Sewage treatment process flow fault diagnosis expert system and application thereof

Abstract

The invention discloses a sewage treatment process flow fault diagnosis expert system and application thereof. The method for constructing the system comprises the following steps: S1, establishing a reference mathematical model based on an activated sludge mechanism on a simulation platform; s2, introducing an expert system technology, and establishing an overall structure design of the fault diagnosis expert system of the activated sludge process sewage treatment process flow by combining a mathematical model based on a mechanism; s3, establishing a knowledge base for representing knowledge according to the production rule; s4, establishing a database; s5, establishing an internal inference engine; and S6, establishing a human-computer interaction interface. The fault diagnosis expert system constructed by the invention can be applied to intelligent control of sewage treatment, the control level of a sewage treatment system can be improved, the sewage treatment efficiency and the stability of effluent quality can be improved, and the operation cost and resource consumption can be reduced; meanwhile, technical reference and optimization and transformation schemes can be provided for other sewage treatment plants, and the method has important practical value and popularization and application prospects.

Description

Technical Field

[0001] The invention relates to the technical field of sewage treatment expert system design, and in particular to a sewage treatment process fault diagnosis expert system and its application. Background Art

[0002] At present, since it is difficult to find solutions to daily operation problems in sewage treatment plants in books, operators have long managed sewage treatment plants based on years of accumulated experience. However, the accumulation of such experience requires a long period of practical operation experience and extensive knowledge, so only a few people have mastered it. If the knowledge of experts in the field and the long-term experience accumulated by operators are gathered, and fault diagnosis technology is applied to the operation and management of sewage treatment plants, it will have a positive significance to solve the current situation of lack of operation and management personnel in my country (especially in some small town sewage treatment plants).

[0003] With the development of artificial intelligence technology, especially the application of knowledge engineering, expert systems and artificial neural networks in the field of fault diagnosis, the diagnostic methods based on system mathematical models and signal detection and processing are gradually being replaced by diagnostic methods based on artificial intelligence. Fault diagnosis technology has entered the stage of intelligent diagnosis technology with knowledge processing as the core and the integration of modeling processing, signal processing and knowledge processing. In this stage, the knowledge of experts in the field will be fully valued, and the research on diagnostic problems will focus on simulating the reasoning process of experts and the ability to use various diagnostic knowledge. At the knowledge level, it is an inevitable trend for the development of intelligent diagnosis systems to realize the integration of dialectical logic and mathematical logic, the unification of symbolic processing and numerical processing, the unification of reasoning process and algorithmic process, and the interaction between knowledge base and database.

[0004] Therefore, this paper proposes a sewage treatment process fault diagnosis expert system and its application. Summary of the invention

[0005] The purpose of the present invention is to provide a sewage treatment process fault diagnosis expert system which can ensure the stable operation of the sewage treatment plant process engineering, ensure that the effluent water quality is stable and meets the standards, improve the intelligent control level of the sewage treatment system and the work ability and efficiency of the staff, provide a new direction for the optimization and transformation of the sewage treatment plant, and provide new technology for the enrichment and development of the expert system.

[0006] In order to achieve the above technical effects, the present invention is implemented by the following technical solutions: a sewage treatment process fault diagnosis expert system, characterized in that the construction of the system includes the following steps:

[0007] S1. Establish a benchmark mathematical model based on the activated sludge mechanism on the simulation platform;

[0008] S2. Introduce expert system technology and combine it with the mechanism-based mathematical model to establish the overall structural design of the fault diagnosis expert system for the activated sludge wastewater treatment process;

[0009] S3. Establish a knowledge base that uses production rules to represent knowledge. It consists of factual knowledge and heuristic knowledge. It is a collection of several expert knowledge in a field or a specific problem. It is used to store fault diagnosis knowledge of the activated sludge wastewater treatment process, including theoretical knowledge and heuristic knowledge. The diagnostic knowledge is represented by production rules and converted into a form that complies with the CLIPS language specification and encoded and stored in the fuzzy knowledge base. This method is simple, clear, easy to understand, easy to retrieve, and easy to reason. Each rule is independent of each other, which facilitates the establishment, expansion and management of the knowledge base. In addition, database technology is used to store knowledge, which is convenient for users to view and edit.

[0010] S4. Establish a database: collect the original symptom information obtained from the fault phenomenon, the intermediate information generated by the diagnosis process, the optimal parameters obtained by the simulation calculation of the mathematical model based on the mechanism, and the inlet and outlet water quality data collected by the sewage plant, and build a database; so that the database can provide necessary data for the reasoning and interpretation of the expert system during the processing process;

[0011] S5. Establish an internal reasoning engine: Type the internal reasoning engine into Visual Basic to achieve mixed programming of Visual Basic and the internal reasoning engine. The reasoning engine is a set of programs used to control and coordinate the methods and strategies of the entire expert system. It uses the knowledge in the knowledge base according to the user's data input and a certain reasoning strategy to solve the current problem, interpret the user's request, and finally draw a conclusion;

[0012] S6. Establish the human-computer interaction interface: Set up the interface in the Visual Basic 6.0 environment. The human-computer interface is used for information exchange between the expert system and the outside world. It is an important way for users and experts to use and maintain the expert system. The friendly interface makes information input and operation simple, and it is also easy to maintain the system.

