Intelligent regulation and control method and device for microwave precipitation system based on sensor group

Abstract

The invention discloses a microwave precipitation system intelligent regulation and control method and device based on a sensor group, and belongs to the field of precipitation systems.The intelligent control method comprises the steps that the sludge concentration in a hydrocyclone is obtained through an online sludge concentration monitor, the initial adding amount of each filter tank is controlled, and an image sensor obtains image data; a turbidity sensor acquires turbidity acquisition data to generate a single-tank purification capacity parameter; a single-tank preparation correction amount is generated according to the single-tank purification capacity parameter; and the preparation addition amount of each filter tank is controlled according to the initial addition amount of each filter tank and the single-tank preparation correction amount. According to the intelligent regulation and control method of the microwave precipitation system based on the sensor group, the addition amount of the preparation can be accurately controlled according to the image and turbidity data of the filter tank, so that the purification effect is improved, and the operation cost of the system is saved.

Description

Technical Field

[0001] The present invention relates to the field of precipitation systems, and in particular to an intelligent control method for a microwave precipitation system based on a sensor group and a fully automatic intelligent control device for the fully automatic intelligent control method. Background Art

[0002] It is difficult for the existing microwave sedimentation system to reasonably regulate the injection amount of coagulants, polymers, and micro-sand according to the degree of raw water pollution. This is because the microwave sedimentation system is often composed of multiple tanks such as coagulation tanks, injection tanks, thermal tanks, and sedimentation tanks, and the changes in the water quality of raw water are difficult to capture, which will cause the system debugger to often set the injection amount of coagulants, polymers, micro-sand, etc. based on experience. Therefore, they tend to ensure the sedimentation effect by injecting excess agents and rely on manual maintenance to adapt to the dynamic changes in the water quality of raw water.

[0003] In order to solve the above problems, Chinese patent document CN216366774U discloses a sedimentation tank sludge control system based on image recognition. The system includes a sedimentation tank, a sludge synchronization device outside the sedimentation tank, an image recognition and analysis device, a sludge discharge device and a sludge treatment device. The sludge synchronization device outside the sedimentation tank synchronizes the water and sludge precipitation states in the sedimentation tank to the outside of the tank respectively. The image recognition and analysis device uses a high-definition camera to continuously shoot the image of the sludge precipitation and transmit it to the computer. By continuously accumulating databases and algorithms, the sludge images of different liquid levels are converted into sludge concentrations and the simulation calculation results are output. The output results guide the sludge discharge device to discharge sludge and sludge reflux, and guide the sludge treatment device to intelligently match the sludge treatment production scheduling operation. The sludge images of different liquid levels in the above patent are converted into sludge concentrations to guide the production scheduling operation, but simply calculating the sludge concentration obviously cannot meet the needs of dynamic monitoring of multiple tanks such as coagulation tanks, injection tanks, thermal tanks and sedimentation tanks in microwave sedimentation systems. Summary of the invention

[0004] In order to overcome the defects of the prior art, the technical problem to be solved by the present invention is to propose an intelligent control method for a microwave precipitation system based on a sensor group, which can accurately control the amount of preparation added according to the image and turbidity data of the filter tank, thereby improving the purification effect and saving the operating cost of the system.

[0005] To achieve this object, the present invention adopts the following technical solutions: The present invention provides a method for intelligently controlling a microwave precipitation system based on a sensor group, comprising: The sludge concentration in the hydrocyclone is obtained through an online sludge concentration monitor; The sedimentation system controller performs an initial classification of the raw water according to the sludge concentration, and the execution unit of the sedimentation system controller controls the initial addition amount of each filter tank according to the initial classification of the preset scheme; The image sensor acquires image data of at least one filter tank, the turbidity sensor acquires turbidity data at the water outlet of at least one filter tank, and the first intelligent calculation unit of the sedimentation system controller generates a single tank purification capacity parameter according to the image data and the turbidity data of at least one filter tank; The first intelligent calculation unit transmits the single tank purification capacity parameters of each filter tank to the decision unit of the sedimentation system controller, and the decision unit generates a single tank preparation correction amount according to the single tank purification capacity parameters; The execution unit controls the amount of preparation added to each of the filter tanks according to the initial amount of preparation added to each of the filter tanks and the single tank preparation correction amount.

[0006] A further technical solution of the present invention is that each of the filter tanks is a coagulation tank, an injection tank, a thermalization tank and a sedimentation tank, the image sensor includes a first image sensor, a second image sensor and a third image sensor, the first image sensor is used to obtain image data of the coagulation tank, the second image sensor is used to obtain image data of the injection tank, and the third image sensor is used to obtain image data of the sedimentation tank, the turbidity sensor includes a first turbidity sensor, a second turbidity sensor and a third turbidity sensor, the first turbidity sensor is used to obtain turbidity collection data of the first channel between the coagulation tank and the injection tank, the second turbidity sensor is used to obtain turbidity collection data of the second channel between the injection tank and the thermalization tank, and the third turbidity sensor is used to obtain turbidity collection data at the water outlet of the sedimentation tank.

