CN116384158A

CN116384158A - Sewage treatment equipment operation monitoring method and system based on big data

Info

Publication number: CN116384158A
Application number: CN202310605506.4A
Authority: CN
Inventors: 陈华; 孟辉
Original assignee: Guangdong Hecheng Environmental Engineering Co ltd
Current assignee: Guangdong Hecheng Environmental Engineering Co ltd
Priority date: 2023-05-26
Filing date: 2023-05-26
Publication date: 2023-07-04
Anticipated expiration: 2043-05-26
Also published as: CN116384158B

Abstract

According to the sewage treatment equipment operation monitoring method and system based on big data, AI description knowledge corresponding to a target operation monitoring data set is determined based on the first data processing model, a matching representative feature list corresponding to the second data processing model is obtained, the matching description knowledge is determined, a state representative feature list corresponding to the third data processing model is obtained, and the involved description knowledge is determined. And the AI description knowledge, the matching description knowledge and the involved description knowledge are interacted to obtain target interaction description knowledge, and the running state of the target equipment is obtained through the running state identification network. The target equipment operation state of the target operation monitoring data set can be accurately identified through the target operation monitoring model obtained through co-optimization. And identifying the target interaction description knowledge obtained through fusion based on an operation state identification network in the target operation monitoring model, so that the accurate equipment operation state corresponding to the target operation monitoring data set can be obtained.

Description

Sewage treatment equipment operation monitoring method and system based on big data

Technical Field

The application relates to the technical field of data processing, in particular to a sewage treatment equipment operation monitoring method and system based on big data.

Background

With the promotion of park cities and green urban lands, the urban environment is treated in urban construction. The sewage treatment aims at changing the property of sewage and separating pollutant to obtain clean water which does not harm the environmental water area. The sewage treatment involves a plurality of links including large pollutant interception, microorganism treatment, sludge treatment and the like, and each link needs to use corresponding sewage treatment equipment. With the development of the Internet of things and big data analysis technology, the state of sewage treatment can be effectively monitored by collecting the running big data of sewage treatment equipment, and intervention is effectively performed when the sewage treatment is abnormal due to equipment faults, improper parameter adjustment and the like, so that the real-time performance and the efficiency can be improved compared with manual monitoring. The background monitoring of sewage treatment often covers the sewage treatment equipment data of whole city, and the data volume is big, and the monitoring degree of difficulty is high, how to carry out accurate equipment running state monitoring is the technical problem that needs to solve.

Disclosure of Invention

In view of this, the present application provides at least one sewage treatment plant operation monitoring method based on big data.

The technical scheme of the application is realized as follows:

on one hand, the application provides a sewage treatment equipment operation monitoring method based on big data, which is applied to an operation monitoring terminal, and the method comprises the following steps:

acquiring a target operation monitoring data set of sewage treatment equipment to be monitored, and determining AI description knowledge corresponding to the target operation monitoring data set based on a first data processing model; the first data processing model is a target operation monitoring model corresponding to the target operation monitoring data set; the target operation monitoring model comprises a second data processing model and a third data processing model which are different from the first data processing model;

acquiring a matching representative feature list corresponding to the second data processing model, and determining matching description knowledge corresponding to the target operation monitoring data set according to the target operation monitoring data set and the matching representative features in the matching representative feature list;

acquiring a state representative feature list corresponding to the third data processing model, and determining the related description knowledge corresponding to the target operation monitoring data set according to the target operation monitoring data set and the related data item representative features in the state representative feature list;

And carrying out knowledge interaction on the AI description knowledge, the matching description knowledge and the involving description knowledge to obtain target interaction description knowledge of the target operation monitoring data set, inputting the target interaction description knowledge into an operation state recognition network of the target operation monitoring model, and obtaining the operation state of target equipment corresponding to the target operation monitoring data set based on the operation state recognition network.

In some embodiments, the acquiring a target operation monitoring data set of the sewage treatment apparatus to be monitored, determining AI description knowledge corresponding to the target operation monitoring data set based on a first data processing model, includes:

performing data item decomposition on the target operation monitoring data set to obtain decomposed data items of the target operation monitoring data set, and performing decomposed data item embedding on the decomposed data items to obtain decomposed data item embedding description knowledge corresponding to the decomposed data items;

determining the data item arrangement of the decomposed data items in the target operation monitoring data set, and encoding arrangement information of the data item arrangement to obtain arrangement information description knowledge corresponding to the data item arrangement;

determining intermittent description knowledge corresponding to the decomposed data item, embedding the decomposed data item into the description knowledge, combining the description knowledge with the arrangement information description knowledge and the intermittent description knowledge, and obtaining to-be-extracted AI description knowledge of the decomposed data item;

Inputting the AI description knowledge to be refined into a first data processing model in a target operation monitoring model, performing knowledge refinement on the AI description knowledge to be refined based on the first data processing model to obtain data set refined description knowledge corresponding to the decomposed data item, and determining the AI description knowledge corresponding to the target operation monitoring data set according to the data set refined description knowledge corresponding to the decomposed data item.

In some, the first data processing model includes a target knowledge refinement network; the target knowledge extraction network comprises an integrated projection focusing module, a first unified arrangement module, a multi-layer perception module and a second unified arrangement module; inputting the AI description knowledge to be refined into a first data processing model in a target operation monitoring model, performing knowledge refinement on the AI description knowledge to be refined based on the first data processing model to obtain data set refinement description knowledge corresponding to the decomposition data item, and determining the AI description knowledge corresponding to the target operation monitoring data set according to the data set refinement description knowledge corresponding to the decomposition data item, wherein the determining comprises the following steps:

inputting the AI description knowledge to be extracted into the integrated projection focusing module in a first data processing model of the target operation monitoring model, and extracting the description knowledge of the AI description knowledge to be extracted based on the integrated projection focusing module to obtain first depth description knowledge corresponding to the AI description knowledge to be extracted;

Inputting the AI description knowledge to be extracted and the first depth description knowledge into the first unified arrangement module, connecting the AI description knowledge to be extracted and the first depth description knowledge in a cross-layer identical manner based on the first unified arrangement module to obtain first optimized description knowledge, and carrying out unified arrangement on the first optimized description knowledge to obtain first unified arrangement description knowledge corresponding to the AI description knowledge to be extracted;

inputting the first unified arrangement description knowledge into the multi-layer sensing module, and extracting the description knowledge of the first unified arrangement description knowledge based on the multi-layer sensing module to obtain second depth description knowledge corresponding to the first unified arrangement description knowledge;

inputting the first unified arrangement description knowledge and the second depth description knowledge into the second unified arrangement module, performing cross-layer identical connection on the first unified arrangement description knowledge and the second depth description knowledge based on the second unified arrangement module to obtain second optimization description knowledge, performing unified arrangement on the second optimization description knowledge to obtain second unified arrangement description knowledge corresponding to the to-be-extracted AI description knowledge, obtaining data set extraction description knowledge corresponding to the decomposed data items according to the second unified arrangement description knowledge, and determining AI description knowledge corresponding to the target operation monitoring data set according to the data set extraction description knowledge corresponding to the decomposed data items.

