Method, device and computer-readable storage medium for determining faults of sewage treatment equipment

Abstract

The present application discloses a method, device and computer-readable storage medium for determining faults in sewage treatment equipment. Among them, a method for determining faults in sewage treatment equipment includes: collecting sewage data of the sewage treatment equipment at multiple consecutive moments in a preset period; extracting features from the sewage data at the first moment according to the correlation relationship between the sewage data at the first moment and the sewage data at the second moment in the preset period through a feature extraction model to generate first feature information, where the second moment is the moment before the first moment in the preset period; and classifying the first feature information through a classification model to determine whether the sewage treatment equipment has a fault.

Description

Technical Field

[0001] The present application relates to the technical field of sewage treatment, and in particular, to a method and device for determining sewage treatment equipment failures and a computer-readable storage medium. Background Art

[0002] In recent years, the progress of urban construction in China has accelerated, and the living standards of residents have been greatly improved. However, at the same time, more and more industrial production and domestic wastewater have been generated in cities, leading to increasingly serious problems such as water pollution. In recent years, with the large-scale construction of sewage treatment plants in various regions, this problem has been alleviated to a certain extent. However, due to the large number of physical, biological, and chemical reactions involved in sewage treatment, the process involves a variety of monitoring, detection, and sensor devices, and it is in a harsh environment for a long time, so failures of sewage treatment plants occur frequently. When a sewage treatment plant fails, it will affect its operation efficiency and cannot ensure the safety of the discharged sewage. At the same time, since sewage treatment plants are usually equipped with large-scale equipment such as pumping stations and blowers, once a failure occurs, it may also pose a threat to the personal safety of the staff. Therefore, it is extremely important to conduct a comprehensive fault diagnosis on the processes and key equipment of the sewage treatment process. At present, most of the control and management processes in sewage treatment plants in China still require manual participation, resulting in low efficiency and high labor costs in sewage treatment plants. For the fault detection and diagnosis during the operation of sewage treatment plants, it mainly relies on experienced technical experts to calculate and analyze the historical data collected in the database, find the abnormal values in the historical operation data, make judgments based on experience, and finally give the diagnosis results and specific countermeasures. With the increase in the number and continuous expansion of the scale of sewage treatment plants, relying solely on experts for manual diagnosis, its efficiency and standards are insufficient to meet the requirements of sewage treatment.

[0003] The publication number is CN112241128A, and the name is a control method, system, device, and medium for sewage treatment equipment anomalies. It includes the following steps: obtaining the anomaly information of the sewage treatment equipment, where the anomaly information includes alarm information and / or maintenance information; obtaining the prompt information corresponding to the anomaly information, where the prompt information includes the anomaly handling method corresponding to the anomaly information; calculating the number of current anomaly information, and if the number of anomaly information is greater than 1, displaying all the prompt information in an overlapping and covering manner; determining whether the anomaly information has been processed; if it has been processed, closing the corresponding prompt information; if it has not been processed, when a close request is detected, responding to the close request to close the prompt information and redisplaying the prompt information.

[0004] Publication number: CN118964833A, title: A method for monitoring faults of sewage treatment equipment based on cloud computing. The specific steps include collecting water signals and electrical signals of the water treatment equipment through sensors and transmitting them to the cloud computing platform, storing the water signals and electrical signals in the distributed database of the cloud computing platform and preprocessing the data, processing and analyzing the water signals and electrical signals, constructing a fault detection model using machine learning algorithms, training labels using the historical fault data and regular operation data of the water treatment equipment, evaluating the performance of the fault detection model through cross-validation, using the monitoring service of the cloud computing platform to monitor the operation status of the equipment in real time, predicting the operation data of the water treatment equipment according to the fault detection model, and triggering an alarm mechanism when a fault anomaly of the water treatment equipment is predicted.

[0005] In view of the technical problems in the above-mentioned prior art, such as low diagnostic efficiency and high cost caused by manual fault diagnosis of equipment in sewage treatment plants, no effective solution has been proposed yet. Summary of the Invention

[0006] Embodiments of the present application provide a method, device and computer-readable storage medium for determining faults of sewage treatment equipment, so as to at least solve the technical problems of low diagnostic efficiency and high cost caused by manual fault diagnosis of equipment in sewage treatment plants in the prior art.

[0007] According to one aspect of the embodiments of the present application, a method for determining faults of sewage treatment equipment is provided, including: collecting sewage data of the sewage treatment equipment at multiple consecutive moments in a preset period; extracting features from the sewage data at the first moment according to the correlation relationship between the sewage data at the first moment and the sewage data at the second moment in the preset period through a feature extraction model to generate first feature information, where the second moment is the moment before the first moment in the preset period; and classifying the first feature information through a classification model to determine whether the sewage treatment equipment has a fault.

[0008] According to another aspect of the embodiments of the present application, a device for determining faults of sewage treatment equipment is further provided, including: a data collection module, configured to collect sewage data of the sewage treatment equipment at multiple consecutive moments in a preset period; an information generation module, configured to extract features from the sewage data at the first moment according to the correlation relationship between the sewage data at the first moment and the sewage data at the second moment in the preset period through a feature extraction model to generate first feature information, where the second moment is the moment before the first moment in the preset period; and a fault determination module, configured to classify the first feature information through a classification model to determine whether the sewage treatment equipment has a fault.

[0009] According to another aspect of the embodiments of the present application, there is also provided a device for determining faults in a sewage treatment device, including: a processor; and a memory connected to the processor for providing instructions for the processor to process the following steps: collecting sewage data of the sewage treatment device at multiple consecutive moments in a preset period; extracting features from the sewage data at the first moment according to the correlation between the sewage data at the first moment and the sewage data at the second moment in the preset period through a feature extraction model to generate first feature information, where the second moment is the moment before the first moment in the preset period; and classifying the first feature information through a classification model to determine whether the sewage treatment device has a fault.