[0013] Furthermore, in S1, the modeling principle of the benchmark mathematical model is:

[0014]

[0015] Where, Vr A is the amount of material consumed in the reaction pool per unit time; Qc0 is the amount of material flowing in; Qc e is the amount of material flowing out; Vdc A / dt is the amount of material accumulated in the reaction tank per unit time.

[0016] Furthermore, in S3, the knowledge base that establishes production rules to represent knowledge is acquired through three specific ways:

[0017] S3.1. Use manual acquisition methods, whereby personnel manually consult a large number of books, documents, manuals, etc. on the activated sludge wastewater treatment process to obtain factual knowledge;

[0018] S3.2. Through repeated communication between data collectors, domain experts and sewage treatment plant operators, specific cases of typical failures are analyzed to obtain the inspiring knowledge accumulated by domain experts in long-term actual operation practices. The acquired knowledge is then analyzed, summarized and organized, and domain experts propose modification and improvement suggestions until a relatively accurate and complete set of domain knowledge is formed.

[0019] S3.3. Extract the water quality data collected by the sewage treatment plant and the optimal control parameters calculated by the mathematical model in the sewage treatment plant system, and automatically obtain and save them by downloading them in the system.

[0020] Furthermore, in S5, the internal inference engine uses CLIPS; all the rules of input data in the inference are converted into a form that conforms to the CLIPS language specification.

[0021] Furthermore, in S6, the human-machine interface adopts a modular design, including four modules: symptom acquisition, fault diagnosis, knowledge base management and system help; a modular software design method is adopted to achieve the separation of the knowledge base and the inference engine, which is convenient for expanding the functions of the system.

[0022] Furthermore, the specific processes of the four modules of symptom acquisition, fault diagnosis, knowledge base management and system help are as follows:

[0023] Symptom acquisition module: According to the abnormal conditions in the activated sludge wastewater treatment process, the symptom facts are obtained from both the fault phenomenon and the operating parameters.

[0024] Fault diagnosis module: first import the acquired operating parameter values, perform fuzzy representation, start CLIPS for uncertain reasoning, search for matching rules, and determine the adjustment direction of the parameters; then enter the extreme value range of the operating parameters, set the severity of the fault, use fuzzy reasoning technology, obtain the adjustment value of the parameters, and calculate the credibility of the conclusion; after the diagnosis is completed, the diagnosis results will be displayed to the user or in the form of a report, and can be saved and printed.

[0025] Knowledge base management module: The knowledge base management module is an interface between knowledge engineers or users and the knowledge base. Its main functions include: knowledge storage, knowledge retrieval, knowledge addition, knowledge deletion, knowledge modification, etc. It can also perform necessary grammar checks on the input knowledge to avoid misoperation during the input process.

[0026] System Help: Provide context-related help for questions that users may have during use, so that users can better use the system.

[0027] Furthermore, the knowledge base in the knowledge base management module is established on the basis of Microsoft Access, which gives full play to the characteristics of simple database entry and clear entries; at the same time, Visual Basic provides a powerful data access object tool that can easily link database files, access the data therein, and operate and manage the database; therefore, when users manage the knowledge base, they do not need to directly access the database, but can directly edit the database through the editing windows of each functional module of the knowledge base, thus overcoming the trouble caused by frequent switching between the database and the knowledge base; when users edit the knowledge base, they only modify the clipboard file, which has no effect on the CLIPS inference engine, thus realizing the separation of the knowledge base and the inference engine and ensuring the transparency of the expert system.

[0028] Another object of the present invention is to provide an application of a sewage treatment process fault diagnosis expert system, characterized in that the sewage treatment process fault diagnosis expert system is applied to intelligent control of sewage treatment.

[0029] The beneficial effects of the present invention are:

[0030] The present invention can simulate the reaction of each stage of the sewage treatment plant, simulate and predict the effluent water quality, and provide control parameters for the operation of the sewage treatment plant by establishing a mathematical model; by establishing an expert system, it can integrate the domain knowledge of sewage treatment plant experts, the experience knowledge of professional and technical personnel, and the mathematical model calculation data analysis knowledge, analyze the causes of failures in the sewage treatment plant, and propose solutions;

[0031] The present invention forms a diagnostic expert system by combining an expert system with a mathematical model, so that the system has predictive and real-time properties; the expert system can analyze data of abnormal water quality for the mathematical model and feed back the abnormal data to the mathematical model after processing; the mathematical model provides data support for the expert system and enriches the knowledge base of the expert system; the combination of the two can solve abnormal operation of the water plant in advance, make the sewage plant operate stably, and accurately improve the effluent water quality of the sewage plant; the expert system knowledge base and the inference engine are transparent; if the staff often checks the expert system, the staff's work ability can be improved; the easy operation performance of the human-computer interaction interface improves the staff's work efficiency, reduces operating costs and resource consumption, improves sewage treatment efficiency and the stability of effluent water quality; provides a safe and secure working environment for the staff; at the same time, the invention can also provide technical references and optimization and transformation plans for other sewage treatment plants, and has important practical value and promotion and application prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for describing the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0033] Figure 1 It is a schematic diagram of the overall construction process of the sewage treatment process fault diagnosis expert system of the present invention;

[0034] Figure 2 A diagram for constructing a mathematical model control system for a sewage treatment plant according to the present invention;

[0035] Figure 3 It is the principle diagram for realizing the expert system of the present invention;

[0036] Figure 4 This is a diagram showing the interactive interface of the expert system of the present invention. DETAILED DESCRIPTION

[0037] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0038] Example 1

[0039] Constructing a sewage treatment process fault diagnosis system includes the following steps:

[0040] (1) Establish a mechanism-based mathematical model;

[0041] (2) Collect water quality data from the sewage treatment plant and input the data into a mathematical model for calculation to obtain initial control parameter values;

[0042] (3) Randomize the initial control parameter values ​​to generate a sample set;

[0043] (4) Input the sample set into the virtual simulation platform for simulation calculation and comparison to obtain the control parameters with low effluent quality and low energy consumption;

[0044] (5) Input the control parameters into the expert system database.

