Injection rate determination device and injection rate determination method

Abstract

This invention provides a technology for more appropriately determining the injection rate of coagulants into sludge in sludge treatment facilities. [Solution] Applicable to a sludge treatment facility that compresses flocculated sludge containing flocs generated by injecting a flocculant into sludge in a flocculation tank into compressed sludge using a screw press, the device determines the injection rate of a flocculant into the sludge, and comprises a moisture content acquisition unit 142 that inputs image data of compressed sludge discharged from a screw press, captured by a compressed sludge imaging device, into a first trained model learned using image data of compressed sludge, and acquires the moisture content of the compressed sludge based on the estimation results of the first trained model, and an injection rate determination unit 143 that determines the flocculant injection rate based on the acquisition results of the moisture content acquisition unit 142.

Description

Technical Field

[0001] The present disclosure is applied to sludge treatment facilities that perform pressing (concentration or dewatering) of sludge, and relates to an injection rate determination device and an injection rate determination method for determining the injection rate of a flocculant into sludge.

Background Art

[0002] Conventionally, as one type of sludge treatment facility for pressing sludge, there is known a facility that injects a flocculant into sludge to generate flocculated sludge containing flocs, and presses the generated flocculated sludge with a screw press. In such sludge treatment facilities, it is desirable to appropriately adjust the injection rate of the flocculant into the sludge in order to improve the sludge treatment performance. Accordingly, the injection rate of the flocculant into the sludge is adjusted based on the state of flocs identified from floc image data (see Patent Document 1).

[0003] Patent Document 1 describes that using the image data of flocculated flocs and information data related to water treatment as explanatory variables, a numerical value as a determination result indicating excess or deficiency of the flocculant injection rate, which is an objective variable indicating the state of flocculated flocs, is output in real time from a trained model, and the injection rate of the flocculant is adjusted based on the determination result.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] As described in Patent Document 1, the injection rate of the coagulant into the sludge can be adjusted appropriately based on the state of the flocs to bring it closer to a desirable state. However, even when this technique is employed, it is difficult to always determine the appropriate injection rate of the coagulant into the sludge. For example, even if the state of the flocs (which can also be called the state of the coagulated sludge) determined from image data is the same, the compressed sludge obtained by compressing the coagulated sludge with a screw press may not be in the same state. In other words, if the injection rate of the coagulant into the sludge is determined based on the state of the flocs, the injection rate of the coagulant into the sludge may not necessarily be appropriate.

[0006] This disclosure is made in view of these circumstances and aims to provide a technology for more appropriately determining the injection rate of a coagulant into sludge in a sludge treatment facility. [Means for solving the problem]

[0007] The injection rate determination device of the present disclosure is applied to a sludge treatment facility that compresses flocculated sludge containing flocs generated by injecting a flocculant into sludge in a flocculation tank into compressed sludge using a screw press, and determines the injection rate of the flocculant into the sludge. The device comprises: a moisture content acquisition unit that inputs image data of the compressed sludge discharged from the screw press, which is captured at predetermined intervals by a compressed sludge imaging device, into a first trained model trained using image data of the compressed sludge, and acquires the moisture content of the compressed sludge based on the estimation results of the first trained model; and an injection rate determination unit that determines the injection rate of the flocculant into the sludge based on the acquisition results of the moisture content acquisition unit. [Effects of the Invention]

[0008] According to the technology disclosed herein, the injection rate of a coagulant into sludge can be determined more appropriately in a sludge treatment facility. Consequently, the sludge treatment performance in the sludge treatment facility can be improved. [Brief explanation of the drawing]

[0009] [Figure 1] This figure shows an example of the overall configuration of a sludge treatment facility including an injection rate determination device according to one embodiment. [Figure 2] This block diagram shows an example of the configuration of an injection rate determination device according to one embodiment. [Figure 3] This figure shows an example of a flocculant injection rate determination table. [Figure 4] This figure shows an example of a flocculant injection rate determination table presented to the user. [Figure 5] This figure shows an example of changes in parameters indicating the flocculation state during the trial operation of a sludge treatment facility. [Figure 6] This diagram illustrates the first and second threshold values ​​set for the moving average of the estimated values. [Figure 7] This figure shows an example of the change in parameters indicating the moisture content during the trial operation of a sludge treatment facility. [Figure 8] This diagram illustrates the third threshold value set for the moving average of the estimated values. [Figure 9] This figure shows another example of a flocculant injection rate determination table presented to the user. [Modes for carrying out the invention]

[0010] One embodiment of this disclosure will be described in detail with reference to the drawings. In all drawings illustrating the embodiment, the same or substantially identical components and elements will be denoted by the same reference numerals. Components and elements that have already been described will generally not be described again.

[0011] In explanations, program-based processing may sometimes be described primarily from the perspective of the program, its functions, or its processing units. However, the main hardware component is the processor, or a controller, device, computer, or system composed of such a processor. The computer, through its processor, executes processing according to the program read into memory, using resources such as memory and communication interfaces as appropriate. This realizes the specified functions and processing units. The processor is composed of semiconductor devices such as a CPU / MPU or GPU. Processing is not limited to software program processing; it can also be implemented using dedicated circuits. FPGAs, ASICs, CPLDs, etc., can be used as dedicated circuits.

[0012] The program may be pre-installed as data on the target computer, or it may be distributed as data to the target computer from the program source. The program source may be a program distribution server on a communication network, or a non-transient computer-readable storage medium, such as a memory card or disk. The program may consist of multiple modules. The computer system may consist of multiple devices. The computer system may consist of a client-server system, a cloud computing system, an IoT system, etc. Various types of data and information are composed of structures such as tables and lists, but are not limited to these. Representations such as identification information, identifiers, IDs, names, and numbers are interchangeable.

[0013] The following describes an injection rate determination device and an injection rate determination method according to one embodiment of this disclosure. The injection rate determination device of this disclosure is applied to a sludge treatment facility in which a flocculant is injected into sludge in a flocculation tank to produce flocculated sludge containing flocs, which is then compressed by a screw press to produce compressed sludge. The device is for determining the injection rate of a flocculant to the sludge. More specifically, the injection rate determination device is for determining the injection rate of a flocculant to the sludge in real time in a sludge treatment facility so that the compressed sludge is in an appropriate state.

[0014] The pressing treatment in a sludge treatment facility refers to subjecting flocculated sludge to concentration treatment or dewatering treatment using a screw press machine. Both the dewatering treatment and the concentration treatment are included in the pressing treatment, but the moisture content of the sludge after the treatment differs. Specifically, when the moisture content of the sludge after treatment is set to X% or higher, the treatment is referred to as a concentration treatment, and the sludge after the concentration treatment is referred to as concentrated sludge. When the moisture content of the sludge after treatment is set to less than X%, the treatment is referred to as a dewatering treatment, and the sludge after the dewatering treatment is referred to as dewatered sludge. A sludge treatment facility in which such concentration treatment is performed is also called a sludge concentration treatment facility, and a sludge treatment facility in which dewatering treatment is performed is also called a sludge dewatering treatment facility. In addition, the screw press machine provided in the sludge concentration treatment facility is also called a screw press concentrator, and the screw press machine provided in the sludge dewatering treatment facility is also called a screw press dewaterer.