[0010] According to another aspect of the embodiments of the present application, there is also provided a computer system, including a memory, a processor, and a computer program stored on the memory, and the processor executes the computer program to implement the steps of the above method.

[0011] According to another aspect of the embodiments of the present application, there is also provided a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the above method are implemented.

[0012] According to another aspect of the embodiments of the present application, there is also provided a computer program product, including a computer program, and when the computer program is executed by a processor, the steps of the above method are implemented.

[0013] In the embodiments of the present application, the diagnostic system collects sewage data at different moments corresponding to each sewage treatment system, realizing comprehensive monitoring of the dynamic changes in the sewage treatment process. And this technical solution uses a feature extraction model to analyze the correlation of sewage data between different moments, so as to extract feature information according to the correlation with the sewage data of the previous moment, thereby paying attention to the influence between the sewage data of different moments and tracing the source of the fault. And this technical solution combines a classification model to efficiently classify the extracted first feature information, improving the speed of fault diagnosis and enhancing the reliability of the diagnosis result. Thus, this technical solution reduces the dependence on manual experience and subjective judgment through model processing, reduces the misjudgment rate caused by human factors, and also reduces the work burden of technicians. And this technical solution can timely warn of potential faults through automated and intelligent fault diagnosis, and take measures in advance to avoid the occurrence of major accidents. Furthermore, it solves the technical problems of low diagnosis efficiency and high cost caused by using manual methods to diagnose faults in the equipment of sewage treatment plants in the prior art. Description of the Drawings

[0014] The accompanying drawings described herein are used to provide a further understanding of the present application and form a part of the present application. The schematic embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation of the present application. In the drawings:

[0015] Figure 1 is a hardware structure block diagram of a computing device for implementing the method according to Embodiment 1 of the present application;

[0016] Figure 2 is a schematic diagram of a fault diagnosis system of a sewage treatment device according to Embodiment 1 of the present application;

[0017] Figure 3 is a schematic flow chart of a method for determining a fault of a sewage treatment device according to Embodiment 1 of the present application;

[0018] Figure 4 is a schematic diagram of a feature extraction model according to Embodiment 1 of the present application;

[0019] Figure 5 is a schematic diagram of a classification model according to Embodiment 1 of the present application;

[0020] Figure 6 is a schematic sequential flow chart of a method for determining a fault of a sewage treatment device according to Embodiment 1 of the present application;

[0021] Figure 7 is a schematic diagram of a device for determining a fault of a sewage treatment device according to Embodiment 2 of the present application; and

[0022] Figure 8 is a schematic diagram of a device for determining a fault of a sewage treatment device according to Embodiment 3 of the present application. Detailed implementation manners

[0023] In order to enable those skilled in the art to better understand the technical solutions of the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.

[0024] It should be noted that the terms "first", "second", etc. in the description, claims and the above-mentioned drawings of this application are used to distinguish similar objects, and do not necessarily have to be used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged under appropriate circumstances, so that the embodiments of this application described here can be implemented in an order different from those illustrated or described here. In addition, the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device comprising a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0025] Embodiment 1

[0026] According to this embodiment, a method embodiment of a method for determining a failure of a sewage treatment device is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. And although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that here.

[0027] The method embodiment provided in this embodiment can be executed on a mobile terminal, a computer terminal, a server or a similar computing device. Figure 1 A hardware structure block diagram of a computing device for implementing a method for determining a failure of a sewage treatment device is shown. As Figure 1 shown, the computing device may include one or more processors (the processor may include, but is not limited to, a processing device such as a microprocessor MCU or a programmable logic device FPGA), a memory for storing data, and a transmission device for communication functions. In addition, it may further include: a display, an input / output interface (I / O interface), a universal serial bus (USB) port (which may be included as one of the ports of the I / O interface), a network interface, a power supply and / or a camera. Those of ordinary skill in the art can understand that Figure 1 the structure shown is only schematic and does not limit the structure of the above-mentioned electronic device. For example, the computing device may further include more or fewer components than Figure 1 shown, or have a different configuration from Figure 1 shown.

[0028] It should be noted that the above one or more processors and / or other data processing circuits can generally be referred to as "data processing circuits" herein. The data processing circuit can be embodied in software, hardware, firmware or any combination thereof, in whole or in part. In addition, the data processing circuit can be a single independent processing module, or be incorporated in whole or in part into any one of other elements in the computing device. As involved in the embodiments of the present application, the data processing circuit is a kind of processor control (such as the selection of a variable resistance terminal path connected to an interface).

[0029] The memory can be used to store software programs and modules of application software, such as the program instructions / data storage device corresponding to the sewage treatment equipment fault determination method in the embodiments of the present application. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory, that is, realizes the sewage treatment equipment fault determination method of the above application program. The memory can include high-speed random access memory, and can also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memories. In some instances, the memory can further include a memory remotely set relative to the processor, and these remote memories can be connected to the computing device through a network. Examples of the above network include but are not limited to the Internet, enterprise intranet, local area network, mobile communication network and their combinations.

[0030] The transmission device is used to receive or send data via a network. Specific examples of the above network can include a wireless network provided by a communication provider of the computing device. In one instance, the transmission device includes a network adapter (Network Interface Controller, NIC), which can be connected to other network devices through a base station and thus can communicate with the Internet. In one instance, the transmission device can be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0031] The display can be, for example, a touch-screen liquid crystal display (LCD), which enables a user to interact with the user interface of the computing device.

[0032] It should be noted here that in some alternative embodiments, the above Figure 1 shown computing device can include hardware elements (including circuits), software elements (including computer code stored on a computer-readable medium), or a combination of both hardware elements and software elements. It should be pointed out that Figure 1 is only an example of a specific specific instance, and is intended to show the types of components that can exist in the above computing device.