[0045] The benchmark mathematical model based on the activated sludge mechanism (ASM1) in this embodiment includes 13 components: inert dissolved organic matter SI, easily degradable dissolved substrates SS, inert non-soluble substrates XI, slowly degradable non-soluble substrates XS, heterotrophic microorganisms XBH, autotrophic microorganisms XBA, microbial metabolic residues XP, dissolved oxygen SO, nitrate nitrogen SNO, ammonia nitrogen SNH, soluble organic nitrogen SND, insoluble organic nitrogen XND, and alkalinity SALK.

[0046] The benchmark mathematical model based on the activated sludge mechanism (ASM1) in this embodiment includes the construction of a virtual platform, which includes 8 basic reaction processes:

[0047] ① Aerobic growth of heterotrophic bacteria;

[0048] ②, Anoxic growth of heterotrophic bacteria;

[0049] ③. Aerobic growth of autotrophic bacteria;

[0050] ④. Decay of heterotrophic bacteria;

[0051] ⑤. Decay of autotrophic bacteria;

[0052] ⑥, ammoniating process;

[0053] ⑦. Slowly degrade the hydrolysis of insoluble organic matter;

[0054] ⑧. Hydrolysis of biodegradable organic nitrogen.

[0055] Assuming that the organic nitrogen is uniformly distributed in the slowly degradable organic matter, the hydrolysis rate of the insoluble organic nitrogen is proportional to the hydrolysis rate of the slowly degradable organic matter.

[0056] It also includes 19 parameters: heterotrophic bacteria yield coefficient YH, autotrophic bacteria yield coefficient YA, nitrogen content ratio in microbial cells iXB, inert particle ratio in microorganisms fP, nitrogen content ratio in microbial products iXP, oxygen half-saturation coefficient of heterotrophic bacteria KO,H, oxygen half-saturation coefficient of autotrophic bacteria KO,A, nitrate nitrogen half-saturation coefficient of heterotrophic bacteria KNO, ammonia half-saturation coefficient of autotrophic bacteria KNH, heterotrophic bacteria half-saturation coefficient KS, half-saturation coefficient of hydrolysis of slowly biodegradable substrates KX, maximum specific growth rate of autotrophic bacteria μ A , maximum specific growth rate of heterotrophic bacteria μ H , attenuation coefficient bA of autotrophic bacteria, attenuation coefficient bH of heterotrophic bacteria, correction factor ηg for heterotrophic growth under anaerobic conditions, correction factor ηh for hydrolysis under anaerobic conditions, maximum specific ammoniation rate ka, maximum specific hydrolysis rate kh.

[0057] The benchmark mathematical model based on the activated sludge mechanism (ASM1) in this embodiment includes sensitivity analysis. The present invention uses conventional sensitivity analysis (Si,j ) to calibrate the model parameters.

[0058]

[0059] In the formula, S i,j is the sensitivity coefficient, y i is the effluent water quality index, x j Set parameters for the system.

[0060] When the sensitivity coefficient S i,j <0.25, indicating that the parameter has no significant effect on the model output results; when 0.25≤S i,j <1, indicating that the parameter has an impact on the model output results; 1≤S i,j <2, indicating that the parameter has a greater impact on the model output results; S i,j ≥2, indicating that the parameter has a significant impact on the model output results. All chemometric parameters and kinetic parameters in the model were adjusted by 10% and the sensitivity of each parameter to the output water index was calculated.

[0061] Through sensitivity analysis, the model was calibrated under steady-state simulation. Combined with relevant domestic and foreign literature and cases, the chemometric parameters of the model were not adjusted. After calibration, the simulated effluent quality was basically consistent with the actual effluent quality, with a small error. The model was dynamically analyzed, and the various simulated effluent indicators were similar to the measured effluent indicators, and the change trends were basically consistent. The software simulated the steady-state of the sewage plant, and the relative errors of the simulated values ​​and the measured values ​​were less than 5%; the dynamic simulation showed that the relative errors of the simulated values ​​and the measured values ​​were less than 10%. The simulated effluent and the actual effluent indicators fit well.

[0062] The specific implementation of the present invention is to mainly use Simulink simulation and supplemented by C language writing. After the C language writing is completed, it is converted into an S function used inside the MATLAB software to achieve the purpose of the virtual sewage treatment plant.

[0063] Example 2

[0064] Constructing a sewage treatment process fault diagnosis system includes the following steps:

[0065] (1) Introducing expert system technology and combining it with a mechanism-based mathematical model, the overall structural design of the fault diagnosis expert system for the activated sludge wastewater treatment process is established. This system adopts a modular software design method, separating the knowledge base from the inference engine to facilitate the expansion of the system's functions.

[0066] (2) Establish a knowledge base that uses production rules to represent knowledge.