[0015] Note that the injection rate determination device may control various apparatuses and devices provided in the sludge treatment facility so as to achieve the determined injection rate of the flocculant. In this case, the injection rate determination device can also be referred to as an injection rate control device. In addition, the injection rate determination device may present information on the determined injection rate of the flocculant to a user. An operator, who is the user, adjusts various apparatuses and devices provided in the sludge treatment facility based on the presented information. Therefore, the injection rate determination device in this case can also be referred to as an injection rate adjustment support device.

[0016] [Overall Configuration of Sludge Treatment Facility] First, an example of the overall configuration of a sludge treatment facility to which the injection rate determination device and the injection rate determination method according to the embodiment of the present disclosure are applied will be described. Figure 1 is a diagram showing an example of the overall configuration of the sludge treatment facility. The sludge treatment facility is a facility for treating sludge (sludge wastewater), and can also be referred to as a water treatment facility. Further, in the present embodiment, a case where the sludge treatment facility is a sludge dewatering treatment facility that performs dewatering treatment of flocculated sludge will be described. Therefore, in the following description, the screw press machine provided in the sludge dewatering treatment facility is referred to as a screw press dewaterer, and the sludge discharged from the screw press dewaterer is referred to as dewatered sludge.

[0017] As shown in Figure 1, the sludge dewatering treatment facility 10 to which the injection rate determination device 100 is applied is configured to include, as an example, a sludge storage tank 20, a sludge relay tank 30, a flocculation tank 50, a screw press dewatering machine (hereinafter also simply referred to as "dewatering machine") 60, and a control panel 80.

[0018] The sludge storage tank 20 stores, for example, sludge (sludge wastewater) discharged from manufacturing factories for food, chemical products, semiconductors and the like. This sludge storage tank 20 is provided with a first level gauge 21 and a first pump 22. The first level gauge 21 is provided to measure the amount of sludge stored in the sludge storage tank 20. The first level gauge, as an example, measures the level of the sludge stored in the sludge storage tank 20, in other words, the water level (height of the water surface) of the sludge wastewater. The measurement result by the first level gauge 21 is sent to, for example, the control panel 80. Connected to the first pump 22 is a first pipe 23 that is a water supply line connecting the sludge storage tank 20 and the sludge relay tank 30. The sludge stored in the sludge storage tank 20 is sent out to the sludge relay tank 30 via the first pipe 23 by the first pump 22.

[0019] The sludge relay tank 30 is a storage tank smaller than the sludge storage tank 20, which is provided between the sludge storage tank 20 and the flocculation tank 50. This sludge relay tank 30 is a so-called service tank, and temporarily stores the sludge sent from the sludge storage tank 20 toward the flocculation tank 50. The first pipe 23 connecting the sludge storage tank 20 and the sludge relay tank 30 is provided with a flow meter 24. The flow meter 24 measures the flow rate of the sludge flowing from the sludge storage tank 20 into the sludge relay tank 30 via the first pipe 23.

[0020] The sludge relay tank 30 is provided with a second level gauge 31 and a second pump 32. The second level gauge 31 measures the level of the sludge stored in the sludge relay tank 30. The measurement result by the second level gauge 31 is sent to, for example, the control panel 80. Connected to the second pump 32 is a second pipe 33 that is a water supply line connecting the sludge relay tank 30 and the flocculation tank 50. The sludge stored in the sludge relay tank 30 is sent out to the flocculation tank 50 via the second pipe 33 by the second pump 32.

[0021] The coagulation tank 50 is equipped with an internal stirring device 51 for stirring the sludge flowing in from the sludge transfer tank 30, and is also called a coagulation mixing tank or stirring tank. The stirring device 51 comprises a stirring blade 52 located inside the coagulation tank 50 and a stirring motor 53 connected to the stirring blade 52. The coagulation tank 50 is also connected to a supply pipe 71 that leads to a coagulant tank 70, and a third pump 72 is provided in this supply pipe 71. The coagulant stored in the coagulant tank 70 is supplied to the coagulation tank 50 via the supply pipe 71 by the third pump 72. In other words, the coagulant stored in the coagulant tank 70 is injected into the sludge in the coagulation tank 50 via the supply pipe 71.

[0022] In the coagulation tank 50, the sludge and coagulant are mixed by the agitator 51. As a result, the sludge in the coagulation tank 50 becomes a state in which so-called flocs are formed. The sludge in the state in which flocs have been formed (hereinafter also referred to as coagulated sludge) is then supplied from the coagulation tank 50 to the dewatering machine 60 via the fourth pipe 54. As described later, the coagulated sludge is dewatered by the dewatering machine 60 to form dewatered sludge, but the water content of this dewatered sludge changes depending on the injection rate of coagulant to the sludge. For this reason, it is desirable that the injection rate of coagulant to the sludge be adjusted as appropriate depending on the situation. In the following explanation, "injection rate of coagulant to sludge" may be written as "coagulant injection rate".

[0023] In this example, the coagulation tank 50 is equipped with multiple stages of tanks, and the operation of the agitator 51 is appropriately adjusted in each tank. More specifically, the rotation speed of the agitator blades 52 is adjusted to be slower in the later stages of the coagulation tank 50. The coagulation tank 50 is also equipped with a third level gauge 55. The third level gauge 55 measures the level of sludge stored in the coagulation tank 50. The measurement results from the third level gauge 55 are sent to, for example, the control panel 80.

[0024] Furthermore, a coagulated sludge imaging device 56 is provided above the coagulation tank 50 for imaging the coagulated sludge. The coagulated sludge imaging device 56 is composed of, for example, a CCD camera installed outside the coagulation tank 50, and images the coagulated sludge inside the coagulation tank 50 from outside the tank 50. The images captured by the coagulated sludge imaging device 56 may be still images or moving images, but in this example, the coagulated sludge imaging device 56 takes images of the coagulated sludge at predetermined intervals, for example, every minute. The coagulated sludge imaging device 56 also functions as an injection rate determination device 100, as will be described later.

[0025] The screw press dewatering machine 60 dewaters the flocculated sludge flowing in from the sludge storage tank 20 through the flocculation tank 50 while conveying it with a screw. The dewatering machine 60 is an existing model and will be briefly described, but it is equipped with a screw 62 installed inside a cylindrical screen 61. A drive motor 64, for example, which is driven by electricity, is connected to one axial end of the screw 62, and the screw 62 rotates at a predetermined rotational speed using the drive motor 64 as a power source. In addition, a pressing plate 65 is provided near the other axial end of the screw 62, that is, the end of the screen 61 opposite to the drive motor 64, on which the sludge conveyed by the screw 62 is pressed. In other words, the pressing plate 65 is provided at the outlet portion of the dewatering machine 60 where the sludge is discharged.

[0026] The flocculated sludge that flows from the flocculation tank 50 to the dewatering machine 60 via the fourth pipe 54 is transported through the screen 61 by a screw 62 that rotates at a predetermined speed by a drive motor 64, and at this time the filtrate is separated through the screen 61. The flocculated sludge from which the filtrate has been separated is further compacted by the outlet resistance of the pressing plate 65 and becomes dewatered.

[0027] The coagulated sludge that flows into the dewatering machine 60 is transported through the screen 61 by the screw 62, and as it is transported, its moisture content gradually decreases, resulting in a dewatered state. The moisture content of the sludge dewatered in the dewatering machine 60 (dewatered sludge) varies depending on, for example, the rotation speed of the screw 62. Specifically, the lower the rotation speed of the screw 62, the lower the moisture content of the dewatered sludge tends to be.

