JP7704595B2 - Floc state control device, sludge treatment facility, floc state control method, and program - Google Patents

Description

The present invention relates to a flock state control device, a sludge treatment facility, a flock state control method, and a program.

Solid-liquid separation devices are used in facilities such as water treatment plants for the purpose of concentrating or dehydrating materials such as sludge. One example of such a solid-liquid separation device is a concentrator that supplies the material, which is sludge to which a flocculant has been added, into a cylindrical filter screen, rotates a screw disposed in the filter screen to filter the material, and discharges the filtrate. In general, increasing the amount of flocculant added tends to promote solid-liquid separation of the material.
In order to control the amount of coagulant injected, for example, there is a method of capturing an image of the sludge in a coagulation mixing tank or the like and controlling the amount of coagulant injected depending on the formation state of the coagulated sludge (e.g., flocs) (e.g., the area of the flocs).

For example, the following Patent Document 1 discloses a technology for evaluating flocculated sludge by photographing flocs of suspended matter in a piping line emerging from a flocculation mixing tank. In this technology, for example, suspended matter contained in the raw liquid supply pipe of a dehydrator is photographed, and the average area of several flocs or the average area per floc is calculated from the binary image of the flocs shown in the photographed image, and the appropriate amount of flocculant to be injected is calculated by comparing this average area with a preset reference area for the flocs.

Patent No. 4238983

However, with the technology described in Patent Document 1, if a floc in the captured image overlaps with another floc, the area of each floc cannot be calculated accurately. In this case, the amount of coagulant to be injected is not calculated properly, and the state of the flocs cannot be controlled properly. In order to reduce the overlapping of flocs in the captured image, it is possible to dilute the concentration of the sludge, but this requires the installation of a separate dilution device. Alternatively, in order to reduce the overlapping of flocs in the captured image, it is necessary to install a television camera in the raw liquid piping line and capture the size of the flocs in a piping where there is little overlapping of flocs and the flow rate drops to an appropriate level without pulsation.

In consideration of the above problems, the object of the present invention is to provide a flock state control device, sludge treatment equipment, flock state control method, and program that can appropriately control the state of flocks at a location where it is easier to photograph than in the downstream piping and where adjustments of the coagulant can be made earlier, without diluting the sludge concentration or adding new equipment to reduce the overlapping of flocks.

In order to solve the above-mentioned problems, a floc state control device according to one aspect of the present invention includes: a flocculated sludge image acquisition unit that acquires a flocculated sludge image obtained by capturing an image of flocculated sludge containing flocs formed by flocculating sludge with a flocculating agent in a flocculation mixing tank; an area determination unit that determines gap areas and other areas in the acquired flocculated sludge image using a machine learning model that determines gap areas and other areas between a plurality of the flocs in the flocculated sludge image; a gap area acquisition unit that calculates an area of the area determined to be the gap area as a gap area indicating the area of the gap; and a chemical injection rate control unit that controls a chemical injection rate of the flocculating agent so that the acquired gap area becomes a target area , and the machine learning model is a model that has been machine-learned to determine whether each pixel of an input flocculated sludge image is a gap area or a other area using semantic segmentation .

The sludge treatment facility according to one aspect of the present invention is equipped with a flock state control device.

A flock state control method according to one aspect of the present invention includes: a flocked sludge image acquisition step in which a flocked sludge image acquisition unit acquires a flocked sludge image containing flocks formed by flocculating sludge with a flocculating agent in a flocculation mixing tank ; an area determination step in which an area determination unit determines gap areas and other areas in the acquired flocked sludge image using a machine learning model that determines gap areas and other areas between a plurality of the flocks in the flocked sludge image; a gap area acquisition step in which a gap area acquisition unit calculates the area of the area determined to be the gap area as a gap area indicating the area of the gap; and a drug injection rate control step in which a drug injection rate control unit controls the drug injection rate of the flocking agent so that the acquired gap area becomes a target area, wherein the machine learning model is a model that has been machine-learned to determine whether each pixel of an input flocked sludge image is a gap area or a other area using semantic segmentation .

A program according to one aspect of the present invention causes a computer to function as: a flocculated sludge image acquisition means for acquiring a flocculated sludge image obtained by capturing an image of flocculated sludge including flocs formed by flocculating sludge with a flocculant in a flocculation mixing tank; an area determination means for determining gap areas and other areas in the acquired flocculated sludge image using a machine learning model for determining gap areas and other areas between a plurality of flocs in the flocculated sludge image; a gap area acquisition means for calculating the area of the area determined to be the gap area as a gap area indicating the area of the gap; and a chemical injection rate control means for controlling the chemical injection rate of the flocculant so that the acquired gap area becomes a target area, wherein the machine learning model is a model that has been machine-learned to determine whether each pixel of an input flocculated sludge image is a gap area or a other area using semantic segmentation .

The present invention makes it possible to control the state of flocs appropriately and earlier without diluting the sludge concentration or adding new equipment.

1 is a diagram showing an example of the configuration of a sludge treatment facility according to an embodiment of the present invention. FIG. 2 is a diagram illustrating an example of a functional configuration of a PC according to an embodiment of the present invention. FIG. 2 is a diagram showing an example of an image of flocculated sludge according to an embodiment of the present invention. FIG. 2 is a diagram showing an example of an image of flocculated sludge according to an embodiment of the present invention. FIG. 2 is a diagram showing an example of an image of flocculated sludge according to an embodiment of the present invention. FIG. 4 is a diagram showing an example of the relationship between the drug injection rate of a flocculant and the gap area according to the embodiment of the present invention. FIG. 11 is a diagram showing an example of the relationship between gap area and heating efficiency according to an embodiment of the present invention. 10 is a flowchart showing an example of a processing flow in a PC according to an embodiment of the present invention. FIG. 1 is a diagram showing a specific example according to an embodiment of the present invention.