[0033] Figure 2It is a schematic diagram of the fault diagnosis system for the sewage treatment equipment according to the present embodiment. Refer to Figure 2 As shown, the system includes: a sewage treatment plant 100 and a diagnosis system 200.

[0034] Among them, the diagnosis system 200 is used to diagnose whether each sewage treatment equipment in the sewage treatment plant 100 fails.

[0035] And the sewage treatment equipment in the sewage treatment plant 100 includes equipment 101 to 10m. The sewage treatment equipment 101 to 10m can be, for example, grille cleaners, grit chambers, anaerobic reactors, aerobic reactors, secondary sedimentation tanks, and sludge dewatering machines, etc.

[0036] Sewage treatment is a complex process, usually including multiple steps and stages, and each stage is performed by specialized equipment to perform specific tasks. The following is the general working sequence of grille cleaners, grit chambers, anaerobic reactors, aerobic reactors, secondary sedimentation tanks, and sludge dewatering machines in treating sewage: After the sewage enters the sewage treatment plant, it first passes through the grille cleaner. The main purpose of the grille cleaner is to filter out large particulate matter in the sewage, such as plastic bags, branches, paper, etc., to prevent these substances from clogging subsequent treatment equipment. The preliminarily filtered sewage then flows into the grit chamber. Here, heavier inorganic particles, such as sand and soil, settle to the bottom of the tank due to the action of gravity and are removed. Next, the sewage enters the anaerobic reactor. At this stage, in an oxygen-deficient environment, through the action of anaerobic microorganisms, complex organic substances in the sewage are decomposed into simple compounds, such as methane and carbon dioxide. Then, the sewage enters the aerobic reactor. Sufficient oxygen is provided here, enabling aerobic microorganisms to further decompose the remaining organic substances and convert them into harmless substances, such as carbon dioxide and water. The sewage after biological treatment flows into the secondary sedimentation tank, where, through the way of gravity sedimentation, the biological flocs (i.e., activated sludge) generated during the treatment process are separated from the water. Finally, the sludge separated from the secondary sedimentation tank will be sent to the sludge dewatering machine for dewatering treatment to reduce the water content of the sludge, facilitating subsequent disposal or utilization.

[0037] And sensors for providing sewage data are installed on each of the sewage treatment equipment 101 to 10m. For example, a level sensor is set on the grille cleaner to monitor the inlet water level and ensure the normal operation of the equipment.

[0038] A suspended solids sensor is set on the grit chamber to monitor the suspended solids concentration in the water and evaluate the grit removal effect.

[0039] A pH sensor is set on the anaerobic reactor to monitor the acidity and alkalinity in the reactor and ensure the optimal survival conditions for anaerobic microorganisms.

[0040] The aerobic reactor is equipped with a dissolved oxygen sensor to monitor the dissolved oxygen concentration and ensure that aerobic microorganisms have sufficient oxygen for organic matter degradation.

[0041] The secondary sedimentation tank is equipped with a suspended solids sensor to monitor the suspended solids concentration in the water and evaluate the sedimentation effect.

[0042] The sludge dewatering machine is equipped with a pressure sensor to monitor the pressure during the dewatering process and ensure the safe operation of the equipment.

[0043] Under the above operating environment, according to the first aspect of this embodiment, a method for determining faults in sewage treatment equipment is provided. This method is implemented by Figure 2 the diagnostic system 200 shown in Figure 3 FIG. shows a schematic flow diagram of this method. Refer to Figure 3 shown, this method includes:

[0044] S302: Collect sewage data of the sewage treatment equipment at multiple consecutive moments in a preset period;

[0045] S304: Through the feature extraction model, according to the correlation relationship between the sewage data at the first moment and the sewage data at the second moment in the preset period, extract features from the sewage data at the first moment to generate first feature information, where the second moment is the moment before the first moment in the preset period; and

[0046] S306: Classify the first feature information through the classification model to determine whether the sewage treatment equipment has a fault.

[0047] Specifically, the diagnostic system 200 determines a preset period. For example, this preset period has a total of n consecutive moments. Thus, the diagnostic system 200 collects the sewage data A1 to A n at n moments in this preset period through the sensors set on each sewage treatment equipment 101~10m. i where i = 1~n, and the sewage data A i at the i-th moment:

[0048] A1 = [a 1,1 , a 2,1 ,..., a m,1 T ;

[0049] A2 = [a 1,2 , a 2,2 ,..., a m,2 T ; ...

[0050] A n = [a 1,n , a 2,n,..., a m,n T 。

[0051] where a j,i represents the sewage data of the j-th sewage treatment device at the i-th moment, where j = 1~m.

[0052] Furthermore, the diagnostic system 200 is pre-set with a feature extraction model. The feature extraction model can be an attention model for extracting features from the sewage data. And the feature extraction model is pre-trained by the diagnostic system 200 according to the sample sewage data through the gradient descent method. Thus, the diagnostic system 200 determines the sewage data at the first moment in the preset period and the sewage data at the second moment corresponding to each first moment. The second moment is the moment before the first moment in the preset period. The first moment is from moment 1 to m, and the second moment corresponding to the i-th first moment is from moment 1 to moment i. Furthermore, the sewage data at the i-th first moment is A i , and the sewage data at the second moment corresponding to the i-th first moment (i.e., from moment 1 to moment i) is A1~A i 。

[0053] For example, the sewage data at the 1st first moment is A1, and the sewage data at the second moment corresponding to the 1st first moment (i.e., moment 1) is A1;

[0054] The sewage data at the 2nd first moment is A2, and the sewage data at the second moment corresponding to the 2nd first moment (i.e., from moment 1 to moment 2) is A1~A2;

[0055] ……

[0056] The sewage data at the n-th first moment is A n , and the sewage data at the second moment corresponding to the n-th first moment (i.e., from moment 1 to moment n) is A1~A n 。

[0057] During the sewage treatment process, the sewage data at the later moment is affected by the sewage data at the previous moment. Thus, the diagnostic system 200 uses the feature extraction model to extract features from the sewage data at the first moment (i.e., the sewage data A g , g = 1~i) at the second moment according to the correlation relationship between the sewage data at the previous moment (i.e., the sewage data A i ) at the second moment and the sewage data at the later moment (i.e., the sewage data A i ) at the first moment, and generates the first feature information.