[0028] In this way, the dewatered sludge, which has been transported and dewatered through the screen 61 by the screw 62, is pushed out from the outer circumference of the pressing plate 65 to the outside of the screen 61, and gradually falls off from the mass of dewatered sludge. In other words, the dewatered sludge that has been pushed out to the outside of the screen 61 eventually falls off the dewatering machine 60. The fallen dewatered sludge is stored in a dewatered sludge container 66 installed outside the dewatering machine 60.

[0029] Although not shown in the diagram, the pressing plate 65 is configured to be movable within a predetermined range in the axial direction of the screw 62. The size of the outlet from which the dewatered sludge is discharged is adjusted by the position of this pressing plate 65, and the outlet resistance changes accordingly. The larger the outlet resistance, the lower the water content of the dewatered sludge tends to be.

[0030] A compressed sludge imaging device for imaging compressed sludge, specifically a dewatered sludge imaging device 67 for imaging dewatered sludge (dewatered agglomerated sludge discharged from the dewatering machine 60), is provided on the outer periphery of the pressing plate 65 of the dewatering machine 60, that is, at the outlet of the dewatering machine 60. The dewatered sludge imaging device 67 is positioned to image the dewatered sludge discharged from the outer periphery of the pressing plate 65. More specifically, the dewatered sludge imaging device 67 is positioned to image the dewatered sludge before it falls out of the dewatering machine 60. The dewatered sludge imaging device 67 can be composed of a CCD camera or the like, similar to the agglomerated sludge imaging device 56, and the images captured by the dewatered sludge imaging device 67 may be still images or moving images, but the agglomerated sludge is imaged at predetermined intervals (every minute in this example). The dewatered sludge imaging device 67 also functions as an injection rate determination device 100 together with the agglomerated sludge imaging device 56.

[0031] The control panel 80 is comprised of, for example, a PLC 81, an inverter 82, a PC 83, and the like. The PLC 81 is a programmable logic controller connected to various devices in the sludge dewatering treatment facility 10 and controls the operation of these devices; it can also be called an operation control device. The PLC 81 appropriately controls the operation of, for example, the first pump 22 of the sludge storage tank 20, the second pump 32 of the sludge transfer tank 30, the stirring motor 53 of the stirring device 51, and the drive motor 64 of the dewatering machine 60 via the inverter 82. Furthermore, the PLC 81 also controls the operation of the third pump 72 for supplying coagulant to the coagulation tank 50.

[0032] PC83 is a personal computer that inputs and outputs control data for various devices and equipment in the sludge dewatering treatment facility 10, and is connected to the injection rate determination device 100 in a communicative manner. For example, PC83 acquires sensor recording data from various sensors installed in the sludge dewatering treatment facility 10 and transmits it to PLC81. However, the sensor recording data from the various sensors may be acquired directly by PLC81 without going through PC83. Also, as an example, PC83 collects image data captured by the coagulated sludge imaging device 56 and the dewatered sludge imaging device 67, and transmits the collected image data to the injection rate determination device 100.

[0033] The injection rate determination device 100 determines the coagulant injection rate based on image data transmitted from the PC 83, as will be described in more detail later. The injection rate determination device 100 is located, for example, in a central monitoring room located away from the control panel 80 of the sludge dewatering treatment equipment 10. However, the injection rate determination device 100 does not necessarily have to be located away from the control panel 80 of the sludge dewatering treatment equipment 10, and may be located near the sludge dewatering treatment equipment 10. Furthermore, the PC 83 provided in the control panel 80 may also function as the injection rate determination device 100.

[0034] Furthermore, PC83 may acquire information such as the coagulant injection rate determined by the injection rate determination device 100 and present the acquired information to the user, the operator, via a monitor or the like. In this case, for example, by operating PC83, the operation control of the third pump 72 is changed via PLC81, and the amount of coagulant injected into the coagulation tank 50 is adjusted as appropriate. In other words, the injection rate of coagulant to the sludge is adjusted as appropriate.

[0035] Furthermore, PC83 may acquire information such as the coagulant injection rate determined by the injection rate determination device 100 and transmit the acquired information to PLC81. In this case, the information transmitted to PLC81 includes, for example, information instructing a change in the operating state of the third pump 72. Therefore, when information is transmitted from PC83 to PLC81, the operating state of the third pump 72 is appropriately changed by PLC81, and the amount of coagulant injected into the coagulation tank 50 is automatically adjusted.

[0036] Furthermore, even when the amount of coagulant injected is automatically adjusted in this manner, it may be possible for an operator to fine-tune the coagulant injection rate to the sludge by operating the PC83. In that case, it is preferable that the PC83 monitor be configured as, for example, a touch panel. This would allow the user to fine-tune the coagulant injection rate on the PC83 monitor, improving work efficiency.

[0037] [Injection rate determination device] An example of the injection rate determination device 100 of this disclosure will be described in more detail below. Figure 2 is a diagram showing an example of the configuration of the injection rate determination device. The injection rate determination device 100 is applied to the sludge dewatering treatment facility 10 as described above and determines the injection rate of the coagulant to the sludge. As shown in Figure 2, the injection rate determination device 100 comprises a calculation device 110, an input device 120, and an output device 130. The calculation device 110 is composed of a computer system (computer) and comprises a processor 140, memory 150, communication interface device 160, input / output interface device 170, etc. The above components of the calculation device 110 are connected to each other by buses, etc., and are able to input / output information and communicate with each other.

[0038] The processor 140 has at least one execution unit that executes processes according to the control program 151 stored in the memory 150. This enables various functions. While the arithmetic unit 110 is running, execution modules corresponding to each function are generated and operate automatically. The processor 140, as will be described in more detail later, includes, as an example of an execution unit that performs various processes, a coagulated sludge state acquisition unit 141, a water content acquisition unit 142, an injection rate determination unit 143, a change instruction unit 144, and a presentation instruction unit 145.

[0039] Memory 150 is configured using internal or external memory resources of the arithmetic unit 110. Memory 150 is configured using, for example, a non-volatile storage device. Memory 150 may also be configured using an external storage device or DB server independent of the arithmetic unit 110. Memory 150 stores, for example, a control program 151, a model 152, image data 153, and also various setting information 154.

[0040] The control program 151 is a program that causes the processor 140 to execute each process, and can also be said to be a program that executes the injection rate determination method of this embodiment. The model 152 is a model used in inference and includes data and information such as setting parameters. In this example, the model 152 includes the first trained model, which is the water content trained model, and the second trained model, which is the coagulated sludge trained model, which will be described later. The image data 153 is image data of coagulated sludge including flocs transmitted from the coagulated sludge imaging device 56, and image data of dewatered sludge transmitted from the dewatered sludge imaging device 67. The setting information 154 can be, for example, user setting information.

[0041] The communication interface device 160 is a device that implements a communication interface with the control panel 80, including the PLC 81. Input / output interface device 170 has externally connected input devices 120 and output devices 130. Examples of input devices 120 include keyboards, mice, touch panels, and microphones. Examples of output devices 130 include displays, touch panels, printers, and speakers. The input devices 120 and output devices 130 may be built into the arithmetic unit 110.

[0042] Returning to the description of the execution unit of the processor 140, the coagulated sludge state acquisition unit 141 acquires the state of the coagulated sludge, including flocs, using a coagulated sludge trained model (second trained model) constructed by machine learning such as deep learning. As described later, the coagulated sludge state acquisition unit 141 acquires the size of the gaps formed in the flocs (hereinafter referred to as the floc gap size) as an example of the state of the coagulated sludge. In addition, the moisture content acquisition unit 142 acquires the moisture content of the dewatered sludge discharged from the screw press dewatering machine 60 using a moisture content trained model (first trained model) constructed by machine learning such as deep learning.