[0007] A further technical solution of the present invention is that, for the coagulation tank, the acquisition unit generates an average equivalent diameter of alum flocs and a mean value of fractal dimensions based on the image data of the first image sensor, the acquisition unit generates effluent turbidity data based on the turbidity acquisition data of the first turbidity sensor, and the first intelligent calculation unit generates a single tank purification capacity parameter of the coagulation tank based on the average equivalent diameter of alum flocs, the mean value of fractal dimensions and the effluent turbidity data; for the injection tank, the acquisition unit generates an average equivalent diameter of alum flocs and a mean value of fractal dimensions based on the image data of the second image sensor, and the acquisition unit generates an average equivalent diameter of alum flocs and a mean value of fractal dimensions based on the turbidity acquisition data of the second turbidity sensor. The data generates outlet turbidity data, and the first intelligent calculation unit generates the single-tank purification capacity parameters of the injection tank according to the average equivalent diameter of the alum flocs, the mean value of the fractal dimension and the outlet turbidity data; for the sedimentation tank, the acquisition unit generates the average equivalent diameter of the alum flocs, the mean value of the fractal dimension and the bottom sedimentation area according to the image data of the third image sensor, and the acquisition unit generates outlet turbidity data according to the turbidity acquisition data of the third turbidity sensor, and the first intelligent calculation unit generates the single-tank purification capacity parameters of the sedimentation tank according to the average equivalent diameter of the alum flocs, the mean value of the fractal dimension, the bottom sedimentation area and the outlet turbidity data.

[0008] A further technical solution of the present invention is that the second intelligent calculation unit of the sedimentation system controller calculates the comprehensive purification capacity parameter Y according to the single tank purification capacity parameter Y1 of the coagulation tank, the single tank purification capacity parameter Y2 of the injection tank and the single tank purification capacity parameter Y3 of the sedimentation tank; The second intelligent calculation unit obtains the sampling mean data X of the SS removal rate through the human-computer interaction unit of the sedimentation system controller, and calculates the correlation coefficient between the comprehensive purification capacity parameter Y and the sampling mean data X. The second intelligent computing unit is based on the correlation coefficient Calculate the real-time SS removal rate.

[0009] A further technical solution of the present invention is that when the decision unit calculates the single-tank preparation correction amount in a query manner, the decision unit obtains the single-tank standard interval and the ideal value from the standard database and determines whether the single-tank purification capacity parameter of a certain filter tank is within the single-tank standard interval. When the single-tank purification capacity parameter is abnormal, the decision unit performs secondary classification of the abnormal degree of the individual filter tank according to the standard deviation of the abnormal value, and generates an ideal value deviation according to the gap between the single-tank purification capacity parameter and the ideal value. The decision unit generates the single-tank preparation correction amount and the secondary interval according to the ideal value of the secondary classification; when the single-tank purification capacity parameter is not abnormal, the decision unit directly generates the single-tank preparation correction amount according to the ideal value of the single-tank standard interval.

[0010] A further technical solution of the present invention is that the sedimentation system controller includes a scheme database for original classification of raw water, and the scheme database contains original classifications and initial addition amounts of multiple preset schemes. When the single-tank purification capacity parameters are abnormal and require secondary classification, the scheme database records the secondary interval corresponding to the secondary classification of the single-tank purification capacity parameters, and associates the original classification and the secondary classification to form a new scheme.

[0011] A further technical solution of the present invention is that when the decision unit calculates the correction amount of a single tank preparation in a learning algorithm, the decision unit uses the image data, the turbidity collection data and the single tank purification capacity as the input layer of the LSTM neural network model, and uses a set of single tank preparation correction amounts as the output layer of the LSTM neural network model, and continuously trains the LSTM neural network model based on the historical data and update data of the standard database, so that the decision unit outputs a set of single tank preparation correction amounts.

[0012] A further technical solution of the present invention is that, for the injection tank, the single tank preparation correction amount includes a high molecular polymer correction amount and a micro sand correction amount, the acquisition unit generates an average equivalent diameter of alum flocs and a mean value of fractal dimension according to the image data of the second image sensor, the acquisition unit generates outlet turbidity data according to the turbidity acquisition data of the second turbidity sensor, and the first intelligent calculation unit generates a single tank purification capacity parameter of the injection tank according to the average equivalent diameter of alum flocs, the mean value of fractal dimension and the outlet turbidity data; The decision unit uses the average equivalent diameter of the alum flowers, the mean value of the fractal dimension, the effluent turbidity data and the single tank purification capacity as the input layer of the LSTM neural network model, and uses the amount of high molecular polymer correction preparation and the amount of micro sand correction preparation as the output layer of the LSTM neural network model.

[0013] The present invention also provides an intelligent control device for a microwave precipitation system based on a sensor group, which is used in the above-mentioned intelligent control method for a microwave precipitation system based on a sensor group, comprising: Online sludge concentration monitor, used to obtain the sludge concentration in the hydrocyclone; Sedimentation system controller, used to calculate and control the amount of preparation added to each filter tank; An image sensor, used to obtain image data of at least one of the filter tanks; The turbidity sensor is used to obtain turbidity data at the water outlet of at least one of the filter tanks.

[0014] The beneficial effects of the present invention are: The intelligent control method of the microwave precipitation system based on the sensor group provided by the present invention can not only set the initial addition amount of each filter tank according to the terminal sludge concentration, but also dynamically correct the preparation addition amount of each filter tank by real-time detection of image data and turbidity collection data by image sensors and turbidity sensors. This not only achieves the purpose of accurately controlling the preparation addition amount to save the operating cost of the system, but also fundamentally solves the problem that the water quality of raw water changes dynamically and the microwave precipitation system cannot respond in time. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 is a flow chart of the fully automatic intelligent control method provided in the specific implementation manner of the present application; Figure 2 It is a functional block diagram of a sedimentation system controller provided in a specific embodiment of the present application; Figure 3 It is a schematic diagram of the structure of the microwave precipitation system provided in the specific implementation manner of the present application.