The following describes in detail an embodiment of the present invention with reference to the drawings.

<1. Configuration of sludge treatment facility>
First, the configuration of a sludge treatment facility according to this embodiment will be described with reference to Fig. 1. Fig. 1 is a diagram showing an example of the configuration of a sludge treatment facility according to an embodiment of the present invention.

As shown in FIG. 1, the sludge treatment equipment of this embodiment includes, for example, a flocculation mixing tank 1, a thickener 2, a dehydrator 3, a camera 10, a PC (Personal Computer) 20, and a PLC (Programmable Logic Controller).

(1) Coagulation mixing tank 1
The flocculation mixing tank 1 is a facility for flocculating sludge by adding a flocculant B to sludge A, such as sewage sludge including mixed raw sludge generated and supplied from a sewage treatment plant.
As shown in Fig. 1, sludge A is supplied to a flocculation mixing tank 1 from a high-speed mixer 1A. A flocculant B (e.g., a polymer flocculant) is supplied to the high-speed mixer 1A by a pump 1B and added to and mixed with the sludge A. The high-speed mixer 1A has a stirring blade 1b attached to a rotating shaft 1a aligned along the central axis of the high-speed mixer 1A, and a motor 1c is provided on the upper part. In the high-speed mixer 1A, the rotating shaft 1a and the stirring blade 1b are rotated by the motor 1c, whereby the sludge A and the flocculant B are stirred and mixed.
The coagulation mixing tank 1 is provided with a bottomed cylindrical coagulation tank 1C having a vertically extending central axis into which sludge A mixed with coagulant B is supplied from the bottom and held. This coagulation tank 1C is provided with agitator 1D, which has a rotor 1d aligned along the central axis of the coagulation tank 1C and an agitator blade 1e attached to it, and which agitates and mixes the sludge A and the coagulant B by rotating the rotor 1d and the agitator blade 1e driven by a motor 1f provided at the top of the coagulation tank 1C. The agitated sludge C, which contains flocs that have been agitated and mixed with the coagulant B by the agitator 1D, is extracted from the top of the coagulation tank 1C and supplied to the thickener 2.

(2) Concentrator 2
The concentrator 2 is a facility for concentrating the flocculated sludge C whose solid content has been flocculated to a certain extent by the flocculation mixing tank 1. The concentrator 2 in this embodiment is a vertical filtration concentrator, and is provided with a bottomed cylindrical thickening tank 2A having a central axis extending in the vertical direction similar to that of the flocculation tank 1C, in which the flocculated sludge C supplied from the flocculation mixing tank 1 is held, and the flocculated sludge C is supplied into the thickening tank 2A from the top of the thickening tank 2A. However, the body of the thickening tank 2A is a thickening filtration screen 2a formed of a wedge wire, punching metal, or the like, and the outer periphery of the thickening filtration screen 2a is a jacket-shaped filtrate chamber 2b.

The thickening tank 2A is provided with a conveying means 2B in which a screw 2d is attached to a rotating shaft 2c aligned along the central axis of the thickening tank 2A, and the rotating shaft 2c and the screw 2d are rotated by a motor 2e provided at the top of the thickening tank 2A to convey the flocculated sludge C. The flocculated sludge C supplied from the top of the thickening tank 2A is conveyed downward by the conveying means 2B, and the water is separated and concentrated by the thickening filter screen 2a. The flocculated sludge C is then extracted from the bottom of the thickening tank 2A and supplied to the thickened sludge supply channel 4 as thickened sludge D.

The bottoms of the thickening tank 2A and the coagulation tank 1C are formed in a truncated cone shape that narrows toward the bottom. The water separated from the coagulated sludge C by the thickening filtration screen 2a is stored in the filtrate chamber 2b and treated as return water E.

Furthermore, in this embodiment, concentrated hot water F having a temperature in the range of 50°C or more and less than 100°C, preferably in the range of 60°C or more and 90°C or less, is supplied to the concentrator 2 by the pump 2C and mixed with the flocculated sludge C supplied from the flocculation mixing tank 1. The concentrated hot water F may be supplied to the concentrator 2 by, for example, a concentrated hot water supply means (not shown), or the wastewater J discharged from the drain pipe 3H of the dehydrator 3 may be supplied as the concentrated hot water F.

Here, the rotating shaft 2c of the conveying means 2B of the concentrator 2 is hollow and cylindrical, and a large number (multiple) of through holes are formed in the cylindrical wall of the rotating shaft 2c. The concentrated hot water F is supplied from the lower end of the rotating shaft 2c into the rotating shaft 2c and ejected from the through holes, and is then supplied to and mixed with the flocculated sludge C that is supplied from the flocculation mixing tank 1 to the concentration tank 2A and held therein. As a result, in this embodiment, the flocculated sludge C is heated in the concentrator 2, the proteins are thermally denatured, and the retained moisture is separated and discharged as return water E together with the concentrated hot water F, and the concentrated sludge D is concentrated to a predetermined concentration.

The thickened sludge D thus thickened is supplied to the thickened sludge supply line 4, which is provided with a pump 4A that sends the thickened sludge D to the dehydrator 3, and a high-speed mixer 4B is provided between the pump 4A and the dehydrator 3. A flocculant G, such as an inorganic flocculant such as polyferric sulfate (PFS) or a polymer flocculant, is supplied to the high-speed mixer 4B by a pump 4C and added to and mixed with the thickened sludge D.