[0043] The injection rate determination unit 143 determines the injection rate of the coagulant to the sludge based on the results obtained by the coagulated sludge state acquisition unit 141 and the moisture content acquisition unit 142. More specifically, the injection rate determination unit 143 determines the injection rate of the coagulant to the sludge based on the state of the coagulated sludge (in this example, the range of floc gap size) obtained based on image data of the coagulated sludge in the coagulation tank 50, and the range of moisture content of the dewatered sludge obtained based on image data of the dewatered sludge discharged from the dewatering machine 60.

[0044] The coagulant injection rate is predetermined using a coagulant injection rate determination table based on the state of the flocculated sludge (also known as the flocculated state) and the moisture content of the dewatered sludge. Figure 3 illustrates an example of a coagulant injection rate determination table set as part of the operation plan. The coagulant injection rate is predetermined in the coagulant injection rate determination table according to the classification of multiple sludge states, which are identified by the state of the flocculated sludge (flocculated state) and the moisture content of the dewatered sludge. In the coagulant injection rate determination table shown in Figure 3, the coagulant injection rate is predetermined for six sludge state classifications C1 to C6, which are identified by the floc pore size (an example of the flocculated state) and the moisture content of the dewatered sludge.

[0045] The injection rate determination unit 143 determines the injection rate of the coagulant to the sludge based on the results obtained by the coagulated sludge state acquisition unit 141 and the moisture content acquisition unit 142, which will be described in detail later, and the coagulant injection rate determination table (map) set in this manner. The coagulant injection rate determination table is created, for example, by having a skilled worker at the sludge dewatering treatment facility 10 input the appropriate coagulant injection rate for each classification C1 to C6 using the input device 120, and is stored in memory 150 or the like.

[0046] The coagulated sludge state acquisition unit 141 acquires the state of coagulated sludge, including flocs, using a trained model of coagulated sludge constructed by machine learning such as deep learning. The trained model of coagulated sludge is constructed using a learning algorithm, for example, a convolutional neural network (CNN), based on multiple image data of the coagulated sludge. The coagulated sludge state acquisition unit 141 inputs image data of the coagulated sludge (which can also be called image data of flocs) captured by the coagulated sludge imaging device 56 into the constructed trained model of coagulated sludge, and acquires the state of the coagulated sludge based on the estimation result, which is the output of the trained model of coagulated sludge.

[0047] Here, the state of the flocculated sludge can be described as follows: for example, the size of the flocs contained in the flocculated sludge (floc size), the size of the gaps between the flocs (floc gap size), the density of the flocs, the viscosity of the flocculated sludge, the transparency of the water present in the floc gaps (floc gap water), and the color of the flocs. In this example, the flocculated sludge state acquisition unit 141 acquires the floc gap size as an example of the state of the flocculated sludge. The floc gap size is calculated, for example, as the average value of the diameter of each floc gap. The diameter of the floc gap is calculated, for example, as an estimated value based on the surface area of ​​the floc gap. However, the method for calculating the floc gap size is not particularly limited.

[0048] The coagulated sludge state acquisition unit 141 acquires, based on the estimation results of a trained coagulated sludge model, which of several pre-set ranges the state of the coagulated sludge falls into. In this example, the coagulated sludge state acquisition unit 141 acquires, based on the estimation results of the trained coagulated sludge model, which acquires, whether the floc gap size falls into a pre-set appropriate range, an excessive range (>appropriate range), or an under-range (<appropriate range). The trained coagulated sludge model estimates, for example, whether the floc gap size falls into an appropriate range, an excessive range, or an under-range for each input image data, and outputs an estimated value according to the estimation result. The estimated value will be described in detail later. Furthermore, the above appropriate range, excessive range, and under-range for the floc gap size are set in advance, for example, by a skilled worker at the sludge dewatering treatment facility 10.

[0049] The flocculated sludge state acquisition unit 141 determines whether the floc void size falls within the appropriate range, excessive range, or under-range based on the moving average of the estimated values ​​output from the flocculated sludge trained model (for example, the moving average from 10 minutes ago to the present). A judgment threshold is pre-set for determining which of the multiple ranges the floc void size falls within, based on the moving average of the estimated values ​​output from the flocculated sludge trained model. In this example, a first judgment threshold E1 and a second judgment threshold E2 are pre-set for determining whether the floc void size falls within the appropriate range, excessive range, or under-range based on the moving average of the estimate. The procedure for setting the first judgment threshold E1 and the second judgment threshold E2 will be described later.

[0050] The flocculated sludge state acquisition unit 141 determines that the flocculated sludge void size is within the appropriate range if the moving average of the estimated values ​​output from the trained flocculated sludge model is within the range of the first judgment threshold E1 and the second judgment threshold E2 (>E1). If the moving average of the estimated values ​​output from the trained flocculated sludge model is greater than the first judgment threshold E1, the flocculated void size is determined to be in the excessive range. Furthermore, if the moving average of the above estimated values ​​is less than or equal to the second judgment threshold E2, the flocculated void size is determined to be in the underrange range. In this example, the trained flocculated sludge model outputs estimated values ​​as described above, but it may also output a determination result for the flocculated void size based on these estimated values.

[0051] The moisture content acquisition unit 142 acquires the moisture content of dewatered sludge discharged from the screw press dewatering machine 60 using a moisture content trained model (first trained model) constructed by machine learning such as deep learning. The moisture content trained model is constructed using a learning algorithm, for example, a convolutional neural network (CNN), based on multiple image data of the dewatered sludge. In the moisture content acquisition unit 142, the image data of the dewatered sludge captured by the dewatered sludge imaging device 67 is input to the constructed moisture content trained model, and the moisture content of the dewatered sludge is acquired based on the estimation results output from the moisture content trained model.

[0052] As an example, the moisture content acquisition unit 142 acquires information on which of several pre-set ranges the moisture content of the dewatered sludge falls into, based on the estimation results of a pre-trained moisture content model. In this example, the moisture content acquisition unit 142 acquires information on which of two ranges, a first range and a second range (> first range), the moisture content of the dewatered sludge falls into, based on the estimation results of a pre-trained moisture content model. The pre-trained moisture content model estimates, for example, whether the moisture content of the dewatered sludge falls into the first range or the second range for each input image data, and outputs an estimated value according to the estimation result. The first range and the second range are set in advance, for example, by a skilled worker at the sludge dewatering treatment facility 10.

[0053] Furthermore, the moisture content acquisition unit 142 determines whether the moisture content of the dewatered sludge falls within the first or second range based on the moving average of the estimated values ​​output from the moisture content trained model (for example, the moving average from 10 minutes ago to the present). A determination threshold is pre-set for the moving average of the estimated values ​​output from the moisture content trained model to determine which of the multiple ranges the moisture content of the dewatered sludge falls within. In this example, a third determination threshold E3 is pre-set for the moving average of the estimated values ​​output from the moisture content trained model to determine whether the moisture content of the dewatered sludge falls within the first or second range. The procedure for setting the third determination threshold E3 will be described later.