[0058] ​Further, the diagnosis system 200 is preset with a classification model, so that the diagnosis system 200 classifies the first feature information through the classification model to respectively determine whether the sewage treatment devices 101 to 10m are faulty.

[0059] As described in the background art, in recent years, the progress of urban construction in China has accelerated, and the living standards of residents have been greatly improved. However, at the same time, more and more industrial production and domestic wastewater have been generated in cities, leading to increasingly serious problems such as water pollution. In recent years, with the large-scale construction of sewage treatment plants in various regions, this problem has been alleviated to a certain extent. However, due to the large number of physical, biological, and chemical reactions involved in sewage treatment, the process involves a variety of monitoring, detection, and sensor devices, and is in a harsh environment for a long time, so faults in sewage treatment plants occur frequently. When a sewage treatment plant fails, its operating efficiency will be affected, and it cannot ensure the safety of the discharged sewage. At the same time, since sewage treatment plants are usually equipped with large-scale equipment such as pumping stations and blowers, a failure may also pose a threat to the personal safety of the staff. Therefore, it is particularly important to conduct a comprehensive fault diagnosis on the processes and key equipment of the sewage treatment process. At present, most of the control and management processes in sewage treatment plants in China still require manual participation, resulting in low efficiency and high labor costs in sewage treatment plants. For the fault detection and diagnosis during the operation of sewage treatment plants, it mainly relies on experienced technical experts to calculate and analyze the historical data collected in the database, find the abnormal values in the historical operation data, make judgments based on experience, and finally give the diagnosis results and specific countermeasures. With the increase in the number and continuous expansion of the scale of sewage treatment plants, relying solely on experts for manual diagnosis, its efficiency and standards are not sufficient to meet the requirements of sewage treatment.

[0060] In view of the above technical problems, through the technical solution of the embodiments of the present application, the diagnosis system collects sewage data at different times corresponding to each sewage treatment system, realizing comprehensive monitoring of the dynamic changes in the sewage treatment process. And this technical solution uses a feature extraction model to analyze the correlation of sewage data between different times, so as to extract feature information according to the correlation with the sewage data of the previous moment, thereby paying attention to the influence between the sewage data at different times and tracing the source of the fault. And this technical solution combines a classification model to perform efficient classification processing on the extracted first feature information, improving the speed of fault diagnosis and enhancing the reliability of the diagnosis result. Thus, through model processing, this technical solution reduces the dependence on manual experience and subjective judgment, reduces the misjudgment rate caused by human factors, and also reduces the workload of technicians. And this technical solution can timely warn of potential faults through automated and intelligent fault diagnosis, and take measures in advance to avoid the occurrence of major accidents. Furthermore, it solves the technical problems of low diagnosis efficiency and high cost existing in the prior art, which uses manual means to diagnose the faults of equipment in sewage treatment plants.

[0061] Optionally, the operation of extracting features from the sewage data at the first moment according to the correlation relationship between the sewage data at the first moment and the sewage data at the second moment in a preset period by the feature extraction model to generate the first feature information includes: linearly transforming the sewage data at multiple consecutive moments in the preset period by the feature extraction model to generate corresponding query vectors, key vectors, and value vectors; calculating the similarity between the query vector at the first moment and the key vector at the second moment by the feature extraction model; normalizing the similarity by the feature extraction model through the softmax function to generate corresponding normalized values; and performing a weighted summation operation on the normalized values as weights and the corresponding value vectors at the second moment by the feature extraction model to generate the first feature information.

[0062] Specifically, the diagnosis system 200 linearly transforms the sewage data A at n moments by using a preset weight matrix through the feature extraction model i to generate corresponding query vectors Q i , key vectors K i and value vectors V i .

[0063] Further, the diagnosis system 200 determines the query vector Q i corresponding to the sewage data A at the first moment i , key vector K i and value vector V i , and determines the query vector Q g corresponding to the sewage data A at the second moment g , key vector Kg Sum vector V g . Refer to Figure 4 As shown, the diagnostic system 200 then calculates the query vector Q at the first moment through the feature extraction model i and the key vector K at the second moment g (i.e., Figure 4 K1 to K in i ) and the similarity s g,i (i.e., Figure 4 s in 1,i ~s i,i ). The method for calculating the similarity can be the way of taking the inner product. The calculation formula for the similarity is as follows:

[0064] s g,i =F(Q i ,K g )=Q i K g .

[0065] For example, for the similarity s g between the query vector Q1 at the first moment and the key vector K g,1 at the second moment, the similarity s 1,1 between the query vector Q1 and the key vector K1 can be calculated as follows:

[0066] s 1,1 =F(Q1,K1)=Q1K1.

[0067] For the similarity s g between the query vector Q2 at the first moment and the key vector K g,2 at the second moment, the similarity s 1,2 between the query vector Q2 and the key vector K1 can be calculated as follows:

[0068] s 1,2 =F(Q2,K1)=Q2K1.

[0069] Calculate the similarity s 2,2 between the query vector Q2 and the key vector K2:

[0070] s 2,2 =F(Q2,K2)=Q2K2.

[0071] For the similarity s g between the query vector Q3 at the first moment and the key vector K g,3 at the second moment, the similarity s 1,3 between the query vector Q3 and the key vector K1 can be calculated as follows:

[0072] s 1,3 =F(Q3,K1)=Q3K1.

[0073] Calculate the similarity s between the query vector Q3 and the key vector K2 2,3 :

[0074] s 2,3 = F(Q3, K2) = Q3K2.

