JPH01315393A - Device for discriminating treating condition in activated sludge process - Google Patents

Abstract

PURPOSE:To easily control the operation of an activated sludge process even by an average operator by using the knowledge of a knowledge base to infer the treating condition of an activated sludge process from the kind of appearing microbes and the population distribution, and diagnosing the process. CONSTITUTION:The relation between the ecological information of the activated sludge microbes and sewage treating condition as the indices is registered in the knowledge base 21 of the expert system 20 with the knowledge renewable. The treating condition is inferred and diagnosed by a inference engine 22 from the input regarding the kind of the appearing microbes the population distribution, the appearance of the process, and environmental factors in the process in an aeration tank 3, a final settling basin 5, etc., and the process is controlled by the monitoring and controlling system 10 based on the diagnosis. By this method, the operation control of the activated sludge process and the prediction of the activated sludge condition and treating condition can be accurately performed even by an average operator, and the treating system can be accurately maintained and controlled.

Description

DETAILED DESCRIPTION OF THE INVENTION A. Industrial Application Field The present invention is used for diagnosing and monitoring and controlling the treatment status of sewage treatment processes using activated sludge. This invention relates to an activated sludge process treatment status determination device that uses information regarding appearance to enable more accurate determination of the treatment status of an activated sludge process.

B4 Summary of the Invention The present invention provides an activated sludge process treatment status determination device that is equipped with a knowledge base that can update the knowledge and an inference mechanism based on that knowledge, and that detects appearance information of activated sludge microorganisms that serve as indicators and the treatment status of the activated sludge process. While accumulating knowledge about
Infer the treatment status of the process from that knowledge and necessary input data such as emerging microorganism data and process appearance data 9
By diagnosing the system and outputting countermeasures if necessary, it is possible to more accurately manage the operation of the activated sludge treatment process using information on the ecology of microorganisms.

C9 Prior Art Conventionally, systems for purifying sewage using activated sludge have been widely adopted for the treatment of urban sewage and the like. Activated sludge is composed of a population of microorganisms that decompose and assimilate organic matter in wastewater.

FIG. 1θ is a configuration diagram of a sewage treatment system showing a conventional activated sludge process. 1 is a settling tank that removes sediment from sewage, 2 is a first settling tank that removes fine suspended matter, and 3 is a blower 4 that mixes air into the sewage to supply oxygen and grow aerobic microorganisms, making the sewage wastewater An aeration tank promotes biological treatment of organic matter therein, and 5 is a final settling tank that performs solid-liquid separation. Sewage is sent by a sewage pump 6, organic matter is removed by activated sludge in the aeration tank 3, and solid-liquid separation is carried out into activated sludge and supernatant water in the final settling tank 5. The supernatant water is disinfected and then used as treated water in the system. Released outside the system. On the other hand, a part of the activated sludge is returned to the aeration tank 3 by the return sludge pump 7 and reused, but the multiplied portion is first supplied as surplus sludge to the settling tank 2 by the surplus sludge pump 8 and collected at the bottom. The sludge is treated in a sludge treatment system by a sludge extraction pump 9 and then incinerated.

In the activated sludge process mentioned above, in order to avoid improving the quality of treated water, operational management is carried out to control the amount of returned sludge and the amount of air mixed in the aeration tank 3 by the blower 4. This was done using dissolved oxygen concentration (DO) and apparent microbial concentration (activated sludge suspended solids concentration MLSS) as indicators.

D0 Problems to be Solved by the Invention It has become clear that the organic matter removal and solid-liquid separation ability (sedimentability of activated sludge) using activated sludge in the above-mentioned conventional technology are greatly influenced by the ecology of microorganisms. However, since it is difficult to quantify the relationship between the types of microorganisms that appear in the system and the purification processing capacity, information on the types of microorganisms and their population distribution is limited to observation by specific operators using a microscope, and Currently, it is hardly used for management purposes.