[0054] In the moisture content acquisition unit 142, if the moving average of the estimated values ​​is less than or equal to the third judgment threshold E3, it is determined that the moisture content of the dewatered sludge falls within the first range. On the other hand, if the moving average of the estimated values ​​is greater than the third judgment threshold E3, it is determined that the moisture content of the dewatered sludge falls within the second range. In this example, the moisture content trained model outputs the above estimated values, but it may also be configured to output a moisture content determination result based on the estimated values.

[0055] The injection rate determination unit 143 determines the injection rate of the coagulant to the sludge based on the results obtained by the coagulated sludge state acquisition unit 141 and the moisture content acquisition unit 142. For example, as shown in the thick frame in Figure 3, if the floc gap size obtained by the coagulated sludge state acquisition unit 141 is within the appropriate range and the moisture content of the dewatered sludge obtained by the moisture content acquisition unit 142 is within the first range, the injection rate determination unit 143 selects a coagulant injection rate set to classification C2 as a coagulant injection rate suitable for the current situation.

[0056] The change instruction unit 144 instructs the system to change (update) the coagulant injection rate as needed, based on the coagulant injection rate determined by the injection rate determination unit 143. For example, the change instruction unit 144 instructs the system to change the operating state of the third pump 72 as needed. More specifically, if the coagulant injection rate determined by the injection rate determination unit 143 is different from the currently set coagulant injection rate, the change instruction unit 144 instructs the system to change the operating state of the third pump 72. The change instruction from the change instruction unit 144 is transmitted to the control panel 80 via the communication interface device 160. The PLC 81 adjusts the control of the third pump 72 as appropriate in response to the change instruction from the change instruction unit 144. In this way, in the system including the injection rate determination device 100, the amount of coagulant injected into the coagulation tank 50 of the sludge dewatering treatment facility 10 is automatically adjusted according to the coagulant injection rate determined by the injection rate determination unit 143.

[0057] The presentation instruction unit 145 issues a presentation instruction to display information (determined information) of the coagulant injection rate determined by the injection rate determination unit 143 to the user. For example, the presentation instruction unit 145 transmits a presentation instruction to the control panel 80 via the communication interface device 160, along with the determined information of the coagulant injection rate by the injection rate determination unit 143, to display this determined information to the user. The transmitted determined information is then displayed to the user via the PC 83 of the control panel 80.

[0058] For example, in PC83, the currently set coagulant injection rate is always displayed and is updated as needed based on the decision information transmitted from the presentation instruction unit 145. Furthermore, the instruction to present the decision information from the presentation instruction unit 145 may be output to the output device 130 via the input / output interface device 170. In other words, the information on the coagulant injection rate determined by the injection rate determination unit 143 may be presented to the user via the output device 130.

[0059] The decision information presented to the user by the PC83, etc., only needs to allow the user to understand the determined coagulant injection rate, and the specific display content is not particularly limited. Figure 4 shows an example of the display of decision information presented to the user. As shown in the example in Figure 4, the decision information presented to the user may also be information on the difference from the currently set coagulant injection rate. In this example, "Maintain current status" is displayed as the coagulant injection rate when sludge condition classification C2 is selected. In this case, the user can understand that the control of the third pump 72 will not be changed. Also, for example, "+5%" is displayed as the coagulant injection rate when classification C1 is selected. In this case, the user can understand that the control of the third pump 72 will be changed and the coagulant injection rate will be increased by 5% from the current value.

[0060] Furthermore, in this example, as described above, the amount of coagulant injected into the coagulation tank 50 is automatically adjusted so that it matches the coagulant injection rate determined by the injection rate determination unit 143, in response to the instructions from the change instruction unit 144. For this reason, the processor 140 does not necessarily have to be equipped with a presentation instruction unit 145. On the other hand, the processor 140 may be equipped only with the presentation instruction unit 145 and without the change instruction unit 144. In this case, the user can check the information presented on the PC 83, etc., and manually change the control of the third pump 72 to adjust the coagulant injection rate as appropriate.

[0061] As described above, in the injection rate determination device 100 of this embodiment, the coagulant injection rate is determined based on the state of the coagulated sludge, which is obtained from image data of the coagulated sludge using a pre-trained model for coagulated sludge, and the moisture content of the dewatered sludge, which is obtained from image data of the dewatered sludge using a pre-trained model for moisture content. This makes it possible to appropriately determine the coagulant injection rate in real time. In particular, by determining the coagulant injection rate based on the moisture content of the dewatered sludge, which is obtained from image data of the dewatered sludge using a pre-trained model for moisture content, the sludge state can be grasped more appropriately, and the coagulant injection rate can be determined more appropriately.

[0062] By the way, as mentioned above, when determining the coagulant injection rate from the image data of the coagulated sludge and the image data of the dewatered sludge using the injection rate determination device 100, certain preliminary preparations are required. In order to determine the coagulant injection rate, it is necessary to create the coagulant injection rate determination table as described above as a preliminary preparation, but other preliminary preparations such as the following are also required. The preliminary preparations will be explained in detail below.

[0063] In this example, as a preliminary step, Model 152 is trained using a convolutional neural network learning algorithm based on multiple image data of coagulated sludge to construct a trained model for coagulated sludge. Furthermore, a judgment threshold is set for determining the state of the coagulated sludge based on the estimation results of the trained model. Similarly, Model 152 is trained using a convolutional neural network learning algorithm based on multiple image data of dewatered sludge to construct a trained model for moisture content. Furthermore, a judgment threshold is set for determining the moisture content of the dewatered sludge based on the estimation results of the trained model.

[0064] [Construction of a trained model for coagulated sludge] First, the sludge dewatering treatment facility 10 is operated experimentally with a coagulant injection rate determined by a skilled worker. The coagulant injection rate is basically determined so that the floc gap size falls within the appropriate range, but it is intentionally increased or decreased temporarily to cause the floc gap size to fall within the excessive or insufficient range. While the sludge dewatering treatment facility 10 is operated experimentally, image data of the coagulated sludge for several days (for example, 3 to 5 days) is collected. In this example, image data of the coagulated sludge, including flocs, is collected using the coagulated sludge imaging device 56. Along with collecting the image data of the coagulated sludge, it is determined whether the floc gap size corresponding to each image data falls within the appropriate range, excessive range, or insufficient range described above. This determination of the floc gap size is performed, for example, by a skilled worker of the sludge dewatering treatment facility 10 visually. Note that the image data of the coagulated sludge in the preliminary preparation does not necessarily have to be captured by the coagulated sludge imaging device 56.

[0065] Next, each collected image data is labeled. For example, each image data is labeled according to which of the predefined ranges the floc gap size corresponding to each collected image data falls within. In this example, each image data is labeled according to whether the floc gap size falls within the appropriate range, the excessive range, or the under-range range. This creates training data for constructing a trained model of flocculated sludge. Then, using the labeled image data of flocculated sludge (training data), model 152 is trained using so-called supervised learning to construct a trained model of flocculated sludge.

[0066] [Setting the first and second threshold values] As described above, when image data of coagulated sludge is input to the constructed coagulated sludge trained model, it estimates whether the floc gap size falls within the appropriate range, the excessive range, or the insufficient range, and outputs an estimated value according to the estimation result. In this example, the coagulated sludge state acquisition unit 141 determines the floc gap size from the moving average of the estimated values ​​output from the coagulated sludge trained model. Therefore, as one of the preliminary steps, a first judgment threshold E1 and a second judgment threshold E2 are set to determine the floc gap size based on the moving average of the estimated values.