In some embodiments, the integrated projection focusing module includes a target self-directed saliency determination module, a first summation module, a knowledge interaction module, and a second summation module corresponding to the target self-directed saliency determination module; the knowledge interaction module is used for carrying out knowledge interaction on the descriptive knowledge obtained based on each self-oriented significance determination module in the integrated projection focusing module; one self-directed saliency determination module corresponds to one first sum module; the inputting the AI description knowledge to be refined into the integrated projection focusing module in the first data processing model of the target operation monitoring model, extracting the AI description knowledge to be refined based on the integrated projection focusing module, and obtaining a first depth description knowledge corresponding to the AI description knowledge to be refined, including:

in a first data processing model of the target operation monitoring model, determining a target self-directed saliency determination module in a plurality of self-directed saliency determination modules of the integrated projection focusing module;

determining first execution information, second execution information and third execution information corresponding to the to-be-extracted AI description knowledge according to the to-be-extracted AI description knowledge and a first summarization module corresponding to the target self-directed significance determination module;

Inputting the first execution information, the second execution information and the third execution information into the target self-oriented significance determination module, and processing the first execution information, the second execution information and the third execution information based on the target self-oriented significance determination module to obtain description knowledge corresponding to the target self-oriented significance determination module;

if all the self-directed saliency determination modules in the integrated projection focusing module are used as the target self-directed saliency determination modules, descriptive knowledge corresponding to each self-directed saliency determination module is obtained, knowledge interaction is carried out on the descriptive knowledge corresponding to each self-directed saliency determination module based on the knowledge interaction module, and saliency interaction descriptive knowledge corresponding to the to-be-extracted AI descriptive knowledge is obtained;

inputting the saliency interaction description knowledge into the second summarization module, and extracting the description knowledge of the saliency interaction description knowledge based on the second summarization module to obtain first depth description knowledge corresponding to the AI description knowledge to be extracted.

In some embodiments, the obtaining a matching representative feature list corresponding to the second data processing model, determining matching description knowledge corresponding to the target operation monitoring dataset according to the target operation monitoring dataset and matching representative features in the matching representative feature list, includes:

Obtaining a matching representative feature list corresponding to the second data processing model, and matching the target operation monitoring data set with the matching representative features in the matching representative feature list to obtain a matching result corresponding to the target operation monitoring data set;

when the matching result shows that the matching representative feature list contains the matching representative feature corresponding to the target operation monitoring data set, determining the matching representative feature corresponding to the target operation monitoring data set as a target matching representative feature;

inputting the matched data item combination corresponding to the target matching representative feature into the second data processing model, and encoding description knowledge of the matched data item combination based on the second data processing model to obtain matched AI description knowledge corresponding to the target matching representative feature;

determining matching description knowledge corresponding to the target operation monitoring data set according to the matching AI description knowledge;

and when the matching result indicates that the matching representative feature list does not contain the matching representative feature corresponding to the target operation monitoring data set, acquiring alternative matching description knowledge corresponding to the matching representative feature list, and determining the alternative matching description knowledge as the matching description knowledge corresponding to the target operation monitoring data set.

In some embodiments, the acquiring a state representative feature list corresponding to the third data processing model, determining, according to the target operation monitoring dataset and related data item representative features in the state representative feature list, the involvement description knowledge corresponding to the target operation monitoring dataset includes:

acquiring a state representative feature list corresponding to the third data processing model, and comparing the target operation monitoring data set with related data item representative features in the state representative feature list to obtain a comparison result corresponding to the target operation monitoring data set;

when the comparison result shows that the state representative feature list contains the relevant data item representative feature corresponding to the target operation monitoring data set, determining the relevant data item representative feature corresponding to the target operation monitoring data set as a target relevant data item representative feature;

inputting the representative features of the target related data items into the third data processing model, and encoding description knowledge of the representative features of the target related data items based on the third data processing model to obtain associated AI description knowledge corresponding to the representative features of the target related data items;

Determining the related description knowledge corresponding to the target operation monitoring data set according to the related AI description knowledge;

and when the comparison result shows that the state representative feature list does not contain the related data item representative feature corresponding to the target operation monitoring data set, acquiring alternative association description knowledge corresponding to the state representative feature list, and determining the alternative association description knowledge as the involvement description knowledge corresponding to the target operation monitoring data set.

In some, the method further comprises:

acquiring an operation monitoring training data set for optimizing an original operation monitoring model and equipment operation state labeling information of the operation monitoring training data set, and determining training description knowledge corresponding to the operation monitoring training data set based on a first training data processing model; the first training data processing model is an original operation monitoring model corresponding to the operation monitoring training data set; the original operational monitoring model includes a second training data processing model and a third training data processing model that are different from the first training data processing model;

acquiring a matching representative feature list corresponding to the second training data processing model, and determining training matching description knowledge corresponding to the running monitoring training data set according to the running monitoring training data set and the matching representative feature in the matching representative feature list;

Acquiring a state representative feature list corresponding to the third data processing model, and determining training involvement description knowledge corresponding to the operation monitoring training data set according to the target operation monitoring data set and related data item representative features in the state representative feature list;

and performing rolling optimization on the original operation monitoring model according to the training description knowledge, the training matching description knowledge, the training involvement description knowledge, the equipment operation state labeling information and the operation state recognition network of the original operation monitoring model, and determining the original operation monitoring model after rolling optimization as a target operation monitoring model.

In some embodiments, the performing rolling optimization on the original operation monitoring model according to the training description knowledge, the training matching description knowledge, the training involvement description knowledge, the equipment operation state labeling information, and the operation state recognition network of the original operation monitoring model, and determining the rolling optimized original operation monitoring model as a target operation monitoring model includes:

performing knowledge interaction on the training description knowledge, the training matching description knowledge and the training involvement description knowledge to obtain training interaction description knowledge of the operation monitoring training data set, inputting the training interaction description knowledge into an operation state recognition network of the original operation monitoring model, and obtaining the estimated equipment operation state corresponding to the operation monitoring training data set based on the operation state recognition network;

Determining a loss value of the original operation monitoring model according to the estimated equipment operation state and the equipment operation state labeling information;

when the loss value of the original operation monitoring model does not accord with the model training cut-off condition, optimizing the configuration variable of the original operation monitoring model according to the loss value which does not accord with the model training cut-off condition;

and determining an original operation monitoring model after optimizing configuration variables as a temporary operation monitoring model, performing rolling optimization on the temporary operation monitoring model until the loss value of the temporary operation monitoring model after rolling optimization meets the model training cut-off condition, and determining the temporary operation monitoring model meeting the model training cut-off condition as the target operation monitoring model.

In some, the method further comprises:

extracting candidate data items from the operation monitoring training data set according to a candidate data item acquisition mode to obtain a candidate data item set corresponding to the operation monitoring training data set, and acquiring target candidate data items to be input into an anomaly identification model from the candidate data item set;

performing anomaly recognition on the target candidate data item based on the anomaly recognition model to obtain an anomaly recognition result of the target candidate data item, and performing candidate data item confirmation on the target candidate data item to obtain a candidate data item confirmation result of the target candidate data item;

And when the abnormal recognition result indicates that the target candidate data item meets the preset condition in the candidate data item acquisition mode, and the candidate data item confirmation result indicates that the target candidate data item meets the confirmation condition in the candidate data item acquisition mode, determining the target candidate data item as a first related data item representative feature in the candidate data item set, and merging the first related data item representative feature into a state representative feature list corresponding to the third training data processing model.