However, while observing microorganisms at the same time among experienced operators, they discovered a certain relationship between the quality of treated water and the type of microorganisms, and there is a movement to apply this relationship to operational management. It has become. The number of these operators is limited and most of them are experts, so it cannot be said that there are many opportunities to systematically put their knowledge to practical use, and in addition, it is difficult to pass on knowledge to new operators. . For this reason, although the significance of operational management based on the types and distribution of microorganisms has been recognized, it has not become widely used. As described above, in the conventional activated sludge process, it has been a challenge to reflect information on the ecology of microorganisms in operational management.

The present invention was devised to solve the above-mentioned problems, and is an activated sludge treatment process that enables more accurate operation management of activated sludge treatment processes using information on the ecology of activated sludge microorganisms and the appearance of the process. An object of the present invention is to provide a sludge process treatment state determination device.

Means for Solving Problem E6 The activated sludge process treatment status determination device of the present invention to achieve the above object has the following features: An indicator of the treatment status of the activated sludge process among activated sludge microorganisms observed in the activated sludge process. A knowledge base in which the relationship between appearance information of indicator microorganisms and the processing status described above can be registered in advance and updated; a man-machine interface for inputting the types and numbers of currently occurring microorganisms; and knowledge of the above knowledge base. The invention is characterized by comprising an inference mechanism for inferring and diagnosing the treatment status of the activated sludge process from the type and population distribution of the microorganisms that appear.

In addition, when diagnosing an activated sludge process, the above reasoning mechanism uses external appearance information and DO value of the process.

It is preferable to also use environmental factor information such as pH value and MSLL value in order to increase the accuracy of diagnosis, and further,
It is suitable for operation management that the above reasoning mechanism outputs a countermeasure based on the diagnosis result.

F. Function The present invention is equipped with a knowledge base that can update the knowledge and an inference mechanism that makes inferences based on this knowledge, and the relationship between the ecological information of activated sludge microorganisms that serves as an indicator and the treatment status of sewage is added to this knowledge base. Register and infer processing conditions from this knowledge and input information about microorganisms and process appearance, environmental factors, and the appearance of microorganisms within the process.

By performing a diagnosis and outputting countermeasures if necessary, anyone can utilize the knowledge about the ecology of activated sludge microorganisms and process treatment status, which has been accumulated through research and experience, to more accurately manage the activated sludge process. This will enable maintenance and management of water and improve the quality of treated water.

G. EXAMPLES Hereinafter, examples of the present invention will be described in detail based on the drawings.

FIG. 1 is a block diagram of an operation management system for an activated sludge process showing an embodiment of the present invention. Aeration tank 3. Blower 4. Final settling tank5. The return sludge pump 7 and the surplus sludge pump 8 are equipment that constitutes the conventional sewage treatment system shown in FIG. 4, and IO is a monitoring and control system for these. 11 is a flow meter for wastewater flowing into the aeration tank 3; 12 is a suspended solids concentration (MLSS) meter;
13 is a dissolved oxygen concentration (DO) meter, 13a is a pH (p
H) Total, 14 is aeration tank 3 by blower 4
15 is a valve for adjusting the air volume, 16 is a return sludge flow meter, 17 is a valve for adjusting the return sludge flow rate, 18 is an excess sludge flow meter, 19 is for adjusting the excess sludge flow rate. It is a valve to be adjusted. Each measuring instrument 11,1
2, 13.13a.

14, 16, and 18 are input to the supervisory control system IO via an appropriate transmission means, and from the supervisory control system IO, the valve 15 is controlled to adjust the air volume, and the return sludge pump 7 and valve 17 are controlled to send the sludge back. The sludge flow rate is adjusted,
Surplus sludge pump8. Valve 19 is controlled to adjust the flow rate of excess sludge.