[0067] (3) Establish a database to store the original symptom information obtained from the fault phenomenon, the intermediate information generated by the diagnostic process, and the optimal parameters obtained by simulation calculation based on the mechanism-based mathematical model, so as to provide necessary data for the reasoning and interpretation of the expert system.

[0068] (4) Establish an internal reasoning engine (CLIPS) and type the internal reasoning engine into Visual Basic to achieve mixed programming of Visual Basic and CLIPS.

[0069] (5) Establish a human-computer interaction interface. Set up the interface in the Visual Basic 6.0 environment. The human-computer interface is used for information exchange between the expert system and the outside world.

[0070] In this embodiment, the construction of the expert system platform includes the overall structural design of the expert system. The structure of the expert system is equivalent to a container, which contains the components of the expert system and determines the mutual connection and working mode of each functional module within the expert system.

[0071] The construction of the expert system platform in this embodiment includes the establishment of the expert system knowledge base. The establishment of the knowledge base mainly includes two parts: knowledge acquisition and knowledge representation. The knowledge base is the core of the expert system. It is composed of factual knowledge and heuristic knowledge. It is a collection of several expert knowledge in a field or a specific problem. Acquisition studies how to extract and summarize the knowledge (including concepts, facts, relationships and methods) used to solve problems in a specific field from the minds of experts or external knowledge sources (documents, books, materials), and convert it into a specific knowledge representation form and compile it into the knowledge base of the expert system. The present invention uses three knowledge acquisition methods: non-automatic, automatic and neural network-based.

[0072] The knowledge acquisition of the present invention mainly includes two parts: fault information and operation parameters of the activated sludge wastewater treatment process.

[0073] (I) Fault analysis of activated sludge wastewater treatment process

[0074] The activated sludge process is not easy to adapt to sudden changes in water quality, water volume and environment. Therefore, many abnormal situations may occur during operation, resulting in deterioration of effluent quality, sludge loss and reduced treatment effect. The main abnormal phenomena that may occur during operation are as follows:

[0075] 1) Sludge bulking

[0076] a. Fault phenomenon: The sludge volume expands, the structure is loose, it is not easy to settle, and it often floats up and flows out with the effluent; the clarified liquid in the secondary sedimentation tank is scarce and the color also changes.

[0077] b. Cause of failure:

[0078] ① Lack of nutrients; ② Insufficient dissolved oxygen; ③ Too low sludge load; ④ Too low pH value; ⑤ Too high water temperature; ⑥ Too long sludge age; ⑦ Corrupted sewage.

[0079] c. Solution:

[0080] ①. Adding chemicals: Adding coagulants such as iron salts and aluminum salts to improve the compactness of sludge and increase the specific gravity of sludge; adding inert substances such as asbestos powder, diatomaceous earth, clay, etc. to reduce the

[0081] Index; add bleaching powder or liquid chlorine (0.3% to 0.6% of dry sludge) to inhibit the reproduction of filamentous bacteria. This method can only be used as a temporary emergency measure, because long-term addition of agents not only increases operating costs, but also causes changes in the growth environment of microorganisms, resulting in reduced treatment effects.

[0082] ②. Improve the growth environment of microorganisms: Add nitrifying sludge or N, P and other components to supplement the lack of nutrients and maintain BOD:N:P at around 100:5:1; increase the aeration volume or reduce the water intake to reduce the oxygen demand and ensure that the dissolved oxygen concentration in the aeration tank is not less than 2 mg / L; add lime, alkaline surface, yellow mud, etc. to adjust the pH value to not less than 6.0; set a high-load contact area (i.e. selector) in the front of the aeration tank to control low-load expansion; add fillers or strengthen aeration in the aeration tank to control high-load expansion; take four measures: pre-aeration (release odor), chemical oxidation (addition of chlorine, hydrogen peroxide or potassium permanganate), and chemical precipitation (addition of ferric chloride) to prevent wastewater corruption.

[0083] 2) Sludge disintegration

[0084] a. Fault symptoms: The treated water is turbid, the sludge flocs are miniaturized, and the treatment effect is poor.

[0085] b. Cause of failure:

[0086] ①. Problems in operation, such as excessive aeration that destroys the balance of biological nutrition in activated sludge, reduces the number of microorganisms and causes them to lose activity, reduces adsorption capacity, and shrinks flocs. Some of them become feather-like sludge that is not easy to settle, resulting in turbidity in treated water. ②. The presence of toxic substances in sewage inhibits or harms microorganisms, reduces or completely stops the purification function, and causes the sludge to lose activity.

[0087] c. Solution:

[0088] ①. For problems in operation, make adjustments by checking the sewage volume, return sludge volume, air volume, sludge discharge status, SV, MLSS, DO, and N; ②. For toxic substances in sewage, it is necessary to find out the source and take corresponding countermeasures, such as stopping water intake, increasing dilution water, reducing toxic concentration, reducing sludge discharge, and increasing sludge return ratio.

[0089] 3) Sludge corruption

[0090] a. Fault phenomenon: The sludge that has been trapped in the dead corner of the secondary sedimentation tank for a long time becomes corrupt and black, produces a foul odor, and floats up in large pieces.

[0091] b. Cause of failure:

[0092] ①. There are a large number of dead zones or the sludge discharge channel is blocked in the secondary sedimentation tank; ②. The aeration time is too long.