In another aspect, the present application provides a sewage treatment apparatus operation monitoring system, including a data acquisition device and an operation monitoring terminal that are communicatively connected to each other, the operation monitoring terminal including a memory and a processor, the memory storing a computer program executable on the processor, the processor implementing the steps in the above method when executing the program.

The beneficial effects of this application: according to the sewage treatment equipment operation monitoring method and system based on big data, when the target operation monitoring data set of the sewage treatment equipment to be monitored is obtained, AI description knowledge corresponding to the target operation monitoring data set is determined based on the first data processing model. The first data processing model is a target operational monitoring model corresponding to the target operational monitoring dataset, the target operational monitoring model further comprising a second data processing model and a third data processing model different from the first data processing model. And acquiring a matching representative feature list corresponding to the second data processing model, and determining matching description knowledge corresponding to the target operation monitoring data set according to the target operation monitoring data set and the matching representative features in the matching representative feature list. And then, acquiring a state representative feature list corresponding to the third data processing model, and determining the related description knowledge corresponding to the target operation monitoring data set according to the target operation monitoring data set and the related data item representative features in the state representative feature list. And performing knowledge interaction on the AI description knowledge, the matching description knowledge and the involved description knowledge to obtain target interaction description knowledge of the target operation monitoring data set, inputting the target interaction description knowledge into an operation state recognition network of the target operation monitoring model, and obtaining the operation state of the target equipment corresponding to the target operation monitoring data set based on the operation state recognition network. According to the target operation monitoring model obtained through co-optimization, the target equipment operation state of the target operation monitoring data set can be accurately identified. For example, AI description knowledge of the target operational monitoring dataset may be extracted based on the first data processing model, the AI description knowledge being indicative of operational state information of the target operational monitoring dataset. And mining to obtain matching description knowledge of the target operation monitoring data set based on the second data processing model, mining to obtain involvement description knowledge of the target operation monitoring data set based on the third data processing model, and using the matching description knowledge and the involvement description knowledge as collaborative information vector expression of the target operation monitoring data set. After the AI description knowledge, the matching description knowledge and the involved description knowledge are subjected to knowledge interaction, the target interaction description knowledge obtained through fusion is identified based on an operation state identification network in a target operation monitoring model, and the accurate equipment operation state corresponding to the target operation monitoring data set is obtained.

It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the aspects of the present application.

Drawings

The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the application and, together with the description, serve to explain the technical aspects of the application.

Fig. 1 is a schematic implementation flow chart of an operation monitoring method of a sewage treatment device based on big data according to an embodiment of the present application.

Fig. 2 is a schematic diagram of a composition structure of an operation monitoring device of a sewage treatment apparatus according to an embodiment of the present application.

Fig. 3 is a schematic hardware entity diagram of an operation monitoring terminal according to an embodiment of the present application.

Detailed Description

For the purpose of making the objects, technical solutions and advantages of the present application more apparent, the technical solutions of the present application are further elaborated below in conjunction with the accompanying drawings and examples, which should not be construed as limiting the present application, and all other embodiments obtained by those skilled in the art without making inventive efforts are within the scope of protection of the present application.

In the following description, reference is made to "some embodiments" which describe a subset of all possible embodiments, but it is to be understood that "some embodiments" can be the same subset or different subsets of all possible embodiments and can be combined with one another without conflict. The term "first/second/third" is merely to distinguish similar objects and does not represent a specific ordering of objects, it being understood that the "first/second/third" may be interchanged with a specific order or sequence, as permitted, to enable embodiments of the present application described herein to be practiced otherwise than as illustrated or described herein.

Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing the present application only and is not intended to be limiting of the present application.

The embodiment of the application provides a sewage treatment equipment operation monitoring method based on big data, which can be executed by a processor of an operation monitoring terminal. The operation monitoring terminal may refer to a server, a notebook computer, a tablet computer, a desktop computer, and other devices with data processing capability. The operation monitoring terminal is in communication connection with a data acquisition device in the sewage treatment system, and the data acquisition device can acquire operation data, such as pressure, temperature, flow, water content, power and the like, of the sewage treatment equipment during operation.

Fig. 1 is a schematic implementation flow chart of a sewage treatment equipment operation monitoring method based on big data according to an embodiment of the present application, as shown in fig. 1, the method includes operations 101 to 104 as follows:

in operation S101, a target operation monitoring data set of the sewage treatment apparatus to be monitored is obtained, and AI description knowledge corresponding to the target operation monitoring data set is determined based on the first data processing model.

The sewage treatment equipment to be monitored can be equipment involved in each link of sewage treatment, and the target operation monitoring data set is obtained by carrying out data acquisition on the equipment based on the equipment by arranging a data acquisition device (such as a sensor) in the equipment. For example, the sewage treatment apparatus may include sewage pretreatment apparatus, sewage biological treatment apparatus, sludge treatment apparatus, and the like. The specific constitution of the equipment in each link is not limited, for example, sewage pretreatment equipment comprises equipment such as sewage blocking, sand setting, precipitation, air floatation and the like, sewage biological treatment equipment comprises equipment such as blast aeration, decanting, biological membrane filtration, anaerobic assistance and the like, and sludge treatment equipment comprises equipment such as sludge discharge, concentration, anaerobic digestion, dehydration and desiccation and the like. Corresponding data acquisition devices are arranged on equipment corresponding to the data link to be focused, and relevant data are acquired, such as a water content detection sensor, a compressor power detection element, a centrifugal pump flowmeter and the like arranged on the dehydration equipment. The data acquisition device sends acquired data to the operation monitoring terminal, and the operation monitoring terminal can sort the data acquired in the same time period into a target operation monitoring data set.

The operation monitoring terminal performs data item decomposition on the target operation monitoring data set (namely splitting each data constituting the operation monitoring data set to obtain a plurality of data items), obtains the decomposed data items of the target operation monitoring data set, performs decomposed data item embedding on the decomposed data items, and obtains decomposed data item embedding description knowledge corresponding to the decomposed data items (namely, a feature vector obtained after the decomposed data items are embedded is a vectorized result, which can be a feature vector, or a matrix and tensor formed by vectors, and various description knowledge related in the embodiment of the application can be understood in the same form, so that the vectorized knowledge is convenient to operate). The operation monitoring terminal determines the data item arrangement of the decomposed data items in the target operation monitoring data set, carries out arrangement information coding on the data item arrangement, and obtains arrangement information description knowledge corresponding to the data item arrangement (namely locks the distribution positions of all the data items in the operation monitoring data set, and completes position coding). Then, the operation monitoring terminal can determine intermittent description knowledge corresponding to the decomposed data item, namely, description knowledge for distinguishing data related to different sewage treatment links, which can be a vector, then embed the decomposed data item into the description knowledge, arrange information description knowledge and intermittent description knowledge to perform description knowledge combination (such as vector combination modes of splicing, adding and the like), and obtain to-be-extracted AI description knowledge (namely, description knowledge needing to be extracted and further encoded) of the decomposed data item. And then, the operation monitoring terminal inputs the AI description knowledge to be refined into a first data processing model in the target operation monitoring model, performs knowledge refinement on the AI description knowledge to be refined based on the first data processing model, encodes to obtain data set refined description knowledge corresponding to the decomposed data item, and determines AI description knowledge corresponding to the target operation monitoring data set according to the data set refined description knowledge corresponding to the decomposed data item. The first data processing model is a target operational monitoring model corresponding to (i.e., associated with) the target operational monitoring dataset, and the target operational monitoring model corresponding to the target operational monitoring dataset further includes a second data processing model and a third data processing model that are different from the first data processing model.