The activated sludge process treatment status determination device of this embodiment is comprised of an expert system 20. The expert system 20 is a device for managing the operation of the sewage treatment system configured as described above using information accumulated through research and experience regarding the relationship between activated sludge microorganisms and process treatment conditions. Information on the appearance of indicator microorganisms that are indicators of the treatment status of the activated sludge process among the activated sludge microorganisms and knowledge regarding the relationship with the treatment status are stored in the knowledge base 2 in the expert system 20.
1 and stored as a rule. This knowledge can be updated by adding, deleting, or changing new information. 22 is an inference system in which an operator or the like matches microbial data on emerging microorganisms obtained through microscopic observation of samples with rules in the knowledge base, and infers and diagnoses the treatment status of the sewage treatment system, water quality, etc. It is an engine (inference mechanism). 23 is a CRTT for a man-machine interface used for displaying input specifications and medical examination forms when an operator inputs the above microorganism data.
It is a display.

FIG. 2 is a block diagram showing an example of the configuration of the expert system 20. As shown in FIG. 21 is a knowledge base, 22 is an inference engine, 24 is a microbial database, 25 is a human-machine interface, 26 is a human-powered device such as a keyboard,
23 is an output device such as a CRT display, and the expert system 20 is composed of the above-mentioned parts.

Knowledge Base 21 is based on the knowledge of microorganism input specifications and diagnosis sheet creation that performs preliminary diagnosis from data such as treated water quality and sludge treated water such as 5VI (sludge volume index), and selects indicator microorganism groups that appear frequently. processing diagnosis knowledge for diagnosing the state of sewage purification treatment from human data such as the state of emerging microorganisms entered manually from the input device 26;
It consists of three parts: knowledge of how to deal with it, which provides guidance on how to deal with it if necessary based on the diagnosis results. The microorganism database 24 stores data such as characteristics and shapes regarding microorganisms that are expected to appear, as image information, and plays an auxiliary role in microorganism identification. These knowledge bases 21
and the microorganism database 24 are added through the man-machine interface 25.

You can freely update, delete, change, etc. Inference engine 22
extracts knowledge (rules) from the knowledge base 21, compares the premise part of the rule with the contents of the input data, executes the conclusion part when they match, and outputs the above-mentioned microbial manual specification/diagnosis sheet to the output device 23, Outputs guidance to the flash fault monitoring and control system. Among the input data, the treated water quality is obtained by examining a sample with an analyzer or the like, and may be input manually via a monitoring control system, or may be input directly from the input device 26. In addition, SVI is calculated by dividing the sludge volume per unit volume of a sample left for 30 minutes by the MLSS value, so it may be calculated on the monitoring and control system side, or it may be calculated on the Exverton stem side from input data. .

The operation of the embodiment configured as above will be described.

FIG. 3 is a flowchart showing the processing procedure in the expert system. First, data such as sludge treated water is sent to the inference engine 2 from the monitoring control system and input device 26.
Enter 2. The inference engine 22 creates an input specification/diagnosis form based on this sewage treatment water data and the microbial input specification/diagnosis form creation knowledge of the knowledge base 21, and displays the input specification/diagnosis form on the CRT display 23. Based on these input specifications, the operator can determine the type and frequency of microorganisms observed using a microscope, etc.

Microorganism data such as population distribution is input, and at this time, if necessary, data regarding the characteristics and shape of the microorganism is called up from the microorganism database 24 so that beginners, general operators, etc. can check it. The inference engine makes inferences based on the microorganism data and processing diagnosis knowledge of the knowledge base 21, performs prediction 1 diagnosis, and provides guidance on the processing status of purification. Examples of such guidance include ``the load (the product of the amount of sewage inflow and the concentration) is too high'', ``the load is too low'',
"Good" etc. When it is diagnosed that the condition is not in good condition or that deterioration is expected, the inference engine 22 infers a countermeasure based on the countermeasure knowledge in the knowledge base 21 and outputs countermeasure guidance.

At this time, if there is insufficient data for diagnosis, the operator is requested to input additional data and a countermeasure is deduced. Guidance on how to deal with this problem includes the amount of valve control to adjust the flow rate of returned sludge, the amount of valve control to control the amount of air mixed in the air lane tank, etc. These guidances are output to the supervisory control system or displayed on the CRT display 23 via the man-machine interface 25, and are used, for example, to instruct the operator on driving operations, as well as to display the guidance on the controlled variable. It is also possible to automatically perform the above control based on the above.