[0075] Calculate the similarity s between the query vector Q3 and the key vector K3 3,3 :

[0076] s 3,3 = F(Q3, K3) = Q3K3.

[0077] And so on, for the query vector Q at the first moment n and the key vector K at the second moment g the similarity s between them g,n , the similarity s between the query vector Q n and the key vector K1 can be calculated 1,n :

[0078] s 1,n = F(Q n , K1) = Q n K1.

[0079] Calculate the similarity s between the query vector Q n and the key vector K2 2,n :

[0080] s 2,n = F(Q n , K2) = Q n K2.

[0081] And so on, calculate the similarity s between the query vector Q n and the key vector K n between them n,n .

[0082] s n,n = F(Q n , K n ) = Q n K n .

[0083] Furthermore, the diagnostic system 200 normalizes the similarity s through the softmax function by the feature extraction model to generate the corresponding normalized value p g,i : g,i :

[0084] .

[0085] where d k represents the query vector Qi Characteristic dimension

[0086] Further, the diagnostic system 200 uses the feature extraction model to normalize the value p g,i (i.e., Figure 4 p in 1,i ~p i,i ) as weights and performs a weighted summation operation with the corresponding value vector V g (i.e., Figure 4 V1~V in i ) to generate the first feature information [WA]=[W1,W2,...,W n . The calculation formula for the first feature information W i is:

[0087] .

[0088] Therefore, this technical solution calculates the correlation between sewage data at different times, extracts key features based on past sewage data, reduces human error, and shortens the fault diagnosis time.

[0089] Optionally, the operation of linearly transforming the sewage data at multiple consecutive times in a preset period through the feature extraction model to generate the corresponding query vector, key vector, and value vector includes: linearly transforming the sewage data at multiple consecutive times in a preset period through the feature extraction model according to a preset weight matrix to generate the query vector, key vector, and value vector.

[0090] Specifically, the diagnostic system 200 is pre-set with a weight matrix B Q , B K , B V . Among them, the weight matrix B Q is used to generate the query vector, the weight matrix B K is used to generate the key vector, and the weight matrix B V is used to generate the value vector.

[0091] Therefore, the diagnostic system 200 uses the feature extraction model to generate the query vector Q i based on the sewage data A Q and the weight matrix B i :

[0092] Q i =A i B Q .

[0093] The diagnostic system 200 uses the feature extraction model to generate the key vector K based on the sewage data A i and the weight matrix B K ​i :

[0094] K i = A i B K 。

[0095] The diagnostic system 200 generates a value vector V based on the sewage data A i and the weight matrix B V : i V

[0096] = A i B i 。 V 。

[0097] Therefore, this technical solution generates a query vector, a key vector, and a value vector according to the weight matrix, dynamically adjusts the focus of attention according to the relative importance of sewage data at different times, and accurately locates abnormal signals in complex data streams.

[0098] Optionally, the operation of classifying the first feature information through a classification model to determine whether a sewage treatment device fails includes: processing the first feature information through the fully connected layer of the classification model to generate second feature information; and classifying the second feature information through multiple binary classifiers to determine whether the sewage treatment device fails.

[0099] Specifically, as shown in Figure 5 , the diagnostic system 200 is pre-set with a classification model, which includes a fully connected layer and multiple binary classifiers. The classification model is pre-trained by the diagnostic system 200 using the feature information output by the feature extraction model as a sample through the gradient descent method.

[0100] The diagnostic system 200 inputs the first feature information [WA] output by the feature extraction model into the classification model. The classification model inputs the first feature information [WA] into the hidden layer through the input layer of the fully connected layer. After the hidden layer processes the first feature information [WA], the output layer outputs the second feature information [PA] corresponding to the first feature information [WA] = [pa1, pa2,..., pa L T 。

[0101] Further, the diagnostic system 200 inputs the second feature information into the binary classifier, and classifies the second feature information through the binary classifier to determine whether each sewage treatment device 101~10m fails.

[0102] Therefore, this technical solution determines whether a sewage treatment device has a fault through a classification model, which not only improves the speed of fault diagnosis, but also enhances the reliability of the diagnosis result.​

[0103] Optionally, the operation of classifying the second feature information through multiple binary classifiers to determine whether a sewage treatment device fails includes: determining the probability of failure of the sewage treatment device according to the second feature information through the binary classifier corresponding to the sewage treatment device; and determining whether the corresponding sewage treatment device fails according to the probability.

[0104] Specifically, referring to Figure 5 as shown, the diagnostic system 200 outputs the second feature information [PA]=[pa1, pa2,..., pa L T through the fully connected layer of the classification model, and the diagnostic system 200 is pre-set with a binary classifier corresponding to the second feature information [PA], where the number of binary classifiers is the same as the number of sewage treatment devices, and each binary classifier is used to classify different sewage treatment devices to determine the probability of failure of the sewage treatment device.

[0105] For example, the second feature information pa1 and pa2 are the feature information of the sewage treatment device 101, and the binary classifier 1 is used to determine the probability of failure of the sewage treatment device 101. Thus, the diagnostic system 200 inputs the second feature information pa1 and pa2 into the binary classifier 1 through the classification model, and the binary classifier 1 determines the probability pb1 of failure of the sewage treatment device 101 according to the second feature information pa1 and pa2.

[0106] The second feature information pa3 and pa4 are the feature information of the sewage treatment device 103, and the binary classifier 2 is used to determine the probability of failure of the sewage treatment device 102. Thus, the diagnostic system 200 inputs the second feature information pa3 and pa4 into the binary classifier 2 through the classification model, and the binary classifier 2 determines the probability pb2 of failure of the sewage treatment device 102 according to the second feature information pa3 and pa4.