According to the data item decomposition, a plurality of data items are split to form a data item set according to corresponding rules, when the operation monitoring terminal obtains the decomposition data item embedded description knowledge corresponding to the decomposition data items and the arrangement information description knowledge corresponding to the data item arrangement, the description knowledge can be combined only by the decomposition data item embedded description knowledge and the arrangement information description knowledge, the intermittent description knowledge is not used, so that the to-be-extracted AI description knowledge of the decomposition data items is obtained, and the first data processing model can be a Google machine translation model. The to-be-refined AI description knowledge of the decomposed data items may be input data of the first data processing model, if the number of the decomposed data items exceeds the executable number of the first data processing model, splitting the target operation monitoring data set, and taking the to-be-refined AI description knowledge corresponding to the executable number of the decomposed data items of the target operation monitoring data set as the input data of the first data processing model. And when the number of the decomposed data items exceeds the executable number of the first data processing model, decomposing the target operation monitoring data set into a plurality of operation monitoring data sets, and sequentially inputting the quasi-refined AI description knowledge of the decomposed data items of the plurality of operation monitoring data sets into the first data processing model.

In the process of acquiring the to-be-extracted AI description knowledge, the to-be-extracted AI description knowledge input by the first data processing model is embedded with description knowledge, arrangement information description knowledge and intermittent description knowledge for the decomposed data item. The first data processing model includes a target knowledge refinement network, which is an Encoder of the machine translation model, including an integrated projection focus module, a first unified collation module, a multi-layer perception module, and a second unified collation module. Then, the process of obtaining the AI description knowledge corresponding to the target operation monitoring data set according to the AI description knowledge to be refined specifically includes: the operation monitoring terminal inputs the to-be-extracted AI description knowledge into an integrated projection focusing module in a first data processing model of the target operation monitoring model, and performs description knowledge extraction on the to-be-extracted AI description knowledge based on the integrated projection focusing module to obtain first depth description knowledge corresponding to the to-be-extracted AI description knowledge, wherein the depth description knowledge is intermediate description knowledge (implicit knowledge) between an input layer and an output layer, and the integrated projection focusing module is a multi-head attention module. The operation monitoring terminal inputs the AI description knowledge to be extracted and the first depth description knowledge into a first unified arrangement module, cross-layer identity connection, namely residual connection, is carried out on the AI description knowledge to be extracted and the first depth description knowledge based on the first unified arrangement module, residual optimization is completed to obtain first optimization description knowledge, the first optimization description knowledge is subjected to unified arrangement to obtain first unified arrangement description knowledge corresponding to the AI description knowledge to be extracted, and the unified arrangement process is, for example, standardized or normalized or centralized (zero-mean processing) operation and the like, and the corresponding unified arrangement module is a standardized module or a normalized module or a centralized module. The operation monitoring terminal can input the first unified arrangement description knowledge into the multi-layer sensing module, description knowledge extraction is carried out on the first unified arrangement description knowledge based on the multi-layer sensing module, a second depth description knowledge corresponding to the first unified arrangement description knowledge is obtained, and the multi-layer sensing module is a feed-forward network. The operation monitoring terminal inputs the first unified arrangement description knowledge and the second depth description knowledge into a second unified arrangement module, performs cross-layer identical connection on the first unified arrangement description knowledge and the second depth description knowledge based on the second unified arrangement module to obtain second optimization description knowledge, performs unified arrangement on the second optimization description knowledge to obtain second unified arrangement description knowledge corresponding to AI description knowledge to be extracted, obtains data set extraction description knowledge corresponding to the decomposition data item according to the second unified arrangement description knowledge, extracts description knowledge according to the data set extraction description knowledge corresponding to the decomposition data item, and determines AI description knowledge corresponding to the target operation monitoring data set. In the target knowledge extraction network, a first unified arrangement module and a second unified arrangement module have the same composition, an integrated projection focusing module in the target knowledge extraction network is a plurality of parallel self-directed saliency determination modules (which are modules for determining saliency characteristics constructed based on a self-attention mechanism), query, key, value (respectively, the quasi-refined AI description knowledge input by a first data processing model) is transformed by the first summation module before data are input into any self-directed saliency determination module, the output of the self-directed saliency determination modules is input into the second summation module for calculation, and the first summation module and the second summation module are all fully connected networks for carrying out summation classification. The integrated projection focusing module comprises a target self-oriented saliency determination module, a first summarization module corresponding to the target self-oriented saliency determination module, a knowledge interaction module and a second summarization module, wherein the knowledge interaction module is used for carrying out knowledge interaction on descriptive knowledge obtained based on each self-oriented saliency determination module in the integrated projection focusing module, for example, adding or weighted averaging the knowledge, and one self-oriented saliency determination module corresponds to one first summarization module. Thus, deriving first depth description knowledge corresponding to the AI description knowledge to be refined from the AI description knowledge to be refined comprises: and the operation monitoring terminal determines a target self-oriented significance determination module in a plurality of self-oriented significance determination modules contained in the integrated projection focusing module in a first data processing model of the target operation monitoring model. The operation monitoring terminal can determine first execution information, second execution information and third execution information corresponding to the to-be-extracted AI description knowledge according to the to-be-extracted AI description knowledge and a first summarizing module corresponding to the target self-directed saliency determination module, the execution information can be input parameters, the operation monitoring terminal can input the first execution information, the second execution information and the third execution information into the target self-directed saliency determination module, and the first execution information, the second execution information and the third execution information are processed based on the target self-directed saliency determination module to obtain the description knowledge corresponding to the target self-directed saliency determination module. If all the self-directed saliency determination modules in the integrated projection focusing module are used as target self-directed saliency determination modules, the operation monitoring terminal can acquire description knowledge corresponding to each self-directed saliency determination module, and the knowledge interaction module carries out knowledge interaction on the description knowledge corresponding to each self-directed saliency determination module to acquire saliency interaction description knowledge corresponding to the AI description knowledge to be extracted (namely, the saliency feature information which is determined to be concerned after interaction fusion). The operation monitoring terminal inputs the saliency interaction description knowledge into a second merging module, and description knowledge extraction is carried out on the saliency interaction description knowledge based on the second merging module, so that first depth description knowledge corresponding to the AI description knowledge to be extracted is obtained.

The integrated projection focusing module may include a plurality of self-directional saliency determination modules, and in a first summation module corresponding to different self-directional saliency determination modules, the self-directional saliency determination modules are operated together by calculating the AI description knowledge to be extracted through a weight array (for example, a two-dimensional matrix), so that different spaces are formed by the different self-directional saliency determination modules.

The target self-directed saliency determination module adds knowledge of the value descriptions of all of the resolved data items when the resolved data items are encoded.

And S102, acquiring a matching representative feature list corresponding to the second data processing model, and determining matching description knowledge corresponding to the target operation monitoring data set according to the target operation monitoring data set and the matching representative features in the matching representative feature list.