[0016] In the figure: 1. Online sludge concentration monitor; 2. Sedimentation system controller; 21. Collection unit; 22. Execution unit; 23. First intelligent computing unit; 24. Second intelligent computing unit; 25. Decision unit; 26. Human-computer interaction unit; 51. Coagulation tank; 52. Injection tank; 53. Thermalization tank; 54. Sedimentation tank; 55. Hydrocyclone; 56. First channel; 57. Second channel; 31. First image sensor; 32. Second image sensor; 33. Third image sensor; 41. First turbidity sensor; 42. Second turbidity sensor; 43. Third turbidity sensor. DETAILED DESCRIPTION

[0017] The technical solution of the present invention is further described below with reference to the accompanying drawings and through specific implementation methods.

[0018] like Figures 1 to 3 As shown, the intelligent control method of the microwave precipitation system based on the sensor group provided in this embodiment includes the following steps: Step S00: Obtain the sludge concentration in the hydrocyclone 55 through the online sludge concentration monitor 1. Specifically, set the detection probe of the online sludge concentration monitor 1 at the water inlet of the hydrocyclone 55. When the sludge concentration at the water inlet of the hydrocyclone 55 is low, it means that there are fewer pollutants in the raw water. At this time, the initial addition amount of each preparation in each filter tank should be reduced; when the water inlet of the hydrocyclone 55 is at a high sludge concentration, the initial addition amount of each preparation should be increased. It is worth noting that the sludge concentration at the hydrocyclone 55 itself is usually high, so the original classification below can usually only be roughly classified. The online sludge concentration monitor 1 obtains the sludge concentration of the hydrocyclone 55 after it is turned on and works stably. In addition, the microwave precipitation system first works according to the factory setting parameters, and then executes the above-mentioned microwave precipitation system intelligent control method after working stably.

[0019] Step S10: The sedimentation system controller 2 performs an initial classification of the raw water according to the sludge concentration. For example, the initial classification is divided into light pollution level, medium pollution level and heavy pollution level. The execution unit 22 of the sedimentation system controller 2 controls the initial addition amount of each filter tank according to the original classification of the preset scheme. The light pollution level, medium pollution level and heavy pollution level correspond to three different preset schemes, that is, correspond to different initial addition amounts. It can be seen from the following that each filter tank includes a coagulation tank 51, an injection tank 52, a thermalization tank 53 and a sedimentation tank 54, among which only the coagulation tank 51 and the injection tank 52 require preparations. The coagulation tank 51 needs to add a coagulant, and the injection tank 52 needs to add a polymer and micro-sand. The above initial addition amount includes the addition amount of the coagulant, the polymer and the micro-sand.

[0020] Step S20: The image sensor obtains image data of at least one filter tank. The image sensor is usually arranged on the side wall of the corresponding filter tank and can record the suspended matter in the filter tank in real time. The turbidity sensor obtains turbidity collection data at the water outlet of at least one filter tank. The turbidity sensor is also called a water quality sensor, which is used to measure the degree of water pollution. The first intelligent calculation unit 23 of the sedimentation system controller 2 generates a single tank purification capacity parameter based on the image data and turbidity collection data of at least one filter tank. Since the image data mainly obtains the agglomeration and flocculation of SS (Suspended Solids) pollutants, and the turbidity collection data mainly obtains the particle size of the process turbidity sensor, the simultaneous use of the detection values ​​of the two can improve the monitoring level of SS pollutants. The more agglomerates and the fewer particulate matter in the water, the better the water quality.

[0021] Step S30: The first intelligent calculation unit 23 transmits the single tank purification capacity parameters of each filter tank to the decision unit 25 of the sedimentation system controller 2. The decision unit 25 generates a single tank preparation correction amount based on the single tank purification capacity parameters. The single tank preparation correction amount can effectively avoid the defect of the original classification being relatively rough. The single tank preparation correction amount can be calculated based on the single tank purification capacity parameters. For details, please refer to the two calculation methods provided below.

[0022] Step S40: The execution unit 22 controls the amount of preparation added to each filter tank according to the initial amount of addition to each filter tank and the correction amount of the preparation for each tank. Therefore, the fully automatic intelligent control method provided in this embodiment can dynamically correct the amount of preparation added to each filter tank by detecting the terminal sludge concentration to set the initial amount of addition to each filter tank, and can also detect image data and turbidity collection data in real time through image sensors and turbidity sensors, which not only achieves the purpose of accurately controlling the amount of preparation added to save the operating cost of the system, but also fundamentally solves the problem that the microwave precipitation system cannot respond in time to the dynamic change of raw water quality.