[0093] c. Solution:

[0094] ①. Install scum removal equipment to prevent sludge from overflowing; ②. Eliminate dead corners in sedimentation tanks; ③. Increase the slope of the pool bottom or improve the pool bottom scraping equipment to prevent sludge from remaining at the bottom of the pool; ④. Shorten the aeration time.

[0095] 4) Sludge floating

[0096] a. Fault phenomenon: The sludge in the secondary sedimentation tank floats up in blocks.

[0097] b. Cause of failure:

[0098] ①. Influent water quality: A large amount of fat and oil flows into the sewage; too high or too low pH value will affect the absorption of nutrients by microorganisms; too high alkalinity will cause the microbial cells to dehydrate and die or burst; too high temperature will cause most of the microorganisms in the activated sludge to die; toxic substrates will cause the microorganisms in the activated sludge to be impacted and destroyed.

[0099] ②. Process operation: Excessive aeration in the aeration tank causes the sludge to be stirred too violently, generating a large number of small bubbles that are adsorbed on the floccules; the sludge age in the aeration tank is too long, the nitrification process is high, and denitrification occurs at the bottom of the secondary sedimentation tank; the sludge return volume is too large, resulting in incomplete gas-water separation.

[0100] ③. Excessive reproduction of filamentous bacteria in activated sludge: effects of surfactants, lipid compounds and mechanical stress; excessive addition of filamentous bacteria inhibitors.

[0101] c. Solution:

[0102] ①. Control the air supply within the limit required for stirring; ②. Increase the sludge return volume or remove the excess sludge in time; ③. Reduce the sludge concentration of the mixed liquor, shorten the sludge age and reduce the dissolved oxygen so that it does not proceed to the nitrification stage; ④. Set up a pH automatic adjustment system consisting of a pH detector, pH automatic adjustment valve, etc. at the inlet of the aeration tank to control the pH value of the aeration tank inlet water within the required range.

[0103] 5) Bubble problem

[0104] a. Fault phenomenon: Foam is generated in the aeration tank.

[0105] b. Cause of failure:

[0106] ①. There are a large number of synthetic detergents or other foaming substances in the sewage; ②. When the temperature is around 20°C, the number of Nocardia (microorganisms that often appear when foam problems occur) increases; ③. When the sludge load is high, Nocardia has an absolute advantage among filamentous bacteria; ④. Excessively high substrate concentration causes Nocardia to multiply in large numbers.

[0107] c. Solution:

[0108] ①. Inject water in stages; ②. Spray the surface of the aeration tank with tap water or treated water; ③. Add defoaming agent (such as engine oil, kerosene, etc., with an amount of about 0.5-1.5 mg / L); ④. Use a fan to mechanically defoam.

[0109] The operating parameters of the activated sludge wastewater treatment process are obtained through direct online detection by sensors and laboratory analysis. They are the direct basis for determining the cause and specific location of failures in the activated sludge treatment process and for providing corresponding solutions.

[0110] (II) Evaluation indicators of activated sludge sewage treatment system mainly include:

[0111] 1) Organic pollution index

[0112] a. Biochemical Oxygen Demand (BOD)

[0113] The amount of dissolved oxygen consumed by microorganisms to oxidize organic matter into inorganic matter under the condition of water temperature of 20℃ is called biochemical oxygen demand.

[0114] b. Chemical Oxygen Demand (COD)

[0115] The amount of oxygen consumed by strong oxidants to oxidize organic matter into CO2 and H2O under acidic conditions is called chemical oxygen demand.

[0116] 2) Indicators of microbial physiological activities

[0117] a. Nutrients (BOD:N:P)

[0118] Microorganisms involved in activated sludge treatment need to continuously absorb nutrients necessary for their life activities from the surrounding sewage, including carbon sources, nitrogen sources, inorganic salts and certain growth hormones. Generally speaking, the BOD:N:P ratio should be maintained at around 100:5:1.

[0119] b. Dissolved Oxygen (DO)

[0120] The microbial community involved in sewage activated sludge treatment is mainly composed of aerobic bacteria. Insufficient or excessive dissolved oxygen will have an adverse effect on the physiological activities of microorganisms. The dissolved oxygen concentration in the aeration tank should generally be maintained at no less than 2 mg / L.

[0121] c. pH value

[0122] The physiological activities of microorganisms are closely related to the pH value of the environment. Only under suitable pH conditions can microorganisms carry out normal biological activities. In the aeration tank, the optimal pH value range is between 6.5 and 8.5.

[0123] d. Water temperature

[0124] Appropriate temperature can promote and strengthen the physiological activities of microorganisms; inappropriate temperature will weaken or even destroy the physiological activities of microorganisms, leading to changes in the morphology and physiological characteristics of microorganisms, and may even cause the death of microorganisms. Generally, the temperature of activated sludge treatment is controlled between 15℃ and 35℃.

[0125] e. Toxic substances

[0126] Toxic substances refer to certain inorganic substances (such as heavy metal ions) and organic substances (such as phenol, cyanide, etc.) that have an inhibitory effect on the physiological activities of microorganisms.

[0127] 3) Activated sludge microbial biomass index

[0128] a. Mixed Liquor Suspended Solids (MLSS)

[0129] The mixed liquor suspended solids concentration is also called the mixed liquor sludge concentration, which indicates the total weight of activated sludge solids contained in the mixed liquor per unit volume of the aeration tank. Generally speaking, MLSS < 3.5g / L is prone to sludge bulking.