(3) Dehydrator 3
The dehydrator 3 is a facility for dehydrating the concentrated sludge D, which has been concentrated by the concentrator 2 and to which a flocculant G has been added and mixed. In the dehydrator 3, a filter screen 3B for filtering the concentrated sludge D is disposed in a casing 3A, and the concentrated sludge D is supplied to a first space 3A1 among a plurality of spaces in the casing 3A divided by the filter screen 3B. The dehydrator 3 in this embodiment is a vertical screw press, and further includes an inner filter screen 3a as the second filter screen 3B, which is disposed in the casing 3A in a cylindrical or conical shape centered on an axis extending in the vertical direction coaxially with the casing 3A. The dehydrator 3 also includes an outer filter screen 3b, which is disposed in the casing 3A in a cylindrical or conical shape coaxially with the inner filter screen 3a, with a gap therebetween outside the inner filter screen 3a. The dehydrator 3 also has a ribbon screw 3d that is housed between the inner filtration screen 3a and the outer filtration screen 3b in a spiral shape twisted around the above-mentioned axis, and is rotated relative to the inner filtration screen 3a and the outer filtration screen 3b around the above-mentioned axis by a motor 3c.

The space between the inner filter screen 3a and the outer filter screen 3b is the first space 3A1, into which the concentrated sludge D is supplied, and the space inside the inner filter screen 3a and the space inside the casing 3A outside the outer filter screen 3b are the second space 3A2. Dehydration hot water H at a temperature in the range of 50°C to 100°C, preferably 60°C to 90°C, is supplied to this second space 3A2 by a pump 3P from a dehydration hot water supply means (not shown). The inner filter screen 3a and the outer filter screen 3b are also made of wedge wire, punched metal, or the like.

The casing 3A is cylindrical with a bottom centered on the axis. The thickened sludge D is supplied to the first space 3A1 from the bottom of the casing 3A via a supply pipe 3f connected to a circular connecting plate 3e that connects the bottoms of the inner filtration screen 3a and the outer filtration screen 3b. The supplied thickened sludge D is transported upward by the relative rotation of the ribbon screw 3d, while the water is separated by the inner filtration screen 3a and the outer filtration screen 3b.

In addition, a circular plate-shaped base plate 3C is disposed at the top of the casing 3A, and the outer filtration screen 3b is attached and fixed to the inner periphery of this base plate 3C. Furthermore, a lid body 3D is disposed at the upper opening of the casing 3A above this base plate 3C, and the inner filtration screen 3a is attached and fixed to this lid body 3D, and the motor 3c is disposed on this lid body 3D and rotates the ribbon screw 3d via a cylindrical screw support covering the upper part of the inner filtration screen 3a. In this embodiment, the inner filtration screen 3a and the outer filtration screen 3b are fixed to the casing in this way, and the ribbon screw 3d is rotated by the motor 3c, but conversely, the ribbon screw 3d may be fixed and the inner filtration screen 3a and the outer filtration screen 3b may be rotated, or the ribbon screw 3d and the inner filtration screen 3a and the outer filtration screen 3b may be rotated in opposite directions.

The upper space in the casing 3A between the base plate 3C and the lid 3D is the discharge chamber 3E, and the annular upper opening of the first space 3A1 in the discharge chamber 3E is the discharge outlet 3F. A squeezing ring 3G having a truncated cone-shaped outer surface centered on the axis that points upward toward the outer periphery is disposed in the discharge outlet 3F. The dehydrated sludge I, which is dewatered from the concentrated sludge D as it is transported upward through the first space 3A1 by the ribbon screw 3d, flows out of the discharge chamber 3E from the discharge outlet 3F while being squeezed by the squeezing ring 3G and discharged.

In this embodiment, the dehydrated hot water H is also supplied from the bottom of the casing 3A to the second space 3A2 in the casing 3A. The dehydrated hot water H thus supplied to the second space 3A2 heats the concentrated sludge D in the first space 3A1, which causes the proteins in the concentrated sludge D to be thermally denatured, separating the water that had been retained therein. The dehydrated hot water H is then filtered by the inner filtration screen 3a and the outer filtration screen 3b, and discharged as wastewater J from the drain pipe 3H extending from the second space 3A2 together with the dehydrated hot water H that has been cooled by heating the concentrated sludge D.

(4) Camera 10
The camera 10 captures an image (still image or video image) of the flocculated sludge C containing flocs formed by flocculation of the sludge A by the flocculant B in the coagulation tank 1C. The camera 10 is communicably connected to the PC 20, and transmits the captured image of the flocculated sludge C (hereinafter also referred to as the “flocculated sludge image”) to the PC 20.
The camera 10 may be provided at any position as long as it can capture an image of the flocculated sludge C. For example, the camera 10 is provided at the top of the flocculation tank 1C as shown in FIG. 1 so as to be able to capture an image of the interior of the flocculation tank 1C. The position at which the camera 10 is provided is not limited to the top of the flocculation tank 1C. The camera 10 may be provided inside the flocculation tank 1C or outside the flocculation tank 1C so long as it can capture an image of the interior of the flocculation tank 1C. Furthermore, the camera 10 may capture an image of the flocculated sludge C at a timing when the agitator blade 1e of the agitator 1D provided in the flocculation tank 1C is at a fixed position. In this case, for example, by capturing an image immediately after the agitator blade 1e of the agitator 1D provided in the flocculation tank 1C passes the capture position of the camera 10, it becomes easier to capture an image of the flocs at rest or moving slowly.

