coarse mineral particle loss and improve a recovery rate while ensuring an ash content of a product. The above are merely preferred specific implementations of the present disclosure, but the protection scope of the present disclosure is not limited thereto. Any modification or replacement easily conceived by those skilled in the art within the technical scope of the present disclosure should fall within the protection scope of the present disclosure.