The operation monitoring terminal can acquire a matching representative feature list (comprising a plurality of matching representative features) corresponding to the second data processing model, and match the target operation monitoring data set with the matching representative features in the matching representative feature list to obtain a matching result corresponding to the target operation monitoring data set. When the matching result indicates that the matching representative feature list contains the matching representative feature corresponding to the target operation monitoring data set, the operation monitoring terminal may determine the matching representative feature corresponding to the target operation monitoring data set as the target matching representative feature. The operation monitoring terminal can input the matched data item combination corresponding to the target matching representative feature (namely, the combination constructed by a plurality of matched data items) into the second data processing model, and perform description knowledge coding on the matched data item combination based on the second data processing model to obtain the matched AI description knowledge corresponding to the target matching representative feature. The operation monitoring terminal can determine the matching description knowledge corresponding to the target operation monitoring data set according to the matching AI description knowledge. The matching representative feature is a feature stored earlier to indicate the matching method for the data item, and has high interpretation. When the matching result indicates that the matching representative feature list does not contain the matching representative feature corresponding to the target operation monitoring data set, the operation monitoring terminal acquires the alternative matching description knowledge corresponding to the matching representative feature list, determines the alternative matching description knowledge as the matching description knowledge corresponding to the target operation monitoring data set, and the alternative matching description knowledge is the description knowledge for assisting in matching, namely, the matching description knowledge not contained in the matching representative feature list. Before matching the target operation monitoring data set and the matching representative features, carrying out description knowledge coding, such as one hot coding, on the combination of the matching data items corresponding to the matching representative features in the matching representative feature list, obtaining AI description knowledge corresponding to each matching representative feature, storing the AI description knowledge corresponding to all the matching representative features in the matching representative feature list, and obtaining a matching description knowledge set, wherein when the operation monitoring terminal determines the target matching representative features corresponding to the target operation monitoring data set in the matching representative feature list, the operation monitoring terminal directly obtains the AI description knowledge corresponding to the target matching representative features in the matching description knowledge set.

For example, in the process of outputting the matching description knowledge, the matching representative feature list includes X matching representative features, including a matching mode a and a matching mode B … matching mode X, and the operation monitoring terminal matches the target operation monitoring data set with all the matching modes. And when the matching representative feature corresponding to the target operation monitoring data set is determined to be the matching mode B in all the matching modes, taking the matching mode B as the target matching representative feature, and taking the matching mode B as the target matching representative feature F. The operation monitoring terminal inputs the target matching representative feature F into a second data processing model, performs description knowledge coding on the matching data item combination corresponding to the target matching representative feature F based on the second data processing model, obtains matching description knowledge K corresponding to the target matching representative feature F, and determines the matching description knowledge K as matching description knowledge corresponding to the target operation monitoring data set.

In operation S103, a state representative feature list corresponding to the third data processing model is acquired, and knowledge of the description of the involvement corresponding to the target operation monitoring dataset is determined according to the target operation monitoring dataset and the related data item representative features in the state representative feature list.

The operation monitoring terminal can acquire a state representing characteristic list corresponding to the third data processing model, and compares the target operation monitoring data set with the related data item representing characteristic in the state representing characteristic list to obtain a comparison result corresponding to the target operation monitoring data set. Optionally, when the comparison result indicates that the state representative feature list includes the relevant data item representative feature corresponding to the target operation monitoring data set, the operation monitoring terminal determines the relevant data item representative feature corresponding to the target operation monitoring data set as the target relevant data item representative feature. And the operation monitoring terminal inputs the representative characteristics of the target related data items into a third data processing model, and carries out description knowledge coding on the representative characteristics of the target related data items based on the third data processing model to obtain associated AI description knowledge corresponding to the representative characteristics of the target related data items. The operation monitoring terminal can determine the related description knowledge corresponding to the target operation monitoring data set according to the associated AI description knowledge, namely the description knowledge with the association relation, and the related data item representative characteristics are determined in the training set. When the comparison result shows that the state representative feature list does not contain the related data item representative feature corresponding to the target operation monitoring data set, the operation monitoring terminal acquires alternative association description knowledge corresponding to the state representative feature list, the alternative association description knowledge is determined to be the involved description knowledge corresponding to the target operation monitoring data set, and the alternative association description knowledge is the description knowledge for assisting in matching and is the matching description knowledge not contained in the state representative feature list. And the operation monitoring terminal determines the related description knowledge corresponding to the target operation monitoring data set according to each associated AI description knowledge in the associated AI description knowledge, for example, the related AI description knowledge is averaged to obtain the related description knowledge.

Before the operation monitoring terminal compares the target operation monitoring data set with the related data item representative features, the operation monitoring terminal can code description knowledge of the related data item representative features in the state representative feature list to obtain associated AI description knowledge corresponding to each related data item representative feature, and store the associated AI description knowledge corresponding to all the related data item representative features in the state representative feature list to obtain a matching description knowledge set. In this way, when the operation monitoring terminal determines the object related data item representative feature corresponding to the object operation monitoring data set in the state representative feature list, the associated AI description knowledge corresponding to the object related data item representative feature is obtained in the matching description knowledge set.

And S104, carrying out knowledge interaction on the AI description knowledge, the matching description knowledge and the involved description knowledge to obtain target interaction description knowledge of the target operation monitoring data set, inputting the target interaction description knowledge into an operation state recognition network of the target operation monitoring model, and obtaining the operation state of the target equipment corresponding to the target operation monitoring data set based on the operation state recognition network.

When the target interaction description knowledge is input into the operation state identification network of the target operation monitoring model, the operation state identification network based on the target operation monitoring model obtains a trusted coefficient (such as a probability value) of an operation state of the device corresponding to the target operation monitoring data set, the operation state of the target device corresponding to the target operation monitoring data set is determined according to the trusted coefficient, the operation state of the device can comprise a first operation state (such as abnormal operation) of the device and a second operation state (such as normal operation) of the device, if the trusted coefficient accords with a preset trusted coefficient, the first operation state of the device is determined to be the operation state of the target device corresponding to the target operation monitoring data set, and if the trusted coefficient does not accord with the preset trusted coefficient, the operation state of the second device is determined to be the operation state of the target device corresponding to the target operation monitoring data set.

According to the sewage treatment equipment operation monitoring method based on big data, when the target operation monitoring data set of the sewage treatment equipment to be monitored is obtained, AI description knowledge corresponding to the target operation monitoring data set is determined based on the first data processing model. The first data processing model is a target operational monitoring model corresponding to the target operational monitoring dataset, the target operational monitoring model further comprising a second data processing model and a third data processing model different from the first data processing model. And acquiring a matching representative feature list corresponding to the second data processing model, and determining matching description knowledge corresponding to the target operation monitoring data set according to the target operation monitoring data set and the matching representative features in the matching representative feature list. And then, acquiring a state representative feature list corresponding to the third data processing model, and determining the related description knowledge corresponding to the target operation monitoring data set according to the target operation monitoring data set and the related data item representative features in the state representative feature list. And performing knowledge interaction on the AI description knowledge, the matching description knowledge and the involved description knowledge to obtain target interaction description knowledge of the target operation monitoring data set, inputting the target interaction description knowledge into an operation state recognition network of the target operation monitoring model, and obtaining the operation state of the target equipment corresponding to the target operation monitoring data set based on the operation state recognition network. According to the method and the device for identifying the target equipment operation state, the target operation monitoring model is obtained through co-optimization, and the target equipment operation state of the target operation monitoring data set can be accurately identified. For example, AI description knowledge of the target operational monitoring dataset may be extracted based on the first data processing model, the AI description knowledge being indicative of operational state information of the target operational monitoring dataset. And mining to obtain matching description knowledge of the target operation monitoring data set based on the second data processing model, mining to obtain involvement description knowledge of the target operation monitoring data set based on the third data processing model, and using the matching description knowledge and the involvement description knowledge as collaborative information vector expression of the target operation monitoring data set. After the AI description knowledge, the matching description knowledge and the involved description knowledge are subjected to knowledge interaction, the target interaction description knowledge obtained through fusion is identified based on an operation state identification network in a target operation monitoring model, and the accurate equipment operation state corresponding to the target operation monitoring data set is obtained.