[0023] In a further embodiment, in order to obtain image data and turbidity collection data of part or all of the filter tanks, each filter tank is specifically divided into a coagulation tank 51, an injection tank 52, a thermalization tank 53 and a sedimentation tank 54. The coagulation tank 51, the injection tank 52, the thermalization tank 53 and the sedimentation tank 54 constitute a common microwave precipitation system, wherein the coagulation tank 51 is used to aggregate colloidal particles and tiny suspended matter in the water by adding chemical agents, the injection tank 52 is used to decompose and transform suspended matter, organic matter and other harmful substances in the sewage, the thermalization tank 53 is used to maintain a constant temperature of the biochemical pool to improve the activity and treatment efficiency of microorganisms, and the sedimentation tank 54 is used to remove suspended matter in the water, thereby purifying the water quality. Usually, the coagulation tank 51 needs to add a coagulant, and the injection tank 52 needs to add a high molecular polymer and micro-sand. Therefore, only the first image sensor 31, the second image sensor 32, the first turbidity sensor 41, and the second turbidity sensor 42 can be set, which can also achieve the purpose of calculating the correction amount of the single tank preparation. However, the third image sensor and the third turbidity sensor 43 can test the purification effect of the entire microwave precipitation system. Therefore, the detection data of the third image sensor and the third turbidity sensor 43 can be used as a more accurate correction factor, which is also conducive to the calculation of the real-time SS removal rate below. Therefore, the image sensor includes a first image sensor 31, a second image sensor 32 and a third image sensor 33. The first image sensor 31 is used to obtain image data of the coagulation tank 51, the second image sensor 32 is used to obtain image data of the injection tank 52, and the third image sensor 33 is used to obtain image data of the sedimentation tank 54. The turbidity sensor includes a first turbidity sensor 41, a second turbidity sensor 42 and a third turbidity sensor 43. The first turbidity sensor 41 is used to obtain turbidity collection data of the first channel 56 between the coagulation tank 51 and the injection tank 52, and the second turbidity sensor 42 is used to obtain turbidity collection data of the injection tank 52. The turbidity collection data of the second channel 57 between the tank 52 and the thermalization tank 53, the third turbidity sensor 43 is used to obtain the turbidity collection data at the water outlet of the sedimentation tank 54, so as to detect the image data and turbidity collection data at the above six locations, which can not only calculate the single tank preparation correction amount, but also can test the purification effect of the entire microwave precipitation system. When the first intelligent calculation unit 23 adopts a neural network model, it can also be known that the third image sensor 33 and the detection data of the third image sensor 33 are used as the input layer data for calculating the single tank preparation correction amount corresponding to the coagulation tank 51 and the injection tank 52, thereby improving the accuracy of the single tank preparation correction amount.

[0024] For the coagulation tank 51, the acquisition unit 21 generates an average equivalent diameter of alum flocs and a mean value of fractal dimension based on the image data of the first image sensor 31. The larger the mean value of the fractal dimension is, the higher the complexity of the alum flocs is. The average equivalent diameter of alum flocs and the mean value of the fractal dimension can be calculated based on the distribution of alum flocs on a single picture. Each group of image data usually contains multiple pictures. The average values ​​of the average equivalent diameter of alum flocs and the fractal dimension are taken to obtain the average equivalent diameter of alum flocs and the mean value of the fractal dimension. The acquisition unit 21 generates effluent turbidity data based on the turbidity acquisition data of the first turbidity sensor 41. The turbidity acquisition data is converted into effluent turbidity data after digital-to-analog conversion. The first intelligent calculation unit 23 generates the average equivalent diameter of alum flocs and the fractal dimension based on the above-generated average equivalent diameter of alum flocs and the fractal dimension. The single-tank purification capacity parameters of the coagulation tank 51 are generated by using the mean fractal dimension and the effluent turbidity data. The single-tank purification capacity parameters only need to relatively measure the purification capacity of the coagulation tank 51. Since the average equivalent diameter of alum flowers and the mean fractal dimension are positively correlated with the purification capacity of the coagulation tank 51, and the effluent turbidity data are negatively correlated with the purification capacity, the least squares method can be used to simulate the linear relationship between the single-tank purification capacity parameters and the average equivalent diameter of alum flowers, the mean fractal dimension and the effluent turbidity data to obtain the single-tank purification capacity parameters. Of course, considering the large error of the linear relationship, the neural network model and the curve model can also be used to calculate the single-tank purification capacity parameters, thereby measuring the relative purification effect of the coagulation tank 51. Similarly, for the injection tank 52, the acquisition unit 21 generates the average equivalent diameter of alum flowers and the mean value of the fractal dimension based on the image data of the second image sensor 32, and the acquisition unit 21 generates the outlet turbidity data based on the turbidity acquisition data of the second turbidity sensor 42. The first intelligent calculation unit 23 generates the single tank purification capacity parameters of the injection tank 52 based on the average equivalent diameter of alum flowers, the mean value of the fractal dimension and the outlet turbidity data generated above. The calculation method of the single tank purification capacity parameters of the injection tank 52 is similar. The single tank purification capacity parameters here are used to measure the relative purification effect of the injection tank 52. For the sedimentation tank 54, the acquisition unit 21 generates the average equivalent diameter of alum flakes, the mean value of fractal dimension and the bottom sedimentation area based on the image data of the third image sensor 33. The bottom sedimentation area is first eliminated from the image data, and then the average equivalent diameter of alum flakes and the mean value of fractal dimension are calculated according to the above method. The acquisition unit 21 generates the effluent turbidity data based on the turbidity acquisition data of the third turbidity sensor 43. The first intelligent calculation unit 23 generates the single tank purification capacity parameters of the sedimentation tank 54 based on the above-generated average equivalent diameter of alum flakes, the mean value of fractal dimension, the bottom sedimentation area and the effluent turbidity data. The single tank purification capacity parameters here are used to measure the relative purification effect of the sedimentation tank 54.