FIG. 4 is a block diagram showing another example of the configuration of the expert system. Expert system 20 of this configuration example
', the configuration of the knowledge base 2bi is different from the configuration example shown in FIG. In other words, the knowledge base 2bi includes treatment diagnosis knowledge for diagnosing the treatment status of the sewage treatment system based on microbial information observed with a microscope, water quality prediction knowledge for predicting the quality of treated water, and information based on these inference results if necessary. It consists of three parts of coping method knowledge that provides guidance on coping methods, and these are stored as knowledge rules. The rest of the configuration is the same as that shown in FIG. 2, and the knowledge and data in the knowledge base 21' can be freely updated by adding, deleting, changing, etc. using the functions of the man-machine interface 25. In this configuration example, the microbial input specifications and diagnosis sheet can be created by using the man-machine interface 25, for example, VAX S, TATION] I (7) manufactured by Digital Equipment Corporation.
The input/output of the computer system 20' is performed on the screen of the CRT display 23.

The processing procedure of the expert system with the above configuration is basically the same as that shown in FIG. 3, but the processing for diagnosing the processing status of the sewage treatment system is performed as shown in the flowchart of the diagnostic processing shown in FIG. First, microorganism data such as the type of microorganism and population distribution, ie, the number of microorganisms appearing or the stage of appearance, is input via a man-machine interface. The processing state is estimated based on the input, and the certainty factor of the processing state is calculated each time the rule is hit, and the certainty factor of the processing state is changed.

On the other hand, the external appearance data of the aeration tank and sedimentation tank was input manually from the man-machine interface, and the pH value (pH) and Do value in the aeration tank were 7ML.
Environmental factor data such as SS values are input from a monitoring control system or the like. This environmental factor data may also be manually input from the human-powered device 26. From the input of these appearance data and environmental factor data, confidence calculations are performed to confirm the treatment status estimated by the microorganism group. The processing status is finally diagnosed from the above two confidence calculations, and the diagnosis results are displayed on a CRT display or the like via a man-machine interface.

Furthermore, the BOD (biochemical oxygen demand) etc. of the treated water can be predicted from this diagnosis result.

The processing status determined by microbial data is (1
) High load and cloudy treated water, (2) Low D0 or presence of anaerobic parts, (3) Low load or nitrification, (4) Disintegration of sludge due to increase in amoeba, etc., and (5) Inflow of toxic substances. It is. In order to estimate these phenomena, 24 species of microorganisms that have indicator properties for each phenomenon are calculated using contribution ratios (including frequency of occurrence).
They are classified into 16 groups taking into account the degree of influence that the number of units has on the phenomenon.

The number of occurrences in each group is set to 0 if it does not appear, and is divided into five stages as the number increases, and an integer from 1 to 5 is assigned to each group. That is, when the appearance stage of microorganism group n is Wn, Wn=(0, L--=5) (n=1-16). For example, the above five stages are given as shown in Table 1.

Table 1 Wn (stage of appearance) of microbial groups In addition, Table 2 shows the contribution rates to phenomena in classification examples of indicator microorganism groups.

(Leaving space below) Table 2 Contribution rate by microorganisms The toxicity contribution rate of microorganism F marked with * in Table 2 is Wn (previous)
- When Wn (current) = 5, contribution rate +0.5 ~ Vn (previous)
- When Wn (current) = 4, the contribution rate is +0.3. Also*
The toxicity contribution rate of the total number of microorganisms marked with * is the value shown in Table 3 based on the total value of the appearance stage Wn.

Table 3 Total contribution rate Table 23 In Table 3, negative numbers represent the degree of negation.

When expressed as a contribution rate Xnm (m is a phenomenon number), the certainty of a phenomenon caused by a certain group of microorganisms is CFnm=WnXXnm.

The confidence level for two types of microorganism groups (n, k) is each confidence level C
Calculate separately according to the sign of F as follows.

CF nm+ CF km+ CF r+mX CF
km (CFrv+>O, CF'km>0) CFnn++CFkn++CFnmXCFkm(CF
The above calculations (nm < O, CF km < 0) are applied in sequence, and finally, the numbers of opposite signs are added to obtain the confidence level based on all microorganisms.