[0067] First, the sludge dewatering treatment facility 10 is operated experimentally for several days (for example, 8 to 10 days) while a skilled worker adjusts the coagulant injection rate so that the floc gap size falls within the appropriate, excessive, and insufficient ranges. Then, based on the image data of the coagulated sludge during this period, the coagulated sludge trained model is used to estimate the floc gap size.

[0068] In this process, the trained model for flocculated sludge will output a predetermined estimated value depending on which range the floc spacing size is estimated to fall within. In this example, if the trained model for flocculated sludge estimates that the floc spacing size falls within the appropriate range, an estimated value of "1" will be output. If the floc spacing size is estimated to fall within the excessive range, an estimated value of "2" will be output. If the floc spacing size is estimated to fall within the underrange, an estimated value of "0" will be output. Note that the above estimated values ​​are merely examples. Each estimated value can be set arbitrarily, as long as the estimated value for the appropriate range is an intermediate value.

[0069] Figure 5 shows the estimated values ​​output from the coagulated sludge trained model, the moving average of the estimated values, and the changes in the skilled worker's operating position when the sludge dewatering equipment is operated for several days. In the example shown in Figure 5, the periods T1, T3, T5, T7, T9, and T11 during which the sludge dewatering equipment 10 is in operation are periods in which skilled workers determined, by visual inspection, etc., that the floc gap size was within the appropriate range, and set the skilled worker's operating value corresponding to the above estimated value to "1". These can be said to be the normal operating period of the sludge dewatering equipment 10. On the other hand, the periods T2, T6, and T8 are periods in which skilled workers determined, by visual inspection, etc., that the floc gap size was within the excessive range, and set the skilled worker's operating value to "2". These can be said to be the off-normal operating period of the sludge dewatering equipment 10. Furthermore, periods T4 and T10 are the periods during which skilled workers determined that the floc gap size was within the under-range and set the skilled worker operating value to "1," and can be considered periods of non-normal operation for the sludge dewatering treatment equipment 10.

[0070] In the example shown in Figure 5, during the normal operating period, the trained model for flocculated sludge primarily outputs an estimated value of "1," but may also output an estimated value of "2" or "0" due to misestimation of the floc pore size. Furthermore, during periods T2, T6, and T8 of the non-normal operating period, the trained model for flocculated sludge primarily outputs an estimated value of "2," but may also output an estimated value of "1" or "0" due to misestimation. In addition, during periods T4 and T10 of the non-normal operating period, the trained model for flocculated sludge primarily outputs an estimated value of "0," but may also output an estimated value of "1" or "2" due to misestimation. Therefore, if the range of the floc pore size is determined based on the estimated value at a certain point in time, there is a risk that the range of the floc pore size will be incorrectly determined. For this reason, in this example, the floc pore size is determined based on the moving average of the estimated values ​​output from the trained model for flocculated sludge.

[0071] As a preliminary step, the first and second judgment thresholds E1 and E2 are set for the moving average of the estimated values ​​output from the trained model for flocculated sludge. Specifically, as described above, the moving average of the above estimated values ​​during the normal operating period within the period in which the sludge dewatering treatment facility 10 is operated is calculated, and the standard deviation σa of that moving average is also calculated. Then, the first and second judgment thresholds E1 and E2 are set based on this standard deviation σa. In this example, the first judgment threshold E1, which is the boundary between the appropriate range and the excessive range, is set by adding three times the standard deviation σa to the estimated value "1" when the trained model for flocculated sludge estimates that the floc gap size is within the appropriate range (1+3σa). In addition, the second judgment threshold E2, which is the boundary between the appropriate range and the under-range, is set by subtracting three times the standard deviation σb from the estimated value "1" (1-3σa).

[0072] In the injection rate determination device 100, the range of floc gap size is determined based on the first and second determination thresholds E1 and E2 set in this manner. As shown in Figure 6 as an example, if the moving average of the estimated values ​​output from the learned coagulated sludge model is less than or equal to the first determination threshold E1 and greater than the second determination threshold E2, the coagulated sludge state acquisition unit 141 determines that the floc gap size at that time falls within the appropriate range. On the other hand, if the moving average of the estimated values ​​output from the learned coagulated sludge model is greater than the first determination threshold E1, the coagulated sludge state acquisition unit 141 determines that the floc gap size at that time falls within the excessive range. Furthermore, if the moving average of the estimated values ​​output from the learned coagulated sludge model is less than or equal to the second determination threshold E2, the coagulated sludge state acquisition unit 141 determines that the floc gap size at that time falls within the under-range. For example, at time Ta in Figure 6, the moving average of the above estimated values ​​is less than or equal to the second determination threshold E2, so the floc gap size is determined to fall within the under-range. At time Tb, the moving average of the above estimates is less than or equal to the first judgment value E1 and greater than the second judgment threshold E2, so the floc gap size is determined to be within the appropriate range. At time Tc, the moving average of the above estimates is greater than the first judgment threshold E1, so the floc gap size is determined to be within the excessive range.

[0073] In short, the coagulated sludge state acquisition unit 141 in this example determines whether the floc void size falls within the appropriate range based on the moving average of the estimated values ​​output from the coagulated sludge trained model. If the floc void size does not fall within the appropriate range, it is determined that the floc void size falls within the excessive or excessive range. More specifically, if the floc void size exceeds the appropriate range, it is determined that the floc void size falls within the excessive range. Also, if the floc void size falls below the appropriate range, it is determined that the floc void size falls within the excessive range.

[0074] By determining the floc pore size based on the moving average of the estimated values ​​output from the trained model of flocculated sludge in this way, the flocculated sludge state acquisition unit 141 can more accurately acquire the floc pore size in real time.

[0075] In this embodiment, an example was described in which the first judgment threshold E1 is (1+3σa). However, the first judgment threshold E1 is not limited to this and can be set appropriately based on the standard deviation σa of the moving average of the estimated values ​​output from the flocculated sludge trained model. For example, the first judgment threshold E1 may be (1+2σa) or (1+σa). Similarly, the second judgment threshold E2 can be set based on the standard deviation σa, for example, (1-2σa) or (1-σa). Furthermore, each threshold can be arbitrarily set by the worker.

[0076] [Building a pre-trained model for moisture content] Next, we will explain the construction of a pre-trained moisture content model, which is carried out as one of the preliminary preparations. Basically, it is the same as the construction of the coagulated sludge pre-trained model described above. First, the sludge dewatering treatment facility 10 is operated experimentally with a coagulant injection rate determined by the operation of a skilled worker. At this time, the coagulant injection rate is basically determined so that the moisture content of the dewatered sludge is within the first range described above, but it is temporarily adjusted so that the moisture content of the dewatered sludge is within the second range described above (> first range). This example can also be said to be an example in which the first range is considered to be the range of appropriate moisture content for dewatered sludge. However, the second range can also be considered to be the range of appropriate moisture content for dewatered sludge.

[0077] During the trial operation of the sludge dewatering treatment facility 10, image data of coagulated sludge for several days (for example, 3 to 5 days) will be collected. In this example, image data of dewatered sludge captured by the dewatered sludge imaging device 67 will be collected. In addition to collecting image data of the dewatered sludge, the moisture content of the dewatered sludge corresponding to each image data will be measured. The method of measuring the moisture content is not particularly limited and can be measured using existing measuring instruments, etc. Alternatively, a skilled worker at the sludge dewatering treatment facility 10 may visually determine the moisture content of the dewatered sludge.