[0107] And so on, the second feature information pa L-1 and pa L are the feature information of the sewage treatment device 10m, and the binary classifier m is used to determine the probability of failure of the sewage treatment device 10m. Thus, the diagnostic system 200 inputs the second feature information pa L-1 and pa L into the binary classifier m through the classification model, and the binary classifier m determines the probability pb L-1 and pa L to determine the probability pb m of failure of the sewage treatment device 10m.

[0108] Furthermore, the diagnostic system 200 determines whether the probability pb​j Whether (j = 1 to m) is greater than a preset threshold (e.g., 0.5). If it is greater than the preset threshold, it is determined that the corresponding sewage treatment equipment has failed.

[0109] Therefore, in this technical solution, by setting a binary classification model for each sewage treatment equipment, it is possible to respectively determine whether each sewage treatment equipment has a failure and accurately locate the sewage treatment equipment with a failure.

[0110] In summary, referring to Figure 6 as shown, the sequential steps for determining whether each sewage treatment equipment has a failure are as follows:

[0111] S601: The diagnosis system 200 collects sewage data of the sewage treatment equipment at multiple consecutive moments in a preset period.

[0112] S602: The diagnosis system 200 linearly transforms the sewage data at multiple consecutive moments in the preset period through the feature extraction model according to a preset weight matrix to generate a query vector, a key vector, and a value vector.

[0113] S603: The diagnosis system 200 calculates the similarity between the query vector at the first moment and the key vector at the second moment through the feature extraction model.

[0114] S604: The diagnosis system 200 normalizes the similarity through the softmax function of the feature extraction model to generate a corresponding normalized value.

[0115] S605: The diagnosis system 200 performs a weighted summation operation on the normalized value as a weight and the corresponding value vector at the second moment through the feature extraction model to generate first feature information.

[0116] S606: The diagnosis system 200 processes the first feature information through the fully connected layer of the classification model to generate second feature information.

[0117] S607: The diagnosis system 200 determines the probability of the sewage treatment equipment having a failure according to the second feature information through a binary classifier corresponding to the sewage treatment equipment.

[0118] S608: The diagnosis system 200 determines whether the corresponding sewage treatment equipment has a failure according to the probability.

[0119] In addition, referring to Figure 1 as shown, according to the second aspect of this embodiment, a computer system is provided, including a memory, a processor, and a computer program stored on the memory. The processor executes the computer program to implement the steps of the above method.

[0120] In addition, referring to Figure 1As shown, according to the third aspect of this embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps of the above method are implemented.

[0121] In addition, referring to Figure 1 As shown, according to the fourth aspect of this embodiment, a computer program product is provided, including a computer program. When the computer program is executed by a processor, the steps of the above method are implemented.

[0122] Thus, according to this embodiment, the diagnosis system collects sewage data at different times corresponding to each sewage treatment system, realizing comprehensive monitoring of the dynamic changes in the sewage treatment process. And this technical solution uses a feature extraction model to analyze the correlation of sewage data between different times, so as to extract feature information according to the correlation with the sewage data of the previous time, thereby paying attention to the influence existing between the sewage data of different times and tracing the source of the fault. And this technical solution combines a classification model to efficiently classify the extracted first feature information, improving the speed of fault diagnosis and enhancing the reliability of the diagnosis result. Thus, through model processing, this technical solution reduces the dependence on manual experience and subjective judgment, reduces the misjudgment rate caused by human factors, and also reduces the work burden of technical personnel. And through automated and intelligent fault diagnosis, this technical solution can timely warn of potential faults and take measures in advance to avoid major accidents. Furthermore, it solves the technical problems of low diagnosis efficiency and high cost caused by using manual methods to diagnose faults in the equipment of sewage treatment plants in the prior art.

[0123] It should be noted that for the foregoing method embodiments, for the sake of simple description, they are all expressed as a series of action combinations. However, those skilled in the art should know that the present invention is not limited by the described action sequence, because according to the present invention, certain steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily essential to the present invention.

[0124] Through the description of the above embodiments, those skilled in the art can clearly understand that the method according to the above embodiments can be implemented by means of software plus a necessary general hardware platform. Of course, it can also be implemented by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes several instructions to enable a terminal device (which can be a mobile phone, computer, server, or network device, etc.) to execute the methods described in various embodiments of the present invention.

[0125] Embodiment 2

[0126] Figure 7 Fig. 700 shows a device for determining faults in a sewage treatment device according to this embodiment. The device 700 corresponds to the method described in the first aspect of Embodiment 1. Refer to Figure 7 As shown, the device 700 includes: a data acquisition module 710 for acquiring sewage data of the sewage treatment device at a plurality of consecutive moments in a preset period; an information generation module 720 for extracting features from the sewage data at the first moment according to the correlation relationship between the sewage data at the first moment and the sewage data at the second moment in the preset period through a feature extraction model to generate first feature information, where the second moment is the moment before the first moment in the preset period; and a fault determination module 730 for classifying the first feature information through a classification model to determine whether the sewage treatment device has a fault.

[0127] Optionally, the information generation module 720 includes: a first generation sub-module for linearly transforming the sewage data at a plurality of consecutive moments in the preset period through a feature extraction model to generate corresponding query vectors, key vectors, and value vectors; a first calculation sub-module for calculating the similarity between the query vector at the first moment and the key vector at the second moment through a feature extraction model; a second generation sub-module for normalizing the similarity through a softmax function through a feature extraction model to generate corresponding normalized values; and a third generation sub-module for performing a weighted summation operation on the normalized values as weights and the corresponding value vectors at the second moment through a feature extraction model to generate first feature information.

[0128] Optionally, the first generation sub-module includes: a first generation unit for linearly transforming the sewage data at a plurality of consecutive moments in the preset period through a feature extraction model according to a preset weight matrix to generate query vectors, key vectors, and value vectors.

[0129] Optionally, the fault determination module 730 includes: a fourth generation sub-module for processing the first feature information through the fully connected layer of the classification model to generate second feature information; and a classification sub-module for classifying the second feature information through a plurality of binary classifiers to determine whether the sewage treatment device has a fault.