The calculations are based on the phenomena estimated by the above microbial groups, environmental factors such as pH value (pH), DO, sludge color, etc.
By considering the apparent state (appearance data) such as the transparency of the sedimentation basin, the reliability of the phenomenon is further calculated. The reliability of this proof calculation is calculated using the reliability shown in Table 4 for each sign of each phenomenon in the same manner as the reliability calculation described above.

Table 4 Supporting Confidence In Table 4, the * mark indicates that black sludge floats to the surface, and the ** mark indicates that brown sludge floats to the surface.

An example of calculating the confidence level is shown below.

Table 5 is an example of manually inputting microbial data, and Table 6 is an example of inputting environmental factors and appearance data.

Table 5: Microbial data (blank below) Table 6: When calculating the confidence level by manually calculating the data beyond the environmental factors and appearance data, the result is as follows.

(1) Possibility of high load 0.73 (2) Possibility of existence of anaerobic part with low Do 0.53 (3)
Low load/possibility of sulfidation -0,67 (4) Possibility of sludge dismantling -0,28 (5) Possibility of toxic substances inflow -
0,24 Next, the operation of the man-machine interface in the above will be described. The man-machine interface performs functions such as human function 1 diagnosis of data necessary for diagnosis and display function of preliminary fA11 results. FIG. 6 is a diagram showing a microbial data input screen, and FIG. 7 is an explanatory diagram of microbial data input. When inputting microorganism data, the operator uses a mouse to move the cursor to the corresponding scratch (for example, Z) in the appearance list of the microorganism name row, and then presses the mouse button. Enter by pressing . As a result, on the screen, as shown in FIG. 7, the * mark moves from the previously input appearance list column to the newly input column Zl::.

In the appearance list, c c ′=r, − correspond to the appearance stages 5 to 1.0 of the microorganisms described above, respectively. Also,
As a help for beginners who do not know the microorganisms that have appeared, by placing the cursor on the microorganism name field (for example, A) and pressing the mouse button, a picture of the microorganism stored in the microorganism database and its characteristics (size and shape of the microorganism) can be displayed.
tc) has the ability to appear on the screen. 8th
The figure shows a data entry screen for appearance data, etc. Microbial data can be input by moving the cursor to the input value and pressing the mouse button, just like when inputting it manually. In the case of numerical data, press the mouse button and then input the numerical value, and in the case of a choice between the two, select one and press the mouse button to change the color of the background.

FIG. 9 shows an output screen for diagnosis results, etc. Windows
.

Window2. Window3. Windows 4 is on the same screen.

In Windows 2, inference is made from microorganism appearance information and input data, and the certainty of the processing state is calculated and changed every time the rule is hit. The difference between the two is that in Windows there is a possibility of any processing state, whereas in Windows,
Window 2 indicates that there is no possibility of any processing state. Windows 3 displays the inference process. The successive processing states of the rules are displayed, and the confidence level CF3 is
Or calculate it with the hit confidence CF of Window 2, CF will change to (CF + CF3 CFXCF3) J
In window 4, the final result of the inference is displayed. Here, the treatment status is displayed in order of certainty, and furthermore, the treatment water reservation from the diagnosis is displayed.

There are many types of microorganisms that appear in the activated sludge process, including protozoa and metazoa that can be observed under an optical microscope, and the ability to remove organic matter and solid-liquid separation differs greatly depending on the type of microorganism. The fate of these microorganisms is greatly influenced by the condition of activated sludge and the water quality environment.

Some professional researchers and experienced treatment system operators predict current treatment conditions and treatment trends by examining the appearance of these microorganisms. The above example is based on the diagnosis of the wastewater treatment ability by emerging microorganisms, which was carried out by some specialized researchers and skilled operators, and the method of operating the treatment system based on the diagnosis results was registered as knowledge in the knowledge base. This allows even general operators and beginners to reflect the information in operational management. Since this knowledge can be updated, for example, by adding data and knowledge specific to the processing system or accumulating new data and knowledge, the knowledge can be organized objectively and the maintenance and management of the processing system can be made more accurate. You will be able to do this.