[0078] Next, each collected image data is labeled. For example, each image data is labeled according to which of the pre-defined ranges the moisture content of the dewatered sludge corresponding to each image data falls within. In this example, each image data is labeled according to whether the moisture content of the dewatered sludge falls within the first range or the second range. This creates training data for constructing a moisture content-trained model. Then, using the labeled image data of dewatered sludge (training data), model 152 is trained using so-called supervised learning to construct a moisture content-trained model.

[0079] For example, a first range is set for dewatered sludge with a moisture content of less than 86% (e.g., 80% or more but less than 86%), and a second range is set for dewatered sludge with a moisture content of 86% or more (e.g., 86% or more but less than 90%). In this example, the moisture content of the dewatered sludge is specified in the range of 80% to 90%, but the range of moisture content of the dewatered sludge to be specified is not particularly limited and may be set appropriately according to the properties of the sludge, etc.

[0080] [Setting the third threshold] As described above, when the constructed moisture content trained model receives image data of dewatered sludge, it estimates whether the moisture content of the dewatered sludge falls within a first or second range, and outputs an estimated value according to the estimation result. In this example, the moisture content acquisition unit 142 determines the range of moisture content of the dewatered sludge from the moving average of the estimated values ​​of the moisture content trained model. Therefore, as one of the preliminary steps, a third determination threshold E3 is set for determining the range of moisture content based on the moving average of the estimated values ​​of the moisture content trained model.

[0081] First, the sludge dewatering treatment facility 10 is operated experimentally for several days (for example, 8 to 10 days) while adjusting the coagulant injection rate by skilled workers so that the moisture content of the dewatered sludge falls within the first and second ranges described above. During this time, the moisture content of the dewatered sludge is estimated using a pre-trained moisture content model based on the image data of the dewatered sludge captured by the dewatered sludge imaging device 67. At this time, if the pre-trained moisture content model estimates that the moisture content of the dewatered sludge is within the first range, the pre-trained moisture content model outputs an estimated value of "0". If the moisture content of the dewatered sludge is estimated to be within the second range, the pre-trained moisture content model outputs an estimated value of "1". Note that the above estimated values ​​are merely examples. The estimated value for the first range should be smaller than the estimated value for the second range.

[0082] Figure 7 shows the changes in estimated values, moving averages of estimated values, and measured values ​​output from a pre-trained moisture content model when a sludge dewatering treatment facility is operated for several days. The measured value here is equivalent to the "estimated value" set based on the measurement results from measuring instruments, etc. For example, if the moisture content is obtained within the first range as a measurement result, the measured value is set to "0". If the moisture content is obtained within the second range as a measurement result, the measured value is set to "1".

[0083] In the example shown in Figure 7, the periods T21, T23, T25, T27, T29, and T31 during which the sludge dewatering treatment equipment 10 is in operation are periods when the measured value is set to "0", and can be said to be the normal operating period of the sludge dewatering treatment equipment 10. On the other hand, the periods T22, T24, T26, T28, and T30 are periods when the measured value is set to "1", and can be said to be the off-normal operating period of the sludge dewatering treatment equipment 10.

[0084] In the example shown in Figure 7, during the normal operating period, the moisture content trained model primarily outputs an estimated value of "0," but may also output an estimated value of "1" due to misestimation. Similarly, during the non-normal operating period, the moisture content trained model primarily outputs an estimated value of "1," but may also output an estimated value of "0" due to misestimation. Therefore, if the moisture content of dewatered sludge is determined based on an estimated value at a given point in time, there is a risk of incorrect determination of the moisture content. Accordingly, in this example, the moisture content of dewatered sludge is determined based on a moving average of the estimated values ​​output from the moisture content trained model.

[0085] As a preliminary step, a third threshold E3 is set for the moving average of the estimated values ​​of the pre-trained moisture content model. Specifically, as described above, the moving average of the above estimated values ​​during the normal operating period within the period in which the sludge dewatering treatment equipment 10 is operated is calculated, and the standard deviation σb of that moving average is also calculated. Then, the third threshold E3 is set based on this standard deviation σb. In this example, the value obtained by adding three times the standard deviation σb to the estimated value "0" when the moisture content of the dewatered sludge is estimated to be within the first range (0+3σb) is set as the third threshold E3, which is the boundary between the first range and the second range.

[0086] In the injection rate determination device 100, the moisture content of the dewatered sludge is determined based on the third judgment threshold E3 set in this manner. As shown in Figure 8 as an example, if the moving average of the estimated values ​​of the moisture content trained model is less than or equal to the third judgment threshold E3, the moisture content acquisition unit 142 determines that the moisture content of the dewatered sludge at that time falls within the first range. On the other hand, if the moving average of the estimated values ​​of the moisture content trained model is greater than the third judgment threshold E3, the moisture content acquisition unit 142 determines that the moisture content of the dewatered sludge at that time falls within the second range. For example, at times Tc, Td, and Te in Figure 8, the moving average of the estimated values ​​of the moisture content trained model is less than or equal to the third judgment threshold E3 in all cases, so the moisture content of the dewatered sludge is determined to fall within the first range.

[0087] In short, the moisture content acquisition unit 142 in this example determines whether the moisture content of the dewatered sludge falls within a first range based on the moving average of the estimated values ​​of the moisture content trained model. If the moisture content of the dewatered sludge does not fall within the first range, it is determined that the moisture content of the dewatered sludge falls within a second range. More specifically, if the moisture content of the dewatered sludge exceeds the first range, it is determined that the moisture content of the dewatered sludge falls within a second range.

[0088] By determining the moisture content of the dewatered sludge based on a moving average of the estimated values ​​from a pre-trained moisture content model, the moisture content acquisition unit 142 can more accurately acquire the moisture content of the dewatered sludge in real time.

[0089] In this embodiment, an example was described in which the third judgment threshold E3 was set to (0+3σb). However, the third judgment threshold E3 is not limited to this and can be set appropriately based on the standard deviation σb of the moving average of the estimated values ​​output from the moisture content trained model. The third judgment threshold E3 may be, for example, (0+2σb) or (0+σb). Furthermore, the threshold may be arbitrarily set by the worker.

[0090] Although one embodiment of the technology disclosed herein has been described above, the technology disclosed herein is not limited to the embodiment described above. It goes without saying that the technology disclosed herein can be modified in various ways without departing from its essence.

[0091] For example, in the above-described embodiment, an example was described in which the injection rate determination unit 143 determines the coagulant injection rate from the coagulant injection rate determination table based on the floc gap size, which is one of the floc states acquired by the coagulated sludge state acquisition unit 141, and the moisture content of the dewatered sludge acquired by the moisture content acquisition unit 142. In other words, an example was described in which the coagulant injection rate determination table sets the coagulant injection rate for each classification of coagulated sludge identified by the floc gap size acquired by the coagulated sludge state acquisition unit 141 and the moisture content of the dewatered sludge acquired by the moisture content acquisition unit 142. However, the method for determining the coagulant injection rate is not limited to this.

[0092] Figure 9 shows another example of a flocculant injection rate determination table presented to the user. As shown in one example in Figure 9, the flocculant injection rate determination table may have flocculant injection rates set for classifications C11 to C19 of flocculated sludge, which are identified by floc size, one of the floc conditions, and the water content of the dewatered sludge.