[0130] b. Mixed Liquor Volatile Suspended Solids (MLVSS) Mixed Liquor Volatile Suspended Solids (MLVSS) is the concentration of organic solids in mixed liquor activated sludge. Compared with MLSS, this indicator is more accurate.

[0131] 4) Activated sludge settling performance indicators

[0132] a. Sludge settling velocity (SV)

[0133] The sludge settling ratio refers to the percentage of the volume of the precipitated sludge formed after the mixed liquid is left in the measuring cylinder for 30 minutes to the original mixed liquid volume. By testing the SV, abnormal phenomena such as sludge expansion can be discovered in time.

[0134] b. Sludge Volume Index (SVI)

[0135] The sludge volume index refers to the volume of the precipitated sludge formed by each gram of dry sludge in the mixed liquid at the outlet of the aeration tank after 30 minutes of static settling. SVI can be used to determine whether sludge bulking has occurred or will occur, which has important guiding significance for operation control. Generally speaking, if SVI>150mLg, sludge bulking is likely to occur.

[0136] c. Sludge Age

[0137] Sludge age is also known as the average residence time of biosolids, that is, the average residence time of activated sludge in the aeration tank. Increasing sludge age will easily lead to an increase in SVI value, which in turn causes sludge bulking.

[0138] d. BOD-sludge load (Ns)

[0139] BOD-sludge load refers to the amount of organic matter that can be accepted and degraded to a predetermined degree per unit weight (kg) of activated sludge in the aeration tank within a unit time (1d). BOD-sludge load is an important factor affecting the degradation of organic pollutants and the growth of activated sludge. When N is between 0.5 and 1.5 kgBOD / (kg MLSS·d), sludge bulking is likely to occur.

[0140] 5) Microbiological examination

[0141] Microorganisms in activated sludge are examined under a microscope. Protozoa and metazoa (collectively referred to as micro-animals) are relatively larger than bacterial individuals, easy to observe, identify and count under a microscope, and are more sensitive to changes in external environmental conditions. They are used as indicator organisms to diagnose the state and performance of activated sludge.

[0142] 3. Knowledge Representation

[0143] Through the analysis of several commonly used knowledge representation methods and their main advantages and disadvantages, combined with the characteristics of activated sludge wastewater treatment process diagnosis knowledge, the knowledge representation method of this system adopts the production rule representation method; it has the following advantages:

[0144] (1) Simple and intuitive expression. The simple expression of rules “IF...THEN” makes it easy for knowledge engineers to explain the structure of knowledge to domain experts.

[0145] (2) It is easy to reason. The process of reasoning using production rules is to match facts with rule antecedents. If the antecedent of a rule matches the current facts, then the rule is satisfied. Therefore, using production rules to express knowledge makes reasoning more efficient and the algorithm simpler.

[0146] (3) Easy to explain to users. By listing the rules in the reasoning process, users will naturally understand how to convert the initial factual conditions into the final diagnosis results.

[0147] 4. Generative Knowledge Representation

[0148] Fault tree is a natural language expression of diagnostic knowledge, and the knowledge in the knowledge base should be expressed in a form that can be stored by computers. There is a one-to-one correspondence between fault tree and production rule representation in the description of knowledge, so the fault tree can be easily converted into the form of "If-Then" rules.

[0149] In this embodiment, the construction of the expert system platform includes the establishment of an expert system database. The database is used to store facts, data, initial states, various intermediate states of the reasoning process, and goals of related field problems, reflecting the main states and characteristics of the problems to be dealt with by the system. It is the object of system operation, mainly the database of sewage plant inlet and outlet water quality data and control parameters calculated by mathematical models.

[0150] In this implementation, the construction of the expert system platform includes the establishment of an expert system inference engine. The inference engine is a set of programs used to control and coordinate the methods and strategies of the entire expert system. It uses the knowledge in the knowledge base based on the user's input data and follows certain inference strategies to solve the current problem, interpret the user's request, and finally draw conclusions.

[0151] The present invention uses CLIPS as the internal inference engine for reasoning, so all rules must be converted into a form that complies with the CLIPS language specification. The general format of the rules is as follows:

[0152] (defrule <rule-name> [ <comment> ]

[0153] [ <declaration> ]

[0154] <conditional-element> *

[0155] ＝>

[0156] <actions>*)

[0157] Rules must be defined using the defrule function. <rule-name>is the rule name. You can use any legal word in CLIPS, but the rule name must be unique. If the rule name you enter has the same name as an existing rule, the new rule will replace the old one. <comment>Used to describe the purpose of the rule or other information the programmer wants; <declaration>A feature used to describe a rule.

[0158] <conditional-element>It is the left-hand side of the rule, consisting of multiple or zero patterns. Each pattern consists of one or more constraints, and its purpose is to match the fields in the custom template fact.

[0159] For the zero-pattern case, the system automatically adds a special pattern (initiate-fact). <actions>It is the right-hand side of the rule. When the condition pattern is met, the action to be performed by the rule is triggered.

[0160] The entire rule must be enclosed in parentheses, and each pattern and action in the rule must also be enclosed in parentheses.