(5) PC20
The PC 20 is a device that controls the state of flocs in the flocculated sludge C formed in the flocculation mixing tank 1, and is an example of a floc state control device. The floc state control device may be, for example, a server device.
The PC 20 receives the flocculated sludge image captured by the camera 10 from the camera 10. The PC 20 according to the present embodiment determines and controls the state of flocs based on the flocculated sludge image received from the camera 10, as an example.
The PC 20 is communicatively connected to the PLC 30. The PC 20 transmits a control signal for controlling the state of the flocks to the PLC 30. The PC 20 according to the present embodiment transmits, as an example, a signal for controlling the rotation speed of the pump 1B or a signal for controlling the rotation speed of the motor 1c of the high-speed mixer 1A to the PLC 30.
By controlling the rotation speed of the pump 1B, the amount of the coagulant B supplied from the pump 1B to the high-speed mixer 1A can be controlled. This makes it possible to control the dosing rate of the coagulant B added to the sludge A.
By controlling the rotation speed of the motor 1c of the high-speed mixer 1A, the rotation of the rotating shaft 1a and the mixing blade 1b of the high-speed mixer 1A can be controlled, thereby controlling the degree of mixing of the sludge A and the flocculant B in the high-speed mixer 1A.

In another embodiment, the PC 20 may transmit a signal to the PLC 30 to control the rotation speed of the motor 1f of the agitator 1D. By controlling the rotation speed of the motor 1f of the agitator 1D, the rotation of the rotating shaft 1d and the agitator blade 1e of the agitator 1D can be controlled. This makes it possible to control the degree of mixing of the sludge A and the flocculant B in the agitator 1D.

(6) PLC30
The PLC 30 is a device that controls the operation of a control object. The PLC 30 controls the operation of the control object based on a control signal received from the PC 20. The control signal includes, for example, a control value indicating the control amount of each control object. The control objects are, for example, the pump 1B and the motor 1c of the high-speed mixer 1A. When the control object is the pump 1B, the PLC 30 controls the rotation speed of the pump 1B based on the control signal received from the PC 20. When the control object is the motor 1c of the high-speed mixer 1A, the PLC 30 controls the rotation speed of the motor 1c of the high-speed mixer 1A based on the control signal received from the PC 20.

<2. Functional configuration of PC>
The configuration of the sludge treatment facility has been described above with reference to Fig. 1. Next, the functional configuration of the PC 20 will be described with reference to Figs. 2 to 7. Fig. 2 is a diagram showing an example of the functional configuration of the PC according to an embodiment of the present invention.
As shown in FIG. 2, the PC 20 includes an input unit 210 , a communication unit 220 , a storage unit 230 , an output unit 240 , and a control unit 250 .

(1) Input unit 210
The input unit 210 has a function of accepting input from a user. The input unit 210 is realized by an input device such as a keyboard, a mouse, a touch panel, etc. The input device may be a device that the PC 20 is provided with in advance as hardware, or may be a device that is externally connected to the PC 20.

(2) Communication unit 220
The communication unit 220 has a function of transmitting and receiving various information. For example, the communication unit 220 receives a flocculated sludge image from the camera 10. The communication unit 220 also transmits a control signal to the PLC 30.

(3) Storage unit 230
The storage unit 230 has a function of storing various information. The storage unit 230 is configured by a storage medium, for example, a hard disk drive (HDD), a network attached storage (NAS), a solid state drive (SSD), a flash memory, an electrically erasable programmable read only memory (EEPROM), a random access read/write memory (RAM), a read only memory (ROM), or any combination of these storage media.

(4) Output unit 240
The output unit 240 has a function of outputting various information. The output unit 240 is realized by, for example, a display device such as a display. The display device may be a device that the PC 20 is provided with in advance as hardware, or may be a device that is externally connected to the PC 20.

(5) Control unit 250
The control unit 250 has a function of controlling the overall operation of the PC 20. The control unit 250 is realized, for example, by causing a central processing unit (CPU) provided as hardware in the PC 20 to execute a program.
As shown in FIG. 2 , the control unit 250 includes a flocculated sludge image acquisition unit 251 , an image selection unit 252 , an area determination unit 253 , a gap area acquisition unit 254 , a chemical injection rate control unit 255 , and a rotation speed control unit 256 .

(5-1) Flocculated sludge image acquisition unit 251
The flocculated sludge image acquisition unit 251 has a function of acquiring a flocculated sludge image. For example, the flocculated sludge image acquisition unit 251 acquires the flocculated sludge image captured by the camera 10, transmitted to the PC 20, and received by the communication unit 220 of the PC 20.

Here, the relationship between the state of flocs and the area of the gaps between flocs will be described with reference to Figures 3 to 5. Figures 3 to 5 are diagrams showing examples of flocculated sludge images.

3 shows a flocculated sludge image 40 in which the average area of each gap is 150. The gaps occurring in the flocculated sludge shown in the flocculated sludge image 40 are, for example, gaps 41.
4 shows a flocculated sludge image 50 in which the average area of each gap is 600. The gaps occurring in the flocculated sludge shown in the flocculated sludge image 50 are, for example, gaps 51.
5 shows a flocculated sludge image 60 in which the average area of each gap is 1000. The gaps occurring in the flocculated sludge shown in the flocculated sludge image 60 are, for example, gaps 61.