[0025] Traditionally, the SS removal rate usually requires collecting clarified water and conducting multiple groups of comparative experiments with raw water, and then obtaining the SS removal rate through statistical analysis, that is, manual comparative experiments are required to draw conclusions, which is obviously not suitable for the automated control of the microwave sedimentation system. In a further embodiment, the second intelligent calculation unit 24 of the sedimentation system controller 2 calculates the comprehensive purification capacity parameter Y based on the single tank purification capacity parameter Y1 of the coagulation tank 51, the single tank purification capacity parameter Y2 of the injection tank 52, and the single tank purification capacity parameter Y3 of the sedimentation tank 54. The second intelligent calculation unit 24 obtains the sampling mean data X of the SS removal rate through the human-computer interaction unit 26 of the sedimentation system controller 2, that is, the SS removal rate is usually positively correlated with the comprehensive purification capacity parameter Y. When the corresponding sampling mean data X of the SS removal rate is manually entered over a period of time, the correlation coefficient between the comprehensive purification capacity parameter Y and the sampling mean data X can be calculated based on the comprehensive purification capacity parameter Y and the sampling mean data X. , which is also calculated by the least squares method value, so that the second intelligent computing unit 24 can Calculate the real-time SS removal rate. When the calculated real-time SS removal rate is abnormal, the precipitation system controller 2 can issue an early warning through the human-computer interaction unit 26, thereby greatly improving the intelligence level of the microwave precipitation system.

[0026] Due to the real-time calculation method, the microwave precipitation system will have relatively large movement fluctuations, which is mainly due to the large fluctuation of SS in the raw water. This will cause the microwave precipitation system to often need to adjust the amount of preparation or the speed of the agitator, which is obviously not conducive to the stable operation of the microwave precipitation system. In a further embodiment, when the decision unit 25 calculates the single tank preparation correction amount in a query manner, the decision unit 25 obtains the single tank standard interval and the ideal value from the standard database and determines whether the single tank purification capacity parameter of a certain filter tank is within the single tank standard interval. The single tank purification capacity parameter of a certain filter tank fluctuates continuously when the calculated value is within a period of time. When the single tank purification capacity parameter is abnormal, the decision unit 25 performs secondary classification of the abnormal degree of the single filter tank according to the standard deviation of the abnormal value, and generates an ideal value deviation according to the gap between the single tank purification capacity parameter and the ideal value, and then uses the degree of the ideal value deviation to find the secondary interval, and the secondary interval also has a corresponding ideal value. For example: the ideal value of a single tank standard interval is Z, and its single tank standard interval is Z 1. If: the ideal value deviation is Z+2.4, Z+2.4-1=Z+1.4, which means that the secondary interval of the secondary classification is the adjacent interval above the single tank standard interval. At this time, the ideal value of this adjacent interval is used as the ideal value of the secondary classification, and the corresponding adjacent interval is the secondary interval. When the ideal value deviation is Z+3.4, Z+3.4-1=Z+2.4, which means that the secondary interval of the secondary classification is the second interval above the single tank standard interval. At this time, the ideal value of the second interval is used as the ideal value of the secondary classification. By analogy, the decision unit 25 generates the single tank preparation correction amount and the secondary interval according to the ideal value of the secondary classification. When the single tank purification capacity parameters are not abnormal, the decision unit directly generates the single tank preparation correction amount according to the ideal value of the single tank standard interval. The microwave precipitation system uses the ideal value for operation instead of directly calculating the detection value of the single tank purification capacity parameter. This is mainly because the detection mean value of the single tank purification capacity parameter is usually a real-time fluctuating curve. In this way, the microwave precipitation system needs to frequently adjust the amount of preparation added to each filter tank, which is obviously not conducive to the stable operation of the microwave precipitation system. Therefore, the above method is adopted to make the equipment for the preparation addition amount equivalent to several fixed working modes, each mode corresponds to an ideal value, and only when it deviates from the single tank standard interval corresponding to the ideal value, the new mode is switched, which greatly improves the stability of the equipment for the preparation addition amount. In addition, for the injection tank 52, the single tank preparation correction amount includes two types: polymer correction amount and micro sand correction amount. At this time, it is necessary to select a factor in the polymer correction amount or micro sand correction amount according to the pollution situation of the raw water to anchor the ideal value, so as to facilitate the calculation of the single tank preparation correction amount in a query manner, thereby simplifying the calculation.

[0027] In a further embodiment, the sedimentation system controller 2 includes a scheme database for original classification of raw water, and the scheme database contains original classifications and initial addition amounts of multiple preset schemes. When the single-tank purification capacity parameters are abnormal and require secondary classification, the case database records the secondary interval corresponding to the secondary classification of the single-tank purification capacity parameters, and associates the original classification and the secondary classification to form a new scheme. Since the raw water in a certain area has certain inherent characteristics over a long period of time, its single-tank purification capacity parameters are relatively stable after monitoring for a period of time, so that the corresponding new scheme is relatively determined, and the sedimentation system controller 2 can directly call the new scheme to run the microwave sedimentation system to improve decision-making efficiency.