[0161] According to the CLIPS rule format, part of the code for the sludge bulking rule is as follows:

[0162] (defrule method1; if sludge bulking occurs, the rules for eliminating the fault from the operating parameters are:

[0163] parameter<-(nitrogen low);

[0164] =>

[0165] (retract parameter);

[0166] (print out t"The nitrogen need to be added."crlf));

[0167] (defrule method2 parameter<-(phosphorus low);

[0168] =>

[0169] (retract parameter)

[0170] (printout t"The phosphorus need to be added."crlf));

[0171] (defrule method3 parameter<-(pH low);

[0172] =>

[0173] (printout t"The temperature need to be increased."crlf));

[0174] (defrule method4 parameter<-(temperature high);

[0175] (retract parameter)

[0176] (printout t"The temperature need to be decreased."crlf));

[0177] (defrule method5 parameter<-(DO low);

[0178] =>

[0179] (retract parameter)

[0180] (printout t"Increase aeration value or decrease inlet volume."crlf));

[0181] (defrule method6 parameter<-(Ns high);

[0182] =>

[0183] (retract parameter)

[0184] (printout t"Increase MLSS."crlf));

[0185] In the above rules, the expression of empirical knowledge is mostly imprecise. For example, the description of the operating parameters of N, P, DO, pH, T, and Ns does not use actual measured values, but uses "high" and "low" to express. Therefore, using this rule for reasoning cannot produce accurate results, but can only roughly give the cause of the fault and the adjustment direction of the operating parameters. In order to obtain the optimal value of the adjustment parameter, fuzzy knowledge and fuzzy reasoning technology are needed.

[0186] Since the fault conditions in the activated sludge wastewater treatment process are difficult to grasp, and the "inverted tree" fault tree requires the diagnosis process to traverse the rule base, this system adopts a top-to-bottom, breadth-first forward reasoning control strategy.

[0187] The human-world interface uses CLIPS embedded in Visual Basic to achieve hybrid programming of the two. The basic steps are as follows:

[0188] ①Install and register the CLIPS control CLIPS Active X Control.OCX in Visual Basic, and then insert the CLIPS control into the VB interface. Then VB can use all the functions and properties of the control.

[0189] ②. Add several global variables in the CLIPS source program to record the information during the inference engine's working process. The main purpose is to use the CLIPS Bind command to implement the constraints on the variable address to transfer the variable value.

[0190] ③. Add a command button in VB to start the CLIPS program. The code is as follows:

[0191] Private Sub ClipsStart_Click()

[0192] Dim retval

[0193] retval=Shell(App.Path&"e: / es / CLIPS win.exe"vbHide)

[0194] ClipsStart_Enabled=False

[0195] End Sub

[0196] ④. Complete the input of questions and the output of answers through shared I / O buffer.

[0197] The input and output results of CLIPS are finally directed to stdout and stdin. By using this feature of CLIPS, the input and output of CLIPS can be redirected. Direct the stdout and stdin streams of CLIPS to the buffer files output.dat and input.dat respectively. The code is as follows:

[0198] freopen("input.dat", "w+", stdin): / / Create an input buffer and redirect stdinfreopen("output.dat", "w+", stdout): / / Create an output buffer and redirect stdout;

[0199] ⑤. Use the Router (path buffer) of the CLIPS control to monitor and record the operation of CLIPS, and display the results in the output window using the wdialog variable. The code is as follows:

[0200] Dim MyLJH As CLIPSLJH / / Define output path buffer

[0201] Set MyLJH=New CLIPSLJH

[0202] MyLJH.Name = "wdialog"

[0203] MyLJH.eatCRLF=True / / path buffer initialization

[0204] MyLJH.Movefirst

[0205] Do OutPut.AddItem(MyLJH.Value)

[0206] MyLJH.MoveNext

[0207] Loop Until MyLJH.EOF=True / / Output the result of the path buffer to the VB interface

[0208] ⑥ After the reasoning is completed, use the CLIPS command function (exit) to exit the CLIPS process.

[0209] In this embodiment, the construction of the expert system platform includes the design of the expert system human-computer interface, which translates the input information of experts or users into a form acceptable to the system, and then hands this information to the corresponding module for processing. On the other hand, the information output by the system to experts or users is converted into a representation form that is easy for humans to understand.

[0210] The human-machine interface of the present invention adopts a modular design, and mainly includes four modules: symptom acquisition, fault diagnosis, knowledge base management and system help.

[0211] After completing the design and development of the entire expert system, in order to verify the correctness and effectiveness of the system diagnosis and whether it can achieve the expected results, some fault signs can be set to perform simulation tests on the system.

[0212] ①. Run this system, click the "Omen Acquisition" button in the main interface to enter the symptom acquisition interface and complete the acquisition of symptom information.

[0213] ②. Click the "Fault Diagnosis" button in the main interface to enter the fault diagnosis interface; perform fault diagnosis. The specific steps are as follows:

[0214] 1) Click the "Import Current Data Value" button to import the acquired parameter value; 2) Click the "Fuzzify" button to fuzzify the parameter value; 3) Click the "Import Maximum and Minimum Values" button to enter the extreme values ​​of the parameters that need to be adjusted; 4) Select the type of fault severity and click "Start CLIPS" to enter the reasoning interface; 5) Click the "Start Reasoning" button to perform fuzzy reasoning to obtain the parameter adjustment value and diagnosis result; 6) Click the "Diagnosis Report" button to generate a report form of the diagnosis result. 7) Click "View Reasoning Log" to display the reasoning process of CLIPS; 8) Click "Exit" to complete the diagnosis; At this time, the diagnosis result has achieved the expected goal, ensuring the accuracy and effectiveness of the system diagnosis.