It goes without saying that the present invention can be applied in various ways in accordance with its gist and can take various embodiments.

For example, the purpose of the present invention can be fully achieved even if the inference mechanism performs inference 1 diagnosis of the treatment status of the activated sludge process based only on human data on the microorganisms that appear. In addition, the output of countermeasures by the inference mechanism and the collation and display of microbial data on the man-machine interface are preferable as they can improve operability, but may be omitted.

If Effects of the Invention As is clear from the above explanation, the activated sludge process treatment state determination device of the present invention provides the following effects.

(1) Even general operators and beginners can accurately manage the operation of an activated sludge process, predict the state of activated sludge, predict the treatment state, etc. using information on the ecology of microorganisms.

(2) Since it is possible to update knowledge of diagnosis and inference,
By adding knowledge specific to the treatment system and accumulating data and knowledge, a more detailed relationship between microorganisms and treatment capacity will be elucidated, and the knowledge will be organized objectively, allowing for more accurate maintenance and management of the treatment system. It becomes like this.

(3) Knowledge can be expressed quantitatively, enabling more accurate predictions.

(4) p1 It is possible to prevent oversight of operation-specific data due to individual differences, judgments based on bias, and erroneous operations.

[Brief explanation of the drawing]

Fig. 1 is a block diagram of an operation management system for an activated sludge process showing an embodiment of the present invention, Fig. 2 is a block diagram showing an example of the configuration of an exvert nostem, and Fig. 3 shows the processing procedure of an expert system. 4 is a block diagram showing another configuration example of the expert system,
Figure 5 is a flowchart of diagnostic processing, Figure 6 is a diagram showing a manual screen for microbial data, Figure 7 is an explanatory diagram for inputting microbial data, Figure 8 is a diagram showing a manual screen for data such as external appearance data, and Figure 9 is a diagram showing a manual screen for microbial data. The figure shows an output screen for diagnosis results, etc., and FIG. 10 is a configuration diagram of a sewage treatment system showing a conventional activated sludge process. 20.20'...Expert system, 21゜2
1'...Knowledge base, 22...Inference engine, 2
3...CRT display (output device), 24...
Microorganism database, 25...Man-machine interface, 26...Man-powered device. Figure 3

Claims (5)

[Claims] (1) A knowledge base that can register and update information on the appearance of indicator microorganisms, which are indicators of the treatment status of the activated sludge process, among the activated sludge microorganisms observed in the activated sludge process, and the relationship between the treatment status. , a man-machine interface for inputting the type and number of currently appearing microorganisms, and an inference mechanism for inferring and diagnosing the treatment status of the activated sludge process from the type and population distribution of the appearing microorganisms using the knowledge of the knowledge base. An activated sludge process treatment status determination device comprising: (2) In the activated sludge process treatment status determination device according to claim 1, the indicator microorganisms are classified into microbial groups based on the contribution rate, and the inference mechanism determines the activity using the number or appearance stage of the microorganism group and the contribution rate. 1. An activated sludge process processing state determination device, which diagnoses the processing state by inferring the processing state of the sludge process and calculating a degree of certainty. (3) In the activated sludge process processing state determination device according to claim 1 or claim 2, the inference mechanism includes information regarding the appearance of the aeration tank and settling tank of the activated sludge process and/or environmental factors constantly measured in the aeration tank. An activated sludge process processing state determination device characterized in that information is used for inference and diagnosis of the processing state of the activated sludge process. (4) The activated sludge process treatment status determination device according to any one of claims 1 to 3, characterized in that the inference mechanism outputs a countermeasure based on the result of diagnosis of the treatment status of the activated sludge process. Activated sludge process treatment status determination device. (5) The activity according to any one of claims 1 to 4, wherein the man-machine interface collates and displays pre-stored characteristics and shapes of activated sludge microorganisms when inputting data for diagnosis. Sludge process treatment status determination device.