[0028] Embodiment 2 The intelligent control method of the microwave precipitation system based on the sensor group provided in the second embodiment includes the following steps: Step S00: Obtain the sludge concentration in the hydrocyclone 55 through the online sludge concentration monitor 1. Specifically, the detection probe of the online sludge concentration monitor 1 is set at the water inlet of the hydrocyclone 55. When the water inlet of the hydrocyclone 55 is at a low sludge concentration, it means that there are fewer pollutants in the raw water. At this time, the initial addition amount of each preparation in each filter tank should be reduced; when the water inlet of the hydrocyclone 55 is at a high sludge concentration, the initial addition amount of each preparation should be increased. It is worth noting that the sludge concentration at the hydrocyclone 55 itself is usually high, so the original classification below can usually only be roughly classified. The online sludge concentration monitor 1 obtains the sludge concentration of the hydrocyclone 55 after it is turned on and works stably. In addition, the microwave precipitation system first works according to the factory setting parameters, and then executes the above-mentioned fully automatic intelligent control method after working stably.

[0029] Step S10: The sedimentation system controller 2 performs an initial classification of the raw water according to the sludge concentration. For example, the initial classification is divided into light pollution level, medium pollution level and heavy pollution level. The execution unit 22 of the sedimentation system controller 2 controls the initial addition amount of each filter tank according to the original classification of the preset scheme. The light pollution level, medium pollution level and heavy pollution level correspond to three different preset schemes, that is, correspond to different initial addition amounts. It can be seen from the following that each filter tank includes a coagulation tank 51, an injection tank 52, a thermalization tank 53 and a sedimentation tank 54, among which only the coagulation tank 51 and the injection tank 52 require preparations. The coagulation tank 51 needs to add a coagulant, and the injection tank 52 needs to add a polymer and micro-sand. The above initial addition amount includes the addition amount of the coagulant, the polymer and the micro-sand.

[0030] Step S20: The image sensor acquires image data of at least one filter tank, which is usually arranged on the side wall of the corresponding filter tank and can record the suspended matter in the filter tank in real time. The turbidity sensor acquires turbidity data at the water outlet of at least one filter tank. The turbidity sensor is also called a water quality sensor, which is used to measure the degree of water pollution. The first intelligent calculation unit 23 of the sedimentation system controller 2 generates a single tank purification capacity parameter based on the image data and turbidity acquisition data of at least one filter tank. Since the image data mainly obtains the agglomeration and flocculation of SS (Suspended Solids) pollutants, and the turbidity acquisition data mainly obtains the particle size of the process turbidity sensor, the simultaneous use of the detection values ​​of the two can improve the monitoring level of SS pollutants. The more agglomerates and the fewer particulate matter in the water, the better the water quality.

[0031] Step S30: The first intelligent calculation unit 23 transmits the single tank purification capacity parameters of each filter tank to the decision unit 25 of the sedimentation system controller 2. The decision unit 25 generates a single tank preparation correction amount based on the single tank purification capacity parameters. The single tank preparation correction amount can effectively avoid the defect of the original classification being relatively rough. The single tank preparation correction amount can be calculated based on the single tank purification capacity parameters. For details, please refer to the two calculation methods provided below.

[0032] Step S40: The execution unit 22 controls the amount of preparation added to each filter tank according to the initial amount of addition to each filter tank and the correction amount of the preparation for each tank. Therefore, the fully automatic intelligent control method provided in this embodiment can dynamically correct the amount of preparation added to each filter tank by detecting the terminal sludge concentration to set the initial amount of addition to each filter tank, and can also detect image data and turbidity collection data in real time through image sensors and turbidity sensors, which not only achieves the purpose of accurately controlling the amount of preparation added to save the operating cost of the system, but also fundamentally solves the problem that the microwave precipitation system cannot respond in time to the dynamic change of raw water quality.

[0033] The difference between the second embodiment and the first embodiment is that: When the decision unit 25 calculates the single-tank preparation correction amount by means of a learning algorithm: the decision unit 25 uses the image data, turbidity collection data and single-tank purification capacity as the input layer of the LSTM neural network model, and uses a set of single-tank preparation correction amounts as the output layer of the LSTM neural network model, and continuously trains the LSTM neural network model based on the historical data and update data of the standard database, so that the decision unit 25 outputs a set of single-tank preparation correction amounts and helps the LSTM neural network model to optimize the accuracy of the single-tank preparation correction amounts. It is worth noting that the use of this method to calculate the tank correction preparation amount has already taken into account the single-tank standard range and ideal value in the LSTM neural network model, that is, it can be considered to be directly associated with the ideal values ​​corresponding to several fixed working modes of the equipment for the preparation addition amount. This method is more accurate after long-term training. In a further embodiment, for the injection tank 52, the single tank preparation correction amount includes a high molecular polymer correction amount and a micro sand correction amount. The acquisition unit 21 generates an average equivalent diameter of alum flowers and a mean value of fractal dimension based on the image data of the second image sensor 32. The acquisition unit 21 generates outlet turbidity data based on the turbidity collection data of the second turbidity sensor 42. The first intelligent calculation unit 23 generates a single tank purification capacity parameter of the injection tank 52 based on the average equivalent diameter of alum flowers, the mean value of fractal dimension and the outlet turbidity data. The decision unit 25 uses the average equivalent diameter of alum flowers, the mean value of fractal dimension, the outlet turbidity data and the single tank purification capacity as the input layer of the LSTM neural network model, and uses the high molecular polymer correction amount and the micro sand correction amount as the output layer of the LSTM neural network model. By utilizing the characteristics of the fuzzy algorithm of the LSTM neural network model, two variables can be calculated at the same time and several modes of fixed operation of the equipment corresponding to multiple preparation addition amounts can be taken into account, so that the calculation is more accurate.