[0130] Optionally, the classification sub-module includes: a first determination unit for determining the probability of the sewage treatment device having a fault according to the second feature information through a binary classifier corresponding to the sewage treatment device; and a second determination unit for determining whether the corresponding sewage treatment device has a fault according to the probability.

[0131] Therefore, according to this embodiment, the diagnostic system collects sewage data at different times corresponding to each sewage treatment system, achieving comprehensive monitoring of the dynamic changes in the sewage treatment process. And this technical solution uses a feature extraction model to analyze the correlation of sewage data between different times, so as to extract feature information based on the correlation with the sewage data of the previous time, thereby paying attention to the influence between the sewage data at different times and tracing the source of the fault. And this technical solution combines a classification model to efficiently classify the extracted first feature information, improving the speed of fault diagnosis and enhancing the reliability of the diagnosis result. Therefore, through model processing, this technical solution reduces the dependence on manual experience and subjective judgment, reduces the misjudgment rate caused by human factors, and also reduces the workload of technicians. And through automated and intelligent fault diagnosis, this technical solution can timely warn of potential faults and take measures in advance to avoid major accidents. Furthermore, it solves the technical problems of low diagnostic efficiency and high cost caused by using manual methods to diagnose equipment faults in sewage treatment plants in the prior art.

[0132] Embodiment 3

[0133] Figure 8 Fig. shows a sewage treatment equipment fault determination device 800 according to this embodiment. The device 800 corresponds to the method according to the first aspect of Embodiment 1. Refer to Figure 8 As shown, the device 800 includes: a processor 810; and a memory 820, connected to the processor 810, for providing instructions for the processor 810 to perform the following processing steps: collecting sewage data of the sewage treatment equipment at multiple consecutive times in a preset period; extracting features from the sewage data of the first time through a feature extraction model according to the correlation between the sewage data of the first time and the sewage data of the second time in the preset period to generate first feature information, where the second time is the time before the first time in the preset period; and classifying the first feature information through a classification model to determine whether the sewage treatment equipment has a fault.

[0134] Optionally, the operation of extracting features from the sewage data at the first moment in a preset period according to the correlation relationship between the sewage data at the first moment and the sewage data at the second moment by a feature extraction model to generate first feature information includes: linearly transforming the sewage data at multiple consecutive moments in the preset period by the feature extraction model to generate corresponding query vectors, key vectors, and value vectors; calculating the similarity between the query vector at the first moment and the key vector at the second moment by the feature extraction model; normalizing the similarity by the softmax function through the feature extraction model to generate corresponding normalized values; and performing a weighted summation operation on the normalized values as weights and the corresponding value vectors at the second moment through the feature extraction model to generate first feature information.

[0135] Optionally, the operation of linearly transforming the sewage data at multiple consecutive moments in a preset period by a feature extraction model to generate corresponding query vectors, key vectors, and value vectors includes: linearly transforming the sewage data at multiple consecutive moments in the preset period by the feature extraction model according to a preset weight matrix to generate query vectors, key vectors, and value vectors.

[0136] Optionally, the operation of classifying the first feature information by a classification model to determine whether a sewage treatment device fails includes: processing the first feature information through the fully connected layer of the classification model to generate second feature information; and classifying the second feature information through multiple binary classifiers to determine whether the sewage treatment device fails.

[0137] Optionally, the operation of classifying the second feature information through multiple binary classifiers to determine whether a sewage treatment device fails includes: determining the probability of failure of the sewage treatment device according to the second feature information by the binary classifier corresponding to the sewage treatment device; and determining whether the corresponding sewage treatment device fails according to the probability.

[0138] Thus, according to this embodiment, the diagnostic system collects sewage data at different times corresponding to each sewage treatment system, achieving comprehensive monitoring of the dynamic changes in the sewage treatment process. And this technical solution uses a feature extraction model to analyze the correlation of sewage data between different times, and thus extracts feature information based on the correlation with the sewage data of the previous time, so as to pay attention to the influence between the sewage data at different times and trace the source of the fault occurrence. And this technical solution combines a classification model to perform efficient classification processing on the extracted first feature information, improving the speed of fault diagnosis and enhancing the reliability of the diagnosis result. Thus, through model processing, this technical solution reduces the dependence on manual experience and subjective judgment, reduces the misjudgment rate caused by human factors, and also reduces the workload of technical personnel. And through automated and intelligent fault diagnosis, this technical solution can timely warn of potential faults and take measures in advance to avoid the occurrence of major accidents. Furthermore, it solves the technical problems of low diagnostic efficiency and high cost in the prior art caused by using manual methods to diagnose the faults of equipment in sewage treatment plants.

[0139] The serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.

[0140] In the above embodiments of the present invention, the descriptions of each embodiment have their own focuses. For the parts not detailed in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0141] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces. The indirect couplings or communication connections of the units or modules can be in electrical or other forms.

[0142] The units described as separate components may or may not be physically separated. The components displayed as units may or may not be physical units, that is, they can be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0143] In addition, in each embodiment of the present invention, the functional units can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of a software functional unit.

[0144] If the above-mentioned integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in each embodiment of the present invention. The foregoing storage medium includes: various media such as USB flash drives, read-only memories (ROMs), random access memories (RAMs), mobile hard disks, magnetic disks, or optical discs that can store program codes.

[0145] The above are only the preferred embodiments of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the principle of the present invention, several improvements and refinements can be made, and these improvements and refinements should also be regarded as the protection scope of the present invention.

