Underground steam pipe network intelligent temperature loss early warning and monitoring device based on ad hoc network

Abstract

The invention discloses an underground steam pipe network intelligent temperature loss early warning and monitoring device based on an ad hoc network, and relates to the technical field of pipe network monitoring, and the method comprises the steps: determining a protection pipeline diagram, and building a temperature loss data collection system; obtaining temperature loss early warning historical data, and extracting text feature vectors and structured fields to establish a temperature loss word list; establishing an alarm type classification and identification model based on the temperature loss word list, obtaining an alarm type sequence, associating temperature loss reasons, and evaluating hazard levels; establishing an association graph; and determining an alarm type, inputting the association graph to obtain a temperature loss reason optimal solution, and performing early warning based on the optimal solution and the key node position. The technical problems that an existing underground steam pipe network temperature loss early warning system is single in feature information, high in false alarm rate and insufficient in alarm accuracy are solved, and the purposes of improving the accuracy and reliability of underground steam pipe network temperature loss early warning, accurately finding potential temperature loss problems, reducing false alarms, rapidly positioning temperature loss reasons and improving the early warning efficiency are achieved. And safe and stable operation of the pipe network is ensured.

Description

Technical Field

[0001] The present application relates to the technical field of pipe network monitoring, and in particular to an intelligent temperature loss early warning monitoring device for underground steam pipe networks based on a self-organizing network. Background Art

[0002] In the operation scenario of underground steam pipelines, the temperature loss problem is particularly prominent, and the contradiction between the demand for efficient and accurate temperature loss early warning monitoring is also more prominent. Realizing intelligent temperature loss early warning monitoring of underground steam pipelines has become a crucial link in ensuring the safe operation of the pipeline network. Traditional underground steam pipeline monitoring is often limited and incomplete, mainly relying on manual inspections and limited sensor data, with insufficient coverage of key nodes in the pipeline network, lack of utilization and in-depth analysis of unstructured data, and difficulty in comprehensively and accurately judging the temperature loss situation. In terms of temperature loss early warning, the accuracy and reliability are insufficient, there is a high false alarm rate, and the determination of the cause of temperature loss is relatively simple, which cannot cope well with the complex and changeable actual operation of the pipeline network.

[0003] At present, relevant technologies still have the technical problems of single characteristic information, high false alarm rate and insufficient alarm accuracy in the temperature loss early warning system of underground steam pipeline network. Summary of the invention

[0004] The present application provides an intelligent temperature loss early warning monitoring device for underground steam pipelines based on a self-organizing network, determines the underground steam pipeline protection pipeline diagram, analyzes the key node positions and establishes a temperature loss data acquisition system; obtains temperature loss early warning historical data, extracts text feature vectors and structured fields to establish a temperature loss vocabulary; establishes an alarm type classification and recognition model based on the temperature loss vocabulary, obtains the alarm type sequence and associates the temperature loss cause and evaluates the hazard level; calculates the alarm type similarity to establish an association graph; analyzes the alarm information of key nodes to determine the alarm type, inputs the association graph to obtain the optimal solution for the temperature loss cause, and warns based on the optimal solution and the key node position, thereby achieving the technical effect of improving the accuracy and reliability of underground steam pipeline temperature loss early warning, accurately discovering potential temperature loss problems, reducing false alarms and quickly locating the cause of temperature loss, and ensuring the safe and stable operation of the pipeline network by introducing unstructured data analysis, damage process time characteristics and transaction logic chain data filtering, and multi-node multi-dimensional alarm prediction mutual verification.

[0005] The present application provides an intelligent temperature loss early warning monitoring device for underground steam pipe network based on a self-organizing network, including: A temperature loss data acquisition system establishment module, the temperature loss data acquisition system establishment module is used to determine the protection pipeline diagram of the underground steam pipeline network, analyze the key node positions of the protection pipeline diagram, and establish a temperature loss data acquisition system for key nodes of the pipeline; a temperature loss vocabulary building module, the temperature loss vocabulary building module is used to obtain temperature loss warning historical data, obtain text feature vectors and structured fields based on the temperature loss warning historical data, the text feature vectors are obtained through unstructured fields, the text feature vectors and the structured fields are analyzed, and a finite set temperature loss vocabulary is established; a classification recognition model building module, the classification recognition model building module is used to establish a self-supervised alarm type classification recognition model based on the finite set temperature loss vocabulary, obtain an alarm type sequence, and associate the each alarm type in the alarm type sequence and multiple temperature loss causes, evaluate the hazard level of each alarm type; a temperature loss cause association graph construction module, the temperature loss cause association graph construction module is used to obtain the alarm type sequence, the hazard level, the multiple temperature loss causes, calculate the alarm type similarity, and establish an alarm type temperature loss cause association graph; a temperature loss cause optimal solution acquisition module, the temperature loss cause optimal solution acquisition module is used to analyze the reported alarm information of multiple key nodes based on the self-supervised alarm type classification and recognition model, determine multiple alarm types, input the multiple alarm types into the alarm type temperature loss cause association graph, and obtain the optimal solution for the temperature loss cause; an early warning module, the early warning module is used to issue an early warning based on the optimal solution for the temperature loss cause and the position of the key node.

[0006] The intelligent temperature loss early warning and monitoring device for underground steam pipeline network based on self-organizing network proposed in this application first determines the protection pipeline diagram of underground steam pipeline network, analyzes the key node positions and establishes a temperature loss data acquisition system; obtains the historical data of temperature loss early warning, extracts text feature vectors and structured fields to establish a temperature loss vocabulary; establishes an alarm type classification and recognition model based on the temperature loss vocabulary, obtains the alarm type sequence and associates the temperature loss cause and evaluates the hazard level; calculates the alarm type similarity to establish an association graph; analyzes the alarm information of key nodes to determine the alarm type, inputs the association graph to obtain the optimal solution for the temperature loss cause, and warns based on the optimal solution and the key node position. By introducing unstructured data analysis, damage process time characteristics and transaction logic chain data filtering, and multi-node and multi-dimensional alarm prediction mutual verification, the accuracy and reliability of underground steam pipeline temperature loss early warning are improved, potential temperature loss problems are accurately discovered, false alarms are reduced, the causes of temperature loss are quickly located, and the technical effect of ensuring safe and stable operation of the pipeline network is achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] In order to more clearly illustrate the technical solution of the embodiment of the present invention, the accompanying drawings of the embodiment of the present invention will be briefly introduced below. A flow chart is used in the present application to illustrate the operations performed by the system according to the embodiment of the present application. It should be understood that the preceding or following operations are not necessarily performed accurately in order. On the contrary, various steps can be processed in reverse order or simultaneously as needed. At the same time, other operations can also be added to these processes, or one or more operations can be removed from these processes.

[0008] Figure 1 A schematic diagram of the structure of an intelligent temperature loss early warning monitoring device for underground steam pipe network based on a self-organizing network provided in an embodiment of the present application; Figure 2 A schematic structural diagram of a temperature loss cause association diagram building module 40 of an ad hoc network-based underground steam pipe network intelligent temperature loss early warning monitoring device provided in an embodiment of the present application.

[0009] Explanation of the accompanying drawings: temperature loss data acquisition system establishment module 10, temperature loss vocabulary construction module 