[0034] Embodiment 3 like Figures 2 to 3 As shown, an intelligent control device for a microwave precipitation system based on a sensor group provided in Example 3 is used for the above-mentioned intelligent control method of a microwave precipitation system based on a sensor group. The intelligent control device for a microwave precipitation system based on a sensor group includes an online sludge concentration monitor 1, an image sensor, a turbidity sensor and a precipitation system controller 2, wherein the online sludge concentration monitor 1 is used to obtain the sludge concentration in the hydrocyclone 55, the precipitation system controller 2 is used to calculate and control the amount of preparation added to each filter tank, the image sensor is used to obtain image data of at least one filter tank, and the turbidity sensor is used to obtain turbidity collection data at the water outlet of at least one filter tank. The online sludge concentration monitor 1, the image sensor, and the turbidity sensor are all electrically connected to the precipitation system controller 2, thereby transmitting the sludge concentration, image data and turbidity collection data to the precipitation system controller 2, so as to control the amount of preparation added to each filter tank.

[0035] The sedimentation system controller 2 includes an acquisition unit 21, an execution unit 22, a first intelligent calculation unit 23, a second intelligent calculation unit 24 and a decision unit 25. The acquisition unit 21 is used to obtain image data and turbidity acquisition data. The acquisition unit 21 acquires image data and turbidity acquisition data and then outputs them to the decision unit 25, which is then indirectly output to the first intelligent calculation unit 23 for calculation through the decision unit 25. That is, the decision unit 25 is the central unit of other units, and is electrically connected to other units and can transmit data. The first intelligent calculation unit 23 is used to generate single-tank purification capacity parameters. The first intelligent calculation unit 23 transmits the single-tank purification capacity parameters of each filter tank to the decision unit 25. The decision unit 25 is used to generate a single-tank preparation correction amount and generate a decision signal based on the single-tank preparation correction amount and transmit it to the execution unit 22 for execution. The execution unit 22 is used to control the preparation addition amount of each filter tank. The decision unit 25 transmits the single-tank purification capacity parameters of each filter tank to the second intelligent calculation unit 24. The second intelligent calculation unit 24 is used to calculate the correlation coefficient and real-time SS removal rate.

[0036] The microwave precipitation system includes a coagulation tank 51, an injection tank 52, a thermalization tank 53, a precipitation tank 54 and a hydrocyclone 55. The image sensor includes a first image sensor 31, a second image sensor 32 and a third image sensor 33. The first image sensor 31 is used to obtain image data of the coagulation tank 51, the second image sensor 32 is used to obtain image data of the injection tank 52, and the third image sensor 33 is used to obtain image data of the precipitation tank 54. The turbidity sensor includes a first turbidity sensor 41, a second turbidity sensor 42 and a third turbidity sensor 43. The first turbidity sensor 41 is used to obtain turbidity collection data of the first channel 56 between the coagulation tank 51 and the injection tank 52, the second turbidity sensor 42 is used to obtain turbidity collection data of the second channel 57 between the injection tank 52 and the thermalization tank 53, and the third turbidity sensor 43 is used to obtain turbidity collection data at the water outlet of the precipitation tank 54.

[0037] The present invention is described by preferred embodiments, and those skilled in the art will appreciate that various changes or equivalent substitutions may be made to these features and embodiments without departing from the spirit and scope of the present invention. The present invention is not limited to the specific embodiments disclosed herein, and other embodiments falling within the claims of this application are within the scope of protection of the present invention.

Claims

1. A method for intelligent control of a microwave precipitation system based on a sensor group, characterized in that: include: The sludge concentration in the hydrocyclone is obtained through an online sludge concentration monitor; The sedimentation system controller performs an initial classification of the raw water according to the sludge concentration, and the execution unit of the sedimentation system controller controls the initial addition amount of each filter tank according to the initial classification of the preset scheme; The image sensor acquires image data of at least one filter tank, the turbidity sensor acquires turbidity data at the water outlet of at least one filter tank, and the first intelligent calculation unit of the sedimentation system controller generates a single tank purification capacity parameter according to the image data and the turbidity data of at least one filter tank; The first intelligent calculation unit transmits the single tank purification capacity parameters of each filter tank to the decision unit of the sedimentation system controller, and the decision unit generates a single tank preparation correction amount according to the single tank purification capacity parameters; The execution unit controls the amount of preparation added to each of the filter tanks according to the initial amount of preparation added to each of the filter tanks and the single tank preparation correction amount.

2. The intelligent control method of microwave precipitation system based on sensor group according to claim 1 is characterized in that: Each of the filtering tanks is a coagulation tank, an injection tank, a thermal tank and a sedimentation tank; The image sensor comprises a first image sensor, a second image sensor and a third image sensor, wherein the first image sensor is used to obtain image data of the coagulation tank, the second image sensor is used to obtain image data of the injection tank, and the third image sensor is used to obtain image data of the sedimentation tank; The turbidity sensor includes a first turbidity sensor, a second turbidity sensor and a third turbidity sensor. The first turbidity sensor is used to obtain turbidity collection data of the first channel between the coagulation tank and the injection tank, the second turbidity sensor is used to obtain turbidity collection data of the second channel between the injection tank and the thermalization tank, and the third turbidity sensor is used to obtain turbidity collection data at the water outlet of the sedimentation tank.