[0093] In this case, the flocculation sludge state acquisition unit 141 acquires whether the floc size (floc state) falls within a preset range of appropriate, excessive (>appropriate range), or insufficient (<appropriate range) based on the estimation results of the flocculation sludge trained model. The flocculation sludge trained model estimates, for example, whether the floc size falls within the appropriate, excessive, or insufficient range for each input image data, and outputs an estimated value according to the estimation result. The above appropriate, excessive, and insufficient ranges for floc size are set in advance, for example, by a skilled worker at the sludge dewatering treatment facility 10.

[0094] Furthermore, in the embodiments described above, the water content of the dewatered sludge was classified into two ranges, a first range and a second range. However, as shown in Figure 9, it may also be classified into three ranges: a first range, a second range (> first range), and a third range (> second range). Moreover, the water content of the dewatered sludge may be classified into four or more ranges. Similarly, the flocculation state may also be classified into four or more ranges, or into two ranges.

[0095] Furthermore, while the above-described embodiment explained an example of presenting information on determining the coagulant injection rate to the user, the information presented to the user is not limited to this. For example, the user may be presented with a graph of the moving average of estimated values ​​output from a pre-trained model for coagulated sludge, or a graph of the moving average of estimated values ​​output from a pre-trained model for water content. This allows the user, who is a skilled worker, to fine-tune the coagulant injection rate as needed based on this information. This enables more appropriate adjustment of the coagulant injection rate.

[0096] Furthermore, in the above-described embodiment, an example was explained in which the operation of the third pump 72 is changed according to the coagulant injection rate determined by the injection rate determination device 100 to adjust the amount of coagulant injected into the coagulation tank 50. However, the method of adjusting the coagulant injection rate is not limited to this. For example, it is also possible to relatively adjust the coagulant injection rate by controlling the operation of the second pump 32, etc., to adjust the amount of inflowing sludge flowing into the coagulation tank 50.

[0097] Furthermore, although the above-described embodiment described an example in which the injection rate determination device 100 is applied to a sludge dewatering treatment facility 10, the sludge treatment facility to which the injection rate determination device 100 is applied is not limited to a sludge dewatering treatment facility 10. The injection rate determination device 100 described above can also be applied to, for example, a sludge thickening treatment facility where sludge thickening is performed. [Explanation of Symbols]

[0098] 10…Sludge dewatering treatment equipment 20...Sludge storage tank 21...Level 1 total 22... Pump No. 1 23...First piping 24…Flow meter 30…Sludge transfer tank 31...Level 2 total 32... Second pump 33...Second piping 50…Flocculation tank (flocculation mixing tank) 51... Stirring device 52…Agitation blade 53… Stirring motor 54...Fourth pipe 55...3rd level total 56…Coagulated sludge imaging device (camera) 60...Screw press machine (screw press dehydrator) 61…Screen 62... Screw 64…Drive motor 65...Pressing plate 66... ​​Dewatered sludge container 67…Dewatered sludge imaging device (camera) 70... Coagulant tank 71…Supply piping 72... Third pump 80... Control panel 81…PLC (Programmable Logic Controller) 82... Inverter 83...PC (Personal Computer) 100...Injection rate determination device 110...Arithmetic device 120...Input device 130…Output device 140… Processor 141...Collated sludge state acquisition unit 142…Moisture content acquisition section 143...Injection rate determination section 144... Change instruction section 145...Presentation instruction section 150...memory 151... Control program 152…Model 153…Image data 154…Settings Information 160...Communication Interface Device 170… Input / Output Interface Device

Claims

1. This device is applied to a sludge treatment facility that uses a screw press to compress flocculated sludge containing flocs generated by injecting a flocculant into sludge in a flocculation tank, and determines the injection rate of the flocculant into the sludge. A moisture content acquisition unit inputs image data of the compressed sludge discharged from the screw press machine, which is captured at predetermined intervals by a compressed sludge imaging device, into a first trained model trained using the image data of the compressed sludge, and acquires the moisture content of the compressed sludge based on the estimation results of the first trained model. The system includes an injection rate determination unit that determines the injection rate of the coagulant into the sludge based on the results obtained by the moisture content acquisition unit. Injection rate determination device.

2. In the injection rate determination device according to claim 1, The moisture content acquisition unit inputs image data of the compressed sludge before it falls out of the screw press machine into the first trained model. Injection rate determination device.

3. In the injection rate determination device according to claim 1, The moisture content acquisition unit determines the moisture content of the compressed sludge based on the moving average of the estimated values ​​output by the first trained model. Injection rate determination device.

4. In the injection rate determination device according to claim 3, The moisture content acquisition unit determines the moisture content of the compressed sludge based on the standard deviation calculated from the moving average. Injection rate determination device.

5. In the injection rate determination device according to claim 1, The system includes a second trained model, which is trained using the image data of the coagulated sludge, and an image data of the coagulated sludge captured in the coagulation tank by a coagulated sludge imaging device, and a coagulated sludge state acquisition unit that acquires the state of the coagulated sludge based on the estimation results of the second trained model. The injection rate determination unit determines the injection rate of the coagulant into the sludge based on the results obtained by the moisture content acquisition unit and the results obtained by the coagulated sludge state acquisition unit. Injection rate determination device.

6. In the injection rate determination device according to claim 5, The coagulated sludge state acquisition unit determines the state of the coagulated sludge based on the moving average of the estimated values ​​output by the second trained model. Injection rate determination device.

7. In the injection rate determination device according to claim 6, The coagulated sludge state acquisition unit determines the state of the coagulated sludge based on the standard deviation calculated from the moving average. Injection rate determination device.

8. In the injection rate determination device according to claim 5, The aforementioned coagulated sludge state acquisition unit acquires the floc size as the state of the coagulated sludge. Injection rate determination device.

9. In the injection rate determination device according to claim 5, The aforementioned coagulated sludge state acquisition unit acquires the floc gap size as the state of the coagulated sludge. Injection rate determination device.

10. In the injection rate determination device according to claim 5, The aforementioned coagulated sludge state acquisition unit acquires the density of the flocs, the viscosity of the coagulated sludge, the transparency of the interpore water between the flocs, or the color of the flocs as the state of the coagulated sludge. Injection rate determination device.

11. In the injection rate determination device according to claim 1, The system includes a presentation instruction unit that instructs the injection rate determination unit to present the determined information to the user. Injection rate determination device.

12. In the injection rate determination device according to claim 1, The system includes a change instruction unit that instructs a change in the amount of coagulant injected into the coagulation tank according to the information determined by the injection rate determination unit. Injection rate determination device.

13. This method is applied to a sludge treatment facility that uses a screw press to compress flocculated sludge containing flocs generated by injecting a flocculant into sludge in a flocculation tank, and determines the injection rate of the flocculant into the sludge. Image data of the compressed sludge discharged from the screw press machine, captured by a compressed sludge imaging device, is input to a first trained model trained using the image data of the compressed sludge, and the water content of the compressed sludge is obtained based on the estimation results of the first trained model. Based on the water content of the compressed sludge obtained, the injection rate of the coagulant into the sludge is determined. Method for determining injection rate.

14. In the injection rate determination method according to claim 13, When obtaining the water content of the compressed sludge, image data of the compressed sludge before it falls out of the screw press is input to the first trained model. Method for determining injection rate.

15. In the injection rate determination method according to claim 13, The image data of the coagulated sludge captured in the coagulation tank is input to a second trained model trained using the image data of the coagulated sludge, and the state of the coagulated sludge is obtained based on the estimation results of the second trained model. Based on the water content of the compressed sludge and the state of the flocculated sludge, the injection rate of the flocculant into the sludge is determined. Method for determining injection rate.

Images