Comparing the flocculated sludge images 40, 50, and 60, it can be seen that increasing the amount of flocculant injected increases the average area of each gap, and decreasing the amount of flocculant injected decreases the average area of each gap. This makes it possible to judge the quality of the floc condition based on the gap area (hereinafter also referred to as "gap area") rather than the area of the flocs themselves.
When a floc overlaps with another floc, the area of the floc cannot be calculated accurately. If an inaccurate floc area is used, the state of the floc cannot be accurately understood. On the other hand, since the gap area is not affected by overlapping flocs, the area of each gap can be calculated with high accuracy. Therefore, by using the gap area instead of the floc area, the state of the floc can be understood and controlled with high accuracy.

(5-2) Image selection unit 252
The image selection unit 252 has a function of selecting an image more suitable for controlling the state of flocs from the flocculated sludge images acquired by the flocculated sludge image acquisition unit 251. For example, the image selection unit 252 performs image processing for detecting the velocity vectors of flocs using optical flow (OF) on the flocculated sludge images acquired by the flocculated sludge image acquisition unit 251. Then, the image selection unit 252 selects the flocculated sludge image according to the detection result of the velocity vectors of the flocs by the optical flow.

Specifically, the image selection unit 252 detects the velocity vectors of the flocs (i.e., the movement of the flocs) using optical flow from the flocculated sludge images (multiple still images or moving images) in the flocculated mixing tank 1 acquired by the flocculated sludge image acquisition unit 251, and acquires the flocculated sludge image when the vector value of the detected floc velocity vector becomes less than a predetermined threshold. The vector value is, for example, the sum of the x components or y components of the multiple detected velocity vectors. Alternatively, it may be the magnitude (absolute value) of the velocity vector, or may be an average value as well as the sum of the x components or y components of the velocity vector. The predetermined threshold is preferably, for example, a value at which it can be determined that the flocs in the flocculated mixing tank 1 are stationary, or a value at which it can be determined that the movement of the flocs is slow.

With this configuration, the image selection unit 252 can select and acquire an image of flocculated sludge when the flocs are stationary or when the movement of the flocs is slow. In other words, the image selection unit 252 can eliminate images of flocculated sludge that are not clear because the sludge is fluid due to the influence of stirring in the flocculation mixing tank 1.
This allows the PC 20 to use an agglomerated sludge image in which the flocs are clearer when controlling the state of the flocs, thereby improving the accuracy of controlling the state of the flocs compared to when an agglomerated sludge image in which the flocs are not clearer is used.

(5-3) Area determination unit 253
The area determination unit 253 has a function of determining gap areas and non-gap areas in the flocculated sludge image acquired by the image selection unit 252. For example, the area determination unit 253 uses a machine learning model that determines gap areas and non-gap areas in the flocculated sludge image, and the image selection unit 252 determines gap areas and non-gap areas in the acquired flocculated sludge image.

The machine learning model used by the area determination unit 253 is, for example, a model that uses semantic segmentation (SS) to perform machine learning to determine whether each pixel of the input flocculated sludge image is a gap area or a non-gap area. This model uses, for example, images in which a person has painted "gap areas" and "non-gap areas" in different colors as training data. This model uses a machine learning technique to learn a neural network model that can classify "gap areas" and "non-gap areas" from the original image in the same way as the training data, and automatically extracts features for classification to create it.

The area determination unit 253 inputs the flocculated sludge image acquired by the image selection unit 252 to a trained model that has undergone machine learning using semantic segmentation. The trained model outputs a result in which each pixel of the input flocculated sludge image is classified into either a "gap area" or a "non-gap area." The area determination unit 253 acquires the result output by the trained model as the determination result.

In addition, the area determination unit 253 may classify each pixel of the flocculated sludge image as either a "gap area" or a "non-gap area" by binarizing the flocculated sludge image without using the above-mentioned trained model.

(5-4) Gap area acquisition unit 254
The gap area acquisition unit 254 has a function of acquiring a gap area indicating the area of gaps occurring between a plurality of flocks based on the flocculated sludge image. For example, the gap area acquisition unit 254 calculates the area of an area determined to be a gap area in the flocculated sludge image based on the determination result acquired by the area determination unit 253 as the gap area. At this time, the gap area acquisition unit 254 calculates the total area of the gap areas or the average area of the gap areas as the gap area.
When multiple gaps are adjacent to each other, the gap area acquisition unit 254 regards the multiple gaps as one gap and calculates the gap area.

The gap area acquisition unit 254 may calculate the total area of the gap areas or the average area of the gap areas by excluding at least one area in order of size from among the areas of the gap areas, or at least one area in order of size from among the areas of the gap areas. When multiple gaps are continuous, the gap area acquisition unit 254 ranks the gaps by area size after regarding the multiple gaps as one gap. Then, the gap area acquisition unit 254 calculates the total area of the gap areas or the average area of the gap areas by excluding the top gap areas (those with large areas) and the bottom gap areas (those with small areas) from the ranked gap areas.
This enables the gap area acquisition unit 254 to reduce judgment errors caused by noise (e.g., debris or bubbles in the sludge) and fluctuations (e.g., the effects of stirring) when imaging the flocculated sludge.

(5-5) Chemical injection rate control unit 255
The dosing rate control unit 255 has a function of controlling the dosing rate of the flocculant. For example, the dosing rate control unit 255 controls the dosing rate of the flocculant so that the gap area acquired by the gap area acquisition unit 254 becomes the target area. For example, the dosing rate control unit 255 determines the rotation speed of the pump 1B for bringing the gap area closer to the target area based on the relationship between the dosing rate of the flocculant and the gap area. Then, the dosing rate control unit 255 transmits a control signal indicating the determined rotation speed of the pump 1B from the communication unit 220 to the PLC 30.