3. The intelligent control method of microwave precipitation system based on sensor group according to claim 2 is characterized in that: For the coagulation tank, the acquisition unit generates an average equivalent diameter of alum flocs and a mean value of fractal dimensions based on the image data of the first image sensor, the acquisition unit generates effluent turbidity data based on the turbidity acquisition data of the first turbidity sensor, and the first intelligent calculation unit generates a single tank purification capacity parameter of the coagulation tank based on the average equivalent diameter of alum flocs, the mean value of fractal dimensions and the effluent turbidity data; For the injection tank, the acquisition unit generates an average equivalent diameter of alum flowers and a mean value of fractal dimensions based on the image data of the second image sensor, the acquisition unit generates outlet turbidity data based on the turbidity acquisition data of the second turbidity sensor, and the first intelligent calculation unit generates a single tank purification capacity parameter of the injection tank based on the average equivalent diameter of alum flowers, the mean value of fractal dimensions and the outlet turbidity data; For the sedimentation tank, the acquisition unit generates an average equivalent diameter of alum flakes, a mean value of the fractal dimension, and a bottom sedimentation area based on the image data of the third image sensor, the acquisition unit generates outlet turbidity data based on the turbidity acquisition data of the third turbidity sensor, and the first intelligent calculation unit generates a single tank purification capacity parameter of the sedimentation tank based on the average equivalent diameter of alum flakes, the mean value of the fractal dimension, the bottom sedimentation area, and the outlet turbidity data.

4. The intelligent control method of microwave precipitation system based on sensor group according to claim 3 is characterized in that: The second intelligent calculation unit of the sedimentation system controller calculates the comprehensive purification capacity parameter Y according to the single tank purification capacity parameter Y1 of the coagulation tank, the single tank purification capacity parameter Y2 of the injection tank and the single tank purification capacity parameter Y3 of the sedimentation tank; The second intelligent calculation unit obtains the sampling mean data X of the SS removal rate through the human-computer interaction unit of the sedimentation system controller, and calculates the correlation coefficient between the comprehensive purification capacity parameter Y and the sampling mean data X. The second intelligent computing unit is based on the correlation coefficient Calculate the real-time SS removal rate.

5. The intelligent control method of microwave precipitation system based on sensor group according to claim 1 is characterized in that: When the decision unit calculates the correction amount of a single tank preparation in a query mode: The decision unit obtains the single-tank standard interval and ideal value from the standard database and determines whether the single-tank purification capacity parameter of a certain filter tank is within the single-tank standard interval. When the single-tank purification capacity parameter is abnormal, the decision unit performs secondary classification of the abnormality of the individual filter tank according to the standard deviation of the abnormal value, and generates an ideal value deviation according to the gap between the single-tank purification capacity parameter and the ideal value. The decision unit generates a single-tank preparation correction amount and a secondary interval according to the ideal value of the secondary classification; when the single-tank purification capacity parameter is not abnormal, the decision unit directly generates a single-tank preparation correction amount according to the ideal value of the single-tank standard interval.

6. The intelligent control method of microwave precipitation system based on sensor group according to claim 5 is characterized in that: The sedimentation system controller includes a scheme database for original classification of raw water, wherein the scheme database contains original classifications and initial addition amounts of multiple preset schemes. When the single-tank purification capacity parameter is abnormal and requires secondary classification, the scheme database records the secondary interval corresponding to the secondary classification of the single-tank purification capacity parameter, and associates the original classification and the secondary classification to form a new scheme.

7. The intelligent control method of microwave precipitation system based on sensor group according to claim 2 is characterized in that: When the decision-making unit calculates the correction amount of a single tank preparation in a learning algorithm: The decision unit uses the image data, the turbidity collection data and the single-tank purification capacity as the input layer of the LSTM neural network model, uses a set of single-tank preparation correction amounts as the output layer of the LSTM neural network model, and continuously trains the LSTM neural network model based on the historical data and update data of the standard database, so that the decision unit outputs a set of single-tank preparation correction amounts.

8. The intelligent control method of microwave precipitation system based on sensor group according to claim 7 is characterized in that: For the injection tank, the single tank preparation correction amount includes a high molecular polymer correction amount and a micro sand correction amount. The acquisition unit generates an average equivalent diameter of alum flowers and a mean value of fractal dimension according to the image data of the second image sensor. The acquisition unit generates outlet turbidity data according to the turbidity acquisition data of the second turbidity sensor. The first intelligent calculation unit generates a single tank purification capacity parameter of the injection tank according to the average equivalent diameter of alum flowers, the mean value of fractal dimension and the outlet turbidity data. The decision unit uses the average equivalent diameter of the alum flowers, the mean value of the fractal dimension, the effluent turbidity data and the single tank purification capacity as the input layer of the LSTM neural network model, and uses the amount of high molecular polymer correction preparation and the amount of micro sand correction preparation as the output layer of the LSTM neural network model.

9. An intelligent control device for a microwave precipitation system based on a sensor group, used in the intelligent control method for a microwave precipitation system based on a sensor group according to any one of claims 1 to 8, characterized in that: include: Online sludge concentration monitor, used to obtain the sludge concentration in the hydrocyclone; Sedimentation system controller, used to calculate and control the amount of preparation added to each filter tank; An image sensor, used to obtain image data of at least one of the filter tanks; The turbidity sensor is used to obtain turbidity data at the water outlet of at least one of the filter tanks.

Images