Claims

1. A method for determining faults in a sewage treatment device, characterized in that, Including: Collecting sewage data of sewage treatment equipment at multiple consecutive moments in a preset period, where the sewage treatment equipment is multiple sewage treatment equipment that complete the sewage treatment work sequence, and the sewage data corresponding to each moment includes different types of sewage data corresponding to each sewage treatment equipment at the corresponding moment respectively, so that the sewage data at each consecutive moment constitutes a vector sequence; Based on the correlation relationship between the sewage data at each first moment and the corresponding sewage data at the second moment in the preset period, the feature extraction model extracts features from the sewage data at the first moment to generate first feature information corresponding to the sewage data at the first moment, where the second moment is the moment before the corresponding first moment in the preset period; and Classifying the first feature information through a classification model to determine whether the sewage treatment equipment fails, where Based on the correlation relationship between the sewage data at the first moment and the sewage data at the second moment in the preset period, the operation of the feature extraction model extracting features from the sewage data at the first moment to generate first feature information includes: The feature extraction model linearly transforms the sewage data at multiple consecutive moments in the preset period to generate corresponding query vectors, key vectors, and value vectors; The feature extraction model calculates the similarity between the query vector at the first moment and the key vector at the corresponding second moment; The feature extraction model normalizes the similarity through a softmax function to generate a corresponding normalized value; and The feature extraction model uses the normalized value as a weight to perform a weighted summation operation with the value vector at the corresponding second moment to generate the first feature information, and where The operation of classifying the first feature information through a classification model to determine whether the sewage treatment equipment fails includes: The fully connected layer of the classification model processes the first feature information to generate second feature information; and Classifying the second feature information through multiple binary classifiers corresponding to each sewage treatment equipment respectively to determine the sewage treatment equipment that fails.

2. The method according to claim 1, characterized in that, The operation of the feature extraction model linearly transforming the sewage data at multiple consecutive moments in the preset period to generate corresponding query vectors, key vectors, and value vectors includes: The feature extraction model linearly transforms the sewage data at multiple consecutive moments in the preset period according to a preset weight matrix to generate the query vector, the key vector, and the value vector.

3. The method according to claim 1, wherein The operation of classifying the second feature information through multiple binary classifiers to determine whether the sewage treatment equipment fails includes: The binary classifier corresponding to the sewage treatment equipment determines the probability of the sewage treatment equipment failing according to the second feature information; and Based on the probability, it is determined whether the corresponding sewage treatment equipment fails.

4. A device for determining faults in a sewage treatment equipment, characterized in that Including: A data acquisition module for acquiring sewage data of sewage treatment equipment at multiple consecutive moments in a preset period, where the sewage treatment equipment is multiple sewage treatment equipment that complete the sewage treatment work sequence, and the sewage data corresponding to each moment includes different types of sewage data corresponding to each sewage treatment equipment at the corresponding moment, so that the sewage data at each consecutive moment constitutes a vector sequence; An information generation module for extracting features from the sewage data at each first moment in the preset period through a feature extraction model according to the correlation relationship between the sewage data at the corresponding second moment, and generating first feature information corresponding to the sewage data at the first moment, where the second moment is the moment before the corresponding first moment in the preset period; And A fault determination module for classifying and processing the first feature information through a classification model to determine whether the sewage treatment equipment has a fault, where the information generation module includes: A first generation sub-module for linearly transforming the sewage data at multiple consecutive moments in the preset period through the feature extraction model to generate corresponding query vectors, key vectors, and value vectors; A first calculation sub-module for calculating the similarity between the query vector at the first moment and the key vector at the corresponding second moment through the feature extraction model; A second generation sub-module for normalizing the similarity through the softmax function through the feature extraction model to generate a corresponding normalized value; and A third generation sub-module for performing a weighted sum operation on the normalized value as a weight and the value vector at the corresponding second moment through the feature extraction model to generate the first feature information, and where The fault determination module includes: processing the first feature information through the fully connected layer of the classification model to generate second feature information; and classifying and processing the second feature information through multiple binary classifiers corresponding to each sewage treatment equipment to determine the sewage treatment equipment with a fault.

5. A device for determining faults in a sewage treatment equipment, characterized in that, Including: A processor; And A memory connected to the processor for providing instructions for the processor to perform the following processing steps: Acquiring sewage data of sewage treatment equipment at multiple consecutive moments in a preset period, where the sewage treatment equipment is multiple sewage treatment equipment that complete the sewage treatment work sequence, and the sewage data corresponding to each moment includes different types of sewage data corresponding to each sewage treatment equipment at the corresponding moment, so that the sewage data at each consecutive moment constitutes a vector sequence; Extracting features from the sewage data at each first moment in the preset period through a feature extraction model according to the correlation relationship between the sewage data at the corresponding second moment, and generating first feature information corresponding to the sewage data at the first moment, where the second moment is the moment before the corresponding first moment in the preset period; and Classify the first feature information through a classification model to determine whether a fault occurs in the sewage treatment equipment, where The operation of extracting features from the sewage data at the first moment in the preset period according to the correlation relationship between the sewage data at the first moment and the sewage data at the second moment through a feature extraction model to generate the first feature information includes: Linearly transform the sewage data at multiple consecutive moments in the preset period through the feature extraction model to generate corresponding query vectors, key vectors, and value vectors; Calculate the similarity between the query vector at the first moment and the key vector at the corresponding second moment through the feature extraction model; Normalize the similarity through the softmax function through the feature extraction model to generate corresponding normalized values; and Use the normalized value as a weight through the feature extraction model to perform a weighted summation operation with the value vector at the corresponding second moment to generate the first feature information, and where The operation of classifying the first feature information through a classification model to determine whether a fault occurs in the sewage treatment equipment includes: A fourth generation sub-module for processing the first feature information through the fully connected layer of the classification model to generate second feature information; and A classification sub-module for classifying the second feature information through multiple binary classifiers corresponding to each sewage treatment equipment to determine the sewage treatment equipment with a fault.

6. A computer system, comprising a memory, a processor, and a computer program stored on the memory, characterized in that, The processor executes the computer program to implement the steps of the method according to claim 1.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the method according to claim 1.

8. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the method according to claim 1.

Images