Here, the relationship between the flocculant dosing rate and the gap area will be described with reference to Fig. 6. Fig. 6 is a diagram showing an example of the relationship between the flocculant dosing rate and the gap area according to an embodiment of the present invention. The horizontal axis of the graph shown in Fig. 6 indicates the flocculant dosing rate (%), and the vertical axis indicates the gap area.
From the graph shown in Figure 6, it can be seen that the gap area increases as the dosing rate increases. However, once the dosing rate of the flocculant reaches a certain level, the gap area decreases when the dosing rate of the flocculant increases further. This is presumably because the flocs are dispersed by overdosing, resulting in a smaller floc diameter.

The drug injection rate control unit 255 controls the drug injection rate of the flocculant so that the gap area acquired by the gap area acquisition unit 254 becomes the target area based on the relationship between the drug injection rate of the flocculant and the gap area as shown in FIG. 6. For example, as shown in FIG. 6, the target area is set to 1300. When the gap area acquired by the gap area acquisition unit 254 is smaller than 1300, the drug injection rate control unit 255 can increase the drug injection rate of the flocculant to bring the gap area closer to the target area. On the other hand, when the gap area acquired by the gap area acquisition unit 254 is larger than 1300, the drug injection rate control unit 255 may increase the drug injection rate of the flocculant to bring the gap area closer to the target area, or may decrease the drug injection rate of the flocculant to bring the gap area closer to the target area.

Here, the setting of the target area will be described. For example, the target area is set in consideration of the operating efficiency of the equipment in the downstream stage of the flocculation mixing tank 1. As an example, the target area is set so as to maximize the operating efficiency.
The equipment downstream of the flocculation mixing tank 1 is, for example, a concentrator 2 as shown in Fig. 1. When the concentrator 2 is a general concentrator, the operating efficiency is the concentration efficiency of the concentrator 2. When the concentrator 2 is a heating concentrator, the operating efficiency is the heating efficiency of the concentrator 2. The heating efficiency is calculated, for example, by the following formula (1).
Heating efficiency (%) = ΔTact / ΔTide × 100 ... (1)
In the above formula (1),
ΔTact = thickened sludge temperature - raw sludge temperature ... (2)
ΔTide=(T1×Q1+T2×Q2)/(Q1+Q2)…(3)
In the above formula (3), T1 is the raw sludge temperature (flocculated sludge temperature), T2 is the hot water temperature, Q1 is the thickened sludge flow rate, and Q2 is the hot water flow rate.

Here, the relationship between the gap area and the heating efficiency will be described with reference to Fig. 7. Fig. 7 is a diagram showing an example of the relationship between the gap area and the heating efficiency according to an embodiment of the present invention. The horizontal axis of the graph shown in Fig. 7 indicates the gap area, and the vertical axis indicates the heating efficiency (%).
From the trends of each graph shown in Figure 7, it can be seen that the heating efficiency tends to increase as the gap area increases. However, once the gap area becomes large to a certain extent, it can be seen that the heating efficiency tends to decrease if the gap area becomes larger than this. This means that the heating efficiency can decrease if the flocs are too large.

The drug injection rate control unit 255 controls the drug injection rate of the coagulant based on the relationship between the gap area and the heating efficiency as shown in FIG. 7 so that the gap area acquired by the gap area acquisition unit 254 is the area at which the heating efficiency is maximized. For example, in the example of the graph shown in FIG. 7, the heating efficiency is maximized when the gap area is around 1250. In other words, the target area is set to around 1250. Then, the drug injection rate control unit 255 controls the drug injection rate of the coagulant in the same manner as in the example of FIG. 6 so that the gap area acquired by the gap area acquisition unit 254 approaches the target area of around 1250.

The target area may also be set according to the concentration of the sludge. For example, the drug injection rate control unit 255 sets a target area according to the concentration of the sludge, and controls the drug injection rate of the coagulant so that the gap area acquired by the gap area acquisition unit 254 approaches the target area. This eliminates the need to adjust the sludge concentration to control the state of the flocs. In other words, the state of the flocs can be controlled without adjusting the sludge concentration by diluting the sludge or adding a new device.

(5-6) Rotational speed control unit 256
The rotation speed control unit 256 controls the rotation speed of the motor 1c of the high-speed mixer 1A that mixes the sludge to be sent to the flocculation mixing tank 1 so that the gap area acquired by the gap area acquisition unit 254 becomes the target area. For example, the rotation speed control unit 256 determines the rotation speed of the motor 1c for bringing the gap area closer to the target area based on the relationship between the rotation speed of the motor 1c and the gap area. Then, the rotation speed control unit 256 transmits a control signal indicating the determined rotation speed of the motor 1c from the communication unit 220 to the PLC 30.

In another embodiment, the rotation speed control unit 256 may control the rotation speed of the motor 1f of the stirring means 1D that stirs the sludge in the flocculation mixing tank so that the gap area acquired by the gap area acquisition unit 254 becomes the target area. In this case, for example, the rotation speed control unit 256 determines the rotation speed of the motor 1f to bring the gap area closer to the target area based on the relationship between the rotation speed of the motor 1f and the gap area. Then, the rotation speed control unit 256 transmits a control signal indicating the determined rotation speed of the motor 1f from the communication unit 220 to the PLC 30.

In this embodiment, in order to bring the gap area closer to the target area, control by the drug injection rate control unit 255 is given priority over control by the rotation speed control unit 256. This is because controlling the drug injection rate is more time-efficient. If the gap area can be brought closer to the target area by control only by the drug injection rate control unit 255, the rotation speed control unit 256 does not need to perform control. On the other hand, if the gap area cannot be brought closer to the target area by control only by the drug injection rate control unit 255, the rotation speed control unit 256 performs control.