As other technical solutions that may be implemented independently, the embodiments of the present application further provide a sewage treatment device operation monitoring method based on big data, including:

in operation S201, operation monitoring training data sets for optimizing an original operation monitoring model and equipment operation state labeling information of the operation monitoring training data sets are obtained, candidate data item extraction is performed on the operation monitoring training data sets according to a candidate data item obtaining mode, a candidate data item set corresponding to the operation monitoring training data sets is obtained, and target candidate data items to be input into an anomaly identification model are obtained in the candidate data item set.

The operation monitoring terminal obtains an operation monitoring training data set for optimizing an original operation monitoring model and equipment operation state labeling information (for example, the operation monitoring training data set can be represented by marks), carries out data item decomposition on the operation monitoring training data set according to a candidate data item obtaining mode to obtain training data items of the operation monitoring training data set, and carries out data item combination on the training data items according to a training data item combination rule to obtain original candidate data items corresponding to the operation monitoring training data set. For example, the operation monitoring terminal may acquire the occurrence number of the original candidate data item in the operation monitoring training data set, determine the original candidate data item whose occurrence number meets the preset number as a temporary candidate data item, determine the association degree (which may be mutual information MI) between the temporary candidate data item and the equipment operation state labeling information, and use the temporary candidate data item whose association degree meets the preset association degree in the candidate data item acquiring manner as the candidate data item to be selected. Then, the operation monitoring terminal can select the candidate data items to be selected, of which the total number of the data items accords with the preset number, from the candidate data items to be selected according to the total number of the data items of the training data items in the candidate data items to be selected. Then, the operation monitoring terminal may generate a candidate data item set corresponding to the operation monitoring training data set according to the selected candidate data item to be selected, obtain a target candidate data item to be input into the anomaly identification model in the candidate data item set, and perform data item combination, such as stitching, on a plurality of training data items based on the position of the training data item in the operation monitoring training data set, where the position of the training data item in the original candidate data item is the same as the position in the operation monitoring training data set when the training data item in the original candidate data item obtained by combination is a plurality of. The explanation of the representative feature of the relevant data item is a feature that occurs a large number of times in the running monitoring training data set, i.e., is distributed a large number of times and has a high degree of association. And determining the original candidate data items which accord with the preset occurrence times as temporary candidate data items, and determining the temporary candidate data items which accord with the preset association degree as candidate data items to be selected.

When the data item combination is carried out on the training data items according to the training data item combination rule, the total number of the data items of the original candidate data items obtained by the data item combination is determined to be in accordance with the preset number, so that when the candidate data items to be selected are obtained according to the two characteristics of the representative characteristics of the related data items, a candidate data item set corresponding to the operation monitoring training data set is generated according to the candidate data items to be selected, and the candidate data items, the total number of which is in accordance with the preset number, are not required to be selected in the candidate data items to be selected.

And S202, carrying out anomaly recognition on the target candidate data item based on the anomaly recognition model to obtain an anomaly recognition result of the target candidate data item, and carrying out candidate data item confirmation on the target candidate data item to obtain a candidate data item confirmation result of the target candidate data item.

The anomaly recognition model is a machine learning model that is optimized to recognize anomalies in the target candidate data item, and may be, for example, a google machine translation model, which may recognize anomalies, such as scoring the degree of anomalies.

In operation S203, when the abnormal recognition result indicates that the target candidate data item meets the preset condition in the candidate data item obtaining mode, and the candidate data item confirmation result indicates that the target candidate data item meets the confirmation condition in the candidate data item obtaining mode, the target candidate data item is determined as the first related data item representative feature in the candidate data item set, and the first related data item representative feature is merged into the state representative feature list. The state representative feature list is used for optimization training of the following third training data processing model.

In operation S204, a candidate operation monitoring data set isolated from the candidate data item set is acquired according to the candidate data item acquisition mode, the candidate data item in the candidate operation monitoring data set is used as a second related data item representative feature, the second related data item representative feature is merged into a state representative feature list, and both the first related data item representative feature and the second related data item representative feature in the state representative feature list are determined as related data item representative features in the state representative feature list.

Since different candidate data items may contain the same operational state information, for data items that continue to appear, new state information may be mapped to related data item representative features whose state information characterizes a close relationship, which may increase the likelihood of new data items being identified.

And the data quantity limited by the operation monitoring training data set is likely to be single, so that the relevant data item representative features extracted from the operation monitoring training data set are acquired, the replacement data item in the replacement operation monitoring data set is taken as a second relevant data item representative feature, and the second relevant data item representative feature is integrated into the state representative feature list. The replacement data items in the replacement operation monitoring data set, that is, the second related data item representative feature and the first related data item representative feature are different, and the operation monitoring terminal uses the replacement data item which is acquired in the replacement operation monitoring data set and is different from the first related data item representative feature as the second related data item representative feature.

In operation S205, training description knowledge corresponding to the running monitoring training data set is determined based on the first training data processing model.

The first training data processing model is an original operational monitoring model corresponding to the operational monitoring training data set, the original operational monitoring model further comprising a second training data processing model and a third training data processing model different from the first training data processing model. The method for determining the training description knowledge corresponding to the operation monitoring training data set by the operation monitoring terminal based on the first training data processing model is the same as the process for determining the AI description knowledge corresponding to the target operation monitoring data set based on the first data processing model.

And S206, acquiring a matching representative feature list corresponding to the second training data processing model, and determining training matching description knowledge corresponding to the operation monitoring training data set according to the operation monitoring training data set and the matching representative features in the matching representative feature list.

The method for determining the training matching description knowledge corresponding to the operation monitoring training data set by the operation monitoring terminal based on the second training data processing model is the same as the process for determining the matching description knowledge corresponding to the target operation monitoring data set based on the second data processing model.

In operation S207, a state representative feature list corresponding to the third data processing model is acquired, and training involvement description knowledge corresponding to the operation monitoring training data set is determined according to the target operation monitoring data set and the related data item representative features in the state representative feature list.

The method for determining the training involvement description knowledge corresponding to the operation monitoring training data set by the operation monitoring terminal based on the third training data processing model is the same as the process for determining the involvement description knowledge corresponding to the target operation monitoring data set based on the third data processing model.

And S208, performing rolling optimization on the original operation monitoring model according to the training description knowledge, the training matching description knowledge, the training involvement description knowledge, the equipment operation state labeling information and the operation state recognition network of the original operation monitoring model, and determining the original operation monitoring model after the rolling optimization as a target operation monitoring model.

Specifically, the operation monitoring terminal carries out knowledge interaction on training description knowledge, training matching description knowledge and training involving description knowledge to obtain training interaction description knowledge of an operation monitoring training data set, inputs the training interaction description knowledge into an operation state recognition network of an original operation monitoring model, and obtains the operation state of the pre-estimated equipment corresponding to the operation monitoring training data set based on the operation state recognition network. The operation monitoring terminal determines a loss value of an original operation monitoring model according to the estimated equipment operation state and the equipment operation state labeling information, if the loss value of the original operation monitoring model does not accord with the model training cut-off condition, if the loss value is smaller than a preset value, the operation monitoring terminal optimizes configuration variables of the original operation monitoring model according to the loss value which does not accord with the model training cut-off condition, the original operation monitoring model after optimizing the configuration variables (namely parameters of the model) is determined to be a temporary operation monitoring model, rolling optimization (namely repeated optimization) is carried out on the temporary operation monitoring model, and when the loss value of the temporary operation monitoring model after rolling optimization accords with the model training cut-off condition, the temporary operation monitoring model which accords with the model training cut-off condition is determined to be a target operation monitoring model.

In operation S209, a target operation monitoring data set of the sewage treatment apparatus to be monitored is acquired, and AI description knowledge corresponding to the target operation monitoring data set is determined based on the first data processing model.

The first data processing model is a target operational monitoring model corresponding to the target operational monitoring dataset, the target operational monitoring model further comprising a second data processing model and a third data processing model different from the first data processing model.

In operation S210, a matching representative feature list corresponding to the second data processing model is acquired, and matching description knowledge corresponding to the target operation monitoring dataset is determined according to the target operation monitoring dataset and the matching representative features in the matching representative feature list.

In operation S211, a state representative feature list corresponding to the third data processing model is acquired, and knowledge of the description of the involvement corresponding to the target operation monitoring dataset is determined according to the target operation monitoring dataset and the related data item representative features in the state representative feature list.

In operation S212, the AI description knowledge, the matching description knowledge and the involved description knowledge are subjected to knowledge interaction to obtain target interaction description knowledge of the target operation monitoring data set, the target interaction description knowledge is input into the operation state recognition network of the target operation monitoring model, and the operation state of the target device corresponding to the target operation monitoring data set is obtained based on the operation state recognition network.

In operation S213, it is identified whether the target device operation state is the first device operation state.

If the equipment operation state is the first equipment operation state, abnormal early warning can be carried out.

Based on the foregoing embodiments, the embodiments of the present application provide an operation monitoring device for a sewage treatment apparatus, where each unit included in the device and each module included in each unit may be implemented by a processor in a computer device; of course, the method can also be realized by a specific logic circuit; in practice, the processor may be a central processing unit (Central Processing Unit, CPU), microprocessor (Microprocessor Unit, MPU), digital signal processor (Digital Signal Processor, DSP) or field programmable gate array (Field Programmable Gate Array, FPGA), etc.

Fig. 2 is a schematic structural diagram of an operation monitoring device for a sewage treatment apparatus according to an embodiment of the present application, and as shown in fig. 2, the operation monitoring device 200 for a sewage treatment apparatus includes:

a monitoring data acquisition module 210, configured to acquire a target operation monitoring data set of a sewage treatment device to be monitored, and determine AI description knowledge corresponding to the target operation monitoring data set based on a first data processing model; the first data processing model is a target operation monitoring model corresponding to the target operation monitoring data set; the target operation monitoring model comprises a second data processing model and a third data processing model which are different from the first data processing model;

A matching knowledge determining module 220, configured to obtain a matching representative feature list corresponding to the second data processing model, and determine matching description knowledge corresponding to the target operation monitoring dataset according to the target operation monitoring dataset and matching representative features in the matching representative feature list;

the involvement knowledge determining module 230 is configured to obtain a state representative feature list corresponding to the third data processing model, and determine, according to the target operation monitoring dataset and the relevant data item representative feature in the state representative feature list, involvement description knowledge corresponding to the target operation monitoring dataset;

the operation state determining module 240 is configured to perform knowledge interaction on the AI description knowledge, the matching description knowledge, and the involved description knowledge to obtain target interaction description knowledge of the target operation monitoring dataset, input the target interaction description knowledge into an operation state recognition network of the target operation monitoring model, and obtain an operation state of a target device corresponding to the target operation monitoring dataset based on the operation state recognition network.

The description of the apparatus embodiments above is similar to that of the method embodiments above, with similar advantageous effects as the method embodiments. In some embodiments, functions or modules included in the apparatus provided in the embodiments of the present application may be used to perform the methods described in the embodiments of the methods, and for technical details that are not disclosed in the embodiments of the apparatus of the present application, please refer to the description of the embodiments of the methods of the present application for understanding.

It should be noted that, in the embodiment of the present application, if the above-mentioned sewage treatment apparatus operation monitoring method based on big data is implemented in the form of a software function module, and sold or used as a separate product, the method may also be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the embodiments of the present application may be essentially or portions contributing to the related art, and the software product may be stored in a storage medium, including several instructions to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to perform all or part of the methods described in the embodiments of the present application. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read Only Memory (ROM), a magnetic disk, an optical disk, or other various media capable of storing program codes. Thus, embodiments of the present application are not limited to any specific hardware, software, or firmware, or to any combination of hardware, software, and firmware.

The embodiment of the application provides an operation monitoring terminal, which comprises a memory and a processor, wherein the memory stores a computer program capable of running on the processor, and the processor realizes part or all of the steps in the method when executing the program.

Embodiments of the present application provide a computer readable storage medium having stored thereon a computer program which, when executed by a processor, performs some or all of the steps of the above-described method. The computer readable storage medium may be transitory or non-transitory.

Embodiments of the present application provide a computer program comprising computer readable code which, when run in a computer device, performs some or all of the steps for implementing the above method.

Embodiments of the present application provide a computer program product comprising a non-transitory computer-readable storage medium storing a computer program which, when read and executed by a computer, performs some or all of the steps of the above-described method. The computer program product may be realized in particular by means of hardware, software or a combination thereof. In some embodiments, the computer program product is embodied as a computer storage medium, in other embodiments the computer program product is embodied as a software product, such as a software development kit (Software Development Kit, SDK), or the like.

It should be noted here that: the above description of various embodiments is intended to emphasize the differences between the various embodiments, the same or similar features being referred to each other. The above description of apparatus, storage medium, computer program and computer program product embodiments is similar to that of method embodiments described above, with similar advantageous effects as the method embodiments. For technical details not disclosed in the embodiments of the apparatus, storage medium, computer program and computer program product of the present application, please refer to the description of the method embodiments of the present application.

Fig. 3 is a schematic hardware entity diagram of an operation monitoring terminal according to an embodiment of the present application, as shown in fig. 3, the hardware entity of the operation monitoring terminal 1000 includes: a processor 1001 and a memory 1002, wherein the memory 1002 stores a computer program executable on the processor 1001, the processor 1001 implementing the steps in the method of any of the embodiments described above when the program is executed.

The memory 1002 stores a computer program executable on a processor, and the memory 1002 is configured to store instructions and applications executable by the processor 1001, and may also cache data (e.g., image data, audio data, voice communication data, and video communication data) to be processed or already processed by each module in the processor 1001 and the operation monitoring terminal 1000, which may be implemented by a FLASH memory (FLASH) or a random access memory (Random Access Memory, RAM).

The processor 1001 performs the steps of the sewage treatment apparatus operation monitoring method based on big data of any one of the above-described steps when executing a program. Processor 1001 generally controls the overall operation of operation monitoring terminal 1000.

Embodiments of the present application provide a computer storage medium storing one or more programs executable by one or more processors to implement the steps of the big data based sewage treatment apparatus operation monitoring method of any of the embodiments above.

It should be noted here that: the description of the storage medium and apparatus embodiments above is similar to that of the method embodiments described above, with similar benefits as the method embodiments. For technical details not disclosed in the embodiments of the storage medium and the apparatus of the present application, please refer to the description of the method embodiments of the present application for understanding. The processor may be at least one of a target application integrated circuit (Application Specific Integrated Circuit, ASIC), a digital signal processor (Digital Signal Processor, DSP), a digital signal processing device (Digital Signal Processing Device, DSPD), a programmable logic device (Programmable Logic Device, PLD), a field programmable gate array (Field Programmable Gate Array, FPGA), a central processing unit (Central Processing Unit, CPU), a controller, a microcontroller, and a microprocessor. It will be appreciated that the electronic device implementing the above-mentioned processor function may be other, and embodiments of the present application are not specifically limited.

The computer storage medium/Memory may be a Read Only Memory (ROM), a programmable Read Only Memory (Programmable Read-Only Memory, PROM), an erasable programmable Read Only Memory (Erasable Programmable Read-Only Memory, EPROM), an electrically erasable programmable Read Only Memory (Electrically Erasable Programmable Read-Only Memory, EEPROM), a magnetic random access Memory (Ferromagnetic Random Access Memory, FRAM), a Flash Memory (Flash Memory), a magnetic surface Memory, an optical disk, or a Read Only optical disk (Compact Disc Read-Only Memory, CD-ROM); but may also be various terminals such as mobile phones, computers, tablet devices, personal digital assistants, etc., that include one or any combination of the above-mentioned memories.

It should be appreciated that reference throughout this specification to "one embodiment" or "an embodiment" means that a particular feature, structure or characteristic described in connection with the embodiment is included in at least one embodiment of the present application. Thus, the appearances of the phrases "in one embodiment" or "in an embodiment" in various places throughout this specification are not necessarily all referring to the same embodiment. Furthermore, the particular features, structures, or characteristics may be combined in any suitable manner in one or more embodiments. It should be understood that, in various embodiments of the present application, the sequence number of each step/process described above does not mean that the execution sequence of each step/process should be determined by the function and the internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application. The foregoing embodiment numbers of the present application are merely for describing, and do not represent advantages or disadvantages of the embodiments. It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.

In the several embodiments provided in this application, it should be understood that the disclosed apparatus and method may be implemented in other ways. The above described device embodiments are only illustrative, e.g. the division of the units is only one logical function division, and there may be other divisions in practice, such as: multiple units or components may be combined or may be integrated into another system, or some features may be omitted, or not performed. In addition, the various components shown or discussed may be coupled or directly coupled or communicatively coupled to each other via some interface, whether indirectly coupled or communicatively coupled to devices or units, whether electrically, mechanically, or otherwise.

The units described above as separate components may or may not be physically separate, and components shown as units may or may not be physical units; can be located in one place or distributed to a plurality of network units; some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

In addition, each functional unit in each embodiment of the present application may be integrated in one processing unit, or each unit may be separately used as one unit, or two or more units may be integrated in one unit; the integrated units may be implemented in hardware or in hardware plus software functional units.

Those of ordinary skill in the art will appreciate that: all or part of the steps for implementing the above method embodiments may be implemented by hardware related to program instructions, and the foregoing program may be stored in a computer readable storage medium, where the program, when executed, performs steps including the above method embodiments; and the aforementioned storage medium includes: a mobile storage device, a Read Only Memory (ROM), a magnetic disk or an optical disk, or the like, which can store program codes.

Alternatively, the integrated units described above may be stored in a computer readable storage medium if implemented in the form of software functional modules and sold or used as a stand-alone product. Based on such understanding, the technical solution of the present application may be embodied essentially or in a part contributing to the related art in the form of a software product stored in a storage medium, including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to perform all or part of the methods described in the embodiments of the present application. And the aforementioned storage medium includes: various media capable of storing program codes, such as a removable storage device, a ROM, a magnetic disk, or an optical disk.

The foregoing is merely an embodiment of the present application, but the protection scope of the present application is not limited thereto, and any person skilled in the art can easily think about changes or substitutions within the technical scope of the present application, and the changes and substitutions are intended to be covered in the protection scope of the present application.

Claims

1. A sewage treatment device operation monitoring method based on big data, which is characterized by being applied to an operation monitoring terminal, the method comprising:

2. The method of claim 1, wherein the acquiring the target operation monitoring dataset of the sewage treatment plant to be monitored, determining AI description knowledge corresponding to the target operation monitoring dataset based on the first data processing model, comprises:

3. The method of claim 2, wherein the first data processing model comprises a target knowledge refinement network; the target knowledge extraction network comprises an integrated projection focusing module, a first unified arrangement module, a multi-layer perception module and a second unified arrangement module; inputting the AI description knowledge to be refined into a first data processing model in a target operation monitoring model, performing knowledge refinement on the AI description knowledge to be refined based on the first data processing model to obtain data set refinement description knowledge corresponding to the decomposition data item, and determining the AI description knowledge corresponding to the target operation monitoring data set according to the data set refinement description knowledge corresponding to the decomposition data item, wherein the determining comprises the following steps:

4. The method of claim 3, wherein the integrated projection focusing module comprises a target self-directed saliency determination module, a first summation module, a knowledge interaction module, and a second summation module corresponding to the target self-directed saliency determination module; the knowledge interaction module is used for carrying out knowledge interaction on the descriptive knowledge obtained based on each self-oriented significance determination module in the integrated projection focusing module; one self-directed saliency determination module corresponds to one first sum module; the inputting the AI description knowledge to be refined into the integrated projection focusing module in the first data processing model of the target operation monitoring model, extracting the AI description knowledge to be refined based on the integrated projection focusing module, and obtaining a first depth description knowledge corresponding to the AI description knowledge to be refined, including:

5. The method of claim 1, wherein the obtaining a list of matching representative features corresponding to the second data processing model, determining matching descriptive knowledge corresponding to the target operational monitoring dataset based on the target operational monitoring dataset and matching representative features in the list of matching representative features, comprises:

6. The method of claim 1, wherein the obtaining a list of state representative features corresponding to the third data processing model, determining the knowledge of the involvement description corresponding to the target operational monitoring dataset based on the target operational monitoring dataset and the relevant data item representative features in the list of state representative features, comprises:

7. The method according to claim 1, wherein the method further comprises:

8. The method of claim 7, wherein the performing rolling optimization on the original operation monitoring model according to the training description knowledge, the training matching description knowledge, the training involvement description knowledge, the equipment operation state labeling information, and the operation state recognition network of the original operation monitoring model, determining the rolling optimized original operation monitoring model as a target operation monitoring model comprises:

9. The method of claim 7, wherein the method further comprises:

10. An operation monitoring system for a sewage treatment plant, comprising a data acquisition device and an operation monitoring terminal in communication with each other, said operation monitoring terminal comprising a memory and a processor, said memory storing a computer program executable on the processor, said processor executing said program to carry out the steps of the method according to any one of claims 1 to 9.