[0215] The specific implementation method of the present invention is to adopt CLIPS and Visual Basic mixed programming, supplemented by the sewage plant data collection system, to achieve the purpose of building a sewage plant expert system.< / actions> < / declaration> < / comment> < / actions> < / conditional-element> < / declaration> < / comment> < / rule-name>

Claims

1. A sewage treatment process fault diagnosis expert system, characterized in that: Building the system involves the following steps: S1. Establish a benchmark mathematical model based on the activated sludge mechanism on the simulation platform; S2. Introduce expert system technology and combine it with the mechanism-based mathematical model to establish the overall structural design of the fault diagnosis expert system for the activated sludge wastewater treatment process; S3. Establish a knowledge base that uses production rules to represent knowledge. It consists of factual knowledge and heuristic knowledge. It is a collection of expert knowledge in a field or a specific problem. It is used to store fault diagnosis knowledge of activated sludge wastewater treatment process, including theoretical knowledge and heuristic knowledge. The diagnostic knowledge is represented by production rules and converted into a form that conforms to the CLIPS language specification and encoded and stored in the fuzzy knowledge base. S4. Establish a database: collect the original symptom information obtained from the fault phenomenon, the intermediate information generated by the diagnosis process, the optimal parameters obtained by the simulation calculation of the mathematical model based on the mechanism, and the inlet and outlet water quality data collected by the sewage plant to build a database; S5. Establish an internal inference engine: Type the internal inference engine into Visual Basic to achieve mixed programming of Visual Basic and the internal inference engine; S6. Establish the human-computer interaction interface: set up the interface in the Visual Basic 6.0 environment.

2. A sewage treatment process fault diagnosis expert system according to claim 1, characterized in that: In S1, the modeling principle of the benchmark mathematical model is: Where, Vr A is the amount of material consumed in the reaction pool per unit time; Qc0 is the amount of material flowing in; Qc e is the amount of material flowing out; Vdc A / dt is the amount of material accumulated in the reaction tank per unit time.

3. A sewage treatment process fault diagnosis expert system according to claim 1, characterized in that: In S3, the knowledge base in which production rules are established to represent knowledge is obtained through three specific ways: S3.

1. Use manual acquisition methods, whereby personnel manually consult a large number of books, documents, manuals, etc. on the activated sludge wastewater treatment process to obtain factual knowledge; S3.

2. Through repeated communication between data collectors, domain experts and sewage treatment plant operators, specific cases of typical failures are analyzed to obtain the inspiring knowledge accumulated by domain experts in long-term actual operation practices. The acquired knowledge is then analyzed, summarized and organized, and domain experts propose modification and improvement suggestions until a relatively accurate and complete set of domain knowledge is formed. S3.

3. Extract the water quality data collected by the sewage treatment plant and the optimal control parameters calculated by the mathematical model in the sewage treatment plant system, and automatically obtain and save them by downloading them in the system.

4. A sewage treatment process fault diagnosis expert system according to claim 1, characterized in that: In S5, the internal inference engine uses CLIPS; all the rules of input data in the inference are converted into a form that conforms to the CLIPS language specification.

5. A sewage treatment process fault diagnosis expert system according to claim 1, characterized in that: In S6, the human-machine interface adopts a modular design method to achieve the separation of the knowledge base and the inference engine, making it convenient to expand the functions of the system.

6. A sewage treatment process fault diagnosis expert system according to claim 5, characterized in that: The specific processes of the four modules of symptom acquisition, fault diagnosis, knowledge base management and system help are as follows: Symptom acquisition module: According to the abnormal conditions in the activated sludge wastewater treatment process, the symptom facts are obtained from both the fault phenomenon and the operating parameters. Fault diagnosis module: first import the acquired operating parameter values, perform fuzzy representation, start CLIPS for uncertain reasoning, search for matching rules, and determine the adjustment direction of the parameters; then enter the extreme value range of the operating parameters, set the severity of the fault, use fuzzy reasoning technology, obtain the adjustment value of the parameters, and calculate the credibility of the conclusion; after the diagnosis is completed, the diagnosis results will be displayed to the user or in the form of a report, and can be saved and printed. Knowledge base management module: The knowledge base management module is an interface between knowledge engineers or users and the knowledge base. Its main functions include: knowledge storage, knowledge retrieval, knowledge addition, knowledge deletion, knowledge modification, etc. It can also perform necessary grammar checks on the input knowledge to avoid misoperation during the input process. System Help: Provide context-related help for questions that users may have during use, so that users can better use the system.

7. A sewage treatment process fault diagnosis expert system according to claim 6, characterized in that: The knowledge base in the knowledge base management module is established on the basis of Microsoft access, which gives full play to the characteristics of simple database entry and clear entries; at the same time, Visual Basic provides a powerful data access object tool that can easily link database files, access the data therein, and operate and manage the database; therefore, when users manage the knowledge base, they do not need to directly access the database, but can directly edit the database through the editing windows of each functional module of the knowledge base, thus overcoming the trouble caused by frequent switching between the database and the knowledge base; when users edit the knowledge base, they only modify the clipboard file, which has no effect on the CLIPS inference engine, thus realizing the separation of the knowledge base and the inference engine and ensuring the transparency of the expert system.

8. An application of a sewage treatment process fault diagnosis expert system, characterized in that: The sewage treatment process fault diagnosis expert system is applied to the intelligent control of sewage treatment.

Images