<3. Processing flow>
The functional configuration of the PC 20 has been described above with reference to Fig. 2 to Fig. 7. Next, the flow of processing in the PC 20 will be described with reference to Fig. 8. Fig. 8 is a flowchart showing an example of the flow of processing in the PC 20 according to an embodiment of the present invention. Note that Fig. 8 shows, as an example, an example in which the rotation speed control unit 256 controls the rotation speed of the motor 1c of the high-speed mixer 1A.

As shown in FIG. 8, first, the flocculated sludge image acquisition unit 251 of the PC 20 acquires the flocculated sludge image captured by the camera 10, transmitted to the PC 20, and received by the communication unit 220 of the PC 20 (step S101).

Next, the image selection unit 252 of the PC 20 performs image processing on the flocculated sludge image acquired by the flocculated sludge image acquisition unit 251 to detect the velocity vector of the flocs using optical flow (step S102).
After detecting the velocity vectors of the flocs, the image selection unit 252 selects an image of flocculated sludge according to the detection result of the velocity vectors of the flocs (step S103).

Next, the area determination unit 253 of the PC 20 uses a machine learning model to determine gap areas and non-gap areas in the flocculated sludge image acquired by the image selection unit 252 (step S104).

Next, the gap area acquisition unit 254 of the PC 20 calculates the area of the area determined to be a gap area in the flocculated sludge image based on the determination result acquired by the area determination unit 253 as the gap area (step S105).

Next, the injection rate control unit 255 of the PC 20 controls the injection rate of the flocculant so that the gap area acquired by the gap area acquisition unit 254 becomes the target area (step S106).
If the gap area approaches the target area and becomes an appropriate value after controlling the drug injection rate (step S107/YES), the process ends. On the other hand, if the gap area after controlling the drug injection rate is not an appropriate value (step S107/NO), the process proceeds to step S108.

When the process proceeds to step S108, the rotation speed control unit 256 controls, for example, the rotation speed of the motor 1c of the high-speed mixer 1A that mixes the sludge to be sent to the flocculation mixing tank 1 so that the gap area acquired by the gap area acquisition unit 254 becomes the target area (step S108), and then ends the process.

4. Specific Examples
The flow of processing in the PC 20 has been described above with reference to Fig. 8. Next, a specific example will be described with reference to Fig. 9. Fig. 9 is a diagram showing a specific example according to an embodiment of the present invention.

9 shows the time series changes in the gap area and the drug injection rate when the drug injection rate is controlled so that the gap area becomes the target area, and the actual measured values of the time series changes in the heating efficiency corresponding to the time series changes are shown. The target area is set to 1250±100.
As shown in FIG. 9, when the drug injection rate was controlled so that the gap area was 1250±100, as shown in the results from 14:50 to 15:50, the drug injection rate was in the range of 0.40% to 0.41%, and the heating efficiency was maintained at a high level of 113% on average.
In addition, an experiment was also conducted in which it was assumed that the drug injection rate was not controlled and that the drug injection rate had fallen to 0.37% due to concentration fluctuations. As shown in the results from 15:55 to 16:20, the gap area fell to about 800 and the heating efficiency fell below 100%.
From the above, it can be said that the heating efficiency can be kept high by controlling the drug injection rate to keep the gap area at an appropriate level. In addition, by keeping the gap area at an appropriate level, it is possible to optimize the drug injection amount, allowing the concentrator 2 to be operated safely and reducing the drug injection amount (reducing costs).

As described above, the PC 20 (flock state control device) according to this embodiment includes the flocculated sludge image acquisition unit 251 , the gap area acquisition unit 254 , and the chemical injection rate control unit 255 .
The flocculated sludge image acquisition unit 251 acquires a flocculated sludge image obtained by imaging flocculated sludge containing flocs formed by mixing a flocculant and sludge in the flocculation mixing tank 1 .
The gap area acquisition unit 254 acquires a gap area indicating the area of the gaps occurring between a plurality of flocs based on the acquired flocculated sludge image.
The drug injection rate control unit 255 controls the drug injection rate of the flocculant so that the acquired gap area becomes the target area.

With this configuration, the PC 20 according to this embodiment can more accurately control the state of the flocs using the gap area that is not affected by overlapping flocs, etc. Because it is not affected by overlapping flocs, etc., it is possible to control the state of the flocs without adjusting the concentration of the sludge by diluting the sludge or adding a new device.

The PC 20 of this embodiment therefore makes it possible to appropriately control the state of flocs in a location where it is easier to photograph than in the downstream piping and where adjustments of the flocculant can be made earlier, without the need to dilute the sludge concentration or add new equipment to reduce the overlap of flocs.

The above describes an embodiment of the present invention. Note that a part or all of the PC 20 in the above embodiment may be realized by a computer. In that case, a program for realizing this function may be recorded on a computer-readable recording medium, and the program recorded on the recording medium may be read into a computer system and executed to realize the function. Note that the term "computer system" here includes hardware such as an OS and peripheral devices. Furthermore, the term "computer-readable recording medium" refers to portable media such as flexible disks, optical magnetic disks, ROMs, and CD-ROMs, and storage devices such as hard disks built into a computer system. Furthermore, the term "computer-readable recording medium" may include a medium that dynamically holds a program for a short period of time, such as a communication line when a program is transmitted via a network such as the Internet or a communication line such as a telephone line, and a medium that holds a program for a certain period of time, such as a volatile memory inside a computer system that is a server or client in such a case. Furthermore, the above program may be a program for realizing a part of the above-mentioned function, or may be a program that can realize the above-mentioned function in combination with a program already recorded in the computer system, or may be a program that is realized using a programmable logic device such as an FPGA (Field Programmable Gate Array).

The above describes in detail an embodiment of the present invention with reference to the drawings, but the specific configuration is not limited to the above, and various design changes can be made without departing from the spirit of the present invention.

1...flocculation mixing tank, 1a...rotating shaft, 1A...high-speed mixer, 1b...mixing blade, 1B...pump, 1c...motor, 1C...flocculation tank, 1d...rotating shaft, 1D...mixing means, 1e...mixing blade, 1f...motor, 2...concentrator, 2a...concentration filtration screen, 2A...concentration tank, 2b...filtrate chamber, 2B...conveying means, 2c...rotating shaft, 2C...pump, 2d...screw, 2e...motor, 3...dehydrator, 3a...inner filtration screen, 3A...casing, 3A1...first space, 3A2...second space, 3b...outer filtration screen, 3B...filtration screen, 3c...motor, 3C...base, 3d...ribbon screw, 3D...Cover, 3e...Connecting plate, 3E...Discharge chamber, 3f...Supply pipe, 3F...Discharge port, 3G...Compression ring, 3H...Drain pipe, 3P...Pump, 4...Concentrated sludge supply path, 4A...Pump, 4B...High-speed mixer, 4C...Pump, 10...Camera, 40...Flocculated sludge image, 41...Gap, 50...Flocculated sludge image, 51...Gap, 60...Flocculated sludge image, 61...Gap, 210...Input unit, 220...Communication unit, 230...Storage unit, 240...Output unit, 250...Control unit, 251...Flocculated sludge image acquisition unit, 252...Image selection unit, 253...Area determination unit, 254...Gap area acquisition unit, 255...Drug injection rate control unit, 256...Rotation speed control unit

Claims (10)

a flocculated sludge image acquisition unit that acquires an image of flocculated sludge including flocs formed by flocculating sludge with a flocculant in the flocculation mixing tank;
an area determination unit that determines the gap areas and the areas other than the gap areas in the acquired flocculated sludge image by using a machine learning model that determines the gap areas and the areas other than the gap areas occurring between the plurality of flocs in the flocculated sludge image;
a gap area acquisition unit that calculates an area of the region determined to be the gap region as a gap area indicating the area of the gap;
A dosing rate control unit that controls the dosing rate of the flocculant so that the acquired gap area becomes a target area;
Equipped with
The flock state control device, wherein the machine learning model is a model that has been machine-learned to determine, using semantic segmentation, whether each pixel of an input flocculated sludge image is a gap region or a non-gap region . an image selection unit that detects a velocity vector of the flocs based on the flocculated sludge image by using an optical flow, and acquires an image of the flocculated sludge when a vector value of the detected velocity vector is less than a predetermined threshold value;
The flock state control device of claim 1 further comprising: The gap area acquisition unit calculates a total area of the gap region or an average area of the gap region as the gap area.
The flock state control device according to any one of claims 1 to 2 . The gap area acquisition unit calculates a total area of the gap regions or an average area of the gap regions by excluding at least one area in order of size from among the areas of the gap regions or at least one area in order of size from among the areas of the gap regions.
The flock state control device according to claim 3 . The target area is set so as to maximize the operating efficiency of the equipment downstream of the flocculation mixing tank.
The flock state control device according to any one of claims 1 to 4 . The target area is set according to the concentration of the sludge.
The flock state control device according to any one of claims 1 to 5 . a rotation speed control unit that controls at least one of the rotation speed of a mixer that stirs the sludge to be sent to the flocculation mixing tank and the rotation speed of an agitator that stirs the flocculated sludge in the flocculation mixing tank so that the acquired gap area becomes the target area;
The flock state control device according to any one of claims 1 to 6 , further comprising: A sludge treatment facility comprising the flock state control device according to any one of claims 1 to 7 . a flocculated sludge image acquisition step in which a flocculated sludge image acquisition unit acquires an image of flocculated sludge including flocs formed by flocculating sludge with a flocculant in a flocculation mixing tank;
an area determination step in which an area determination unit determines the gap areas and other areas in the acquired flocculated sludge image by using a machine learning model that determines gap areas occurring between the plurality of flocs in the flocculated sludge image and other areas;
a gap area acquisition step in which a gap area acquisition unit calculates an area of the region determined to be the gap region as a gap area indicating the area of the gap;
A dosing rate control process in which a dosing rate control unit controls the dosing rate of the flocculant so that the acquired gap area becomes a target area;
Including,
The flock state control method, wherein the machine learning model is a model that has been machine-learned to use semantic segmentation to determine whether each pixel of an input flocculated sludge image is a gap region or a non-gap region . Computer,
a flocculated sludge image acquisition means for acquiring an image of flocculated sludge including flocs formed by flocculation of sludge with a flocculant in the flocculation mixing tank;
an area determination means for determining the gap areas and the non-gap areas in the acquired flocculated sludge image by using a machine learning model for determining the gap areas and the non-gap areas occurring between the plurality of flocs in the flocculated sludge image;
a gap area acquisition means for calculating an area of the region determined to be the gap region as a gap area indicating the area of the gap;
A chemical injection rate control means for controlling the chemical injection rate of the flocculant so that the acquired gap area becomes a target area;
Functioning as a
The machine learning model is a program that uses semantic segmentation to perform machine learning to determine whether each pixel of an input flocculated sludge image is a gap area or a non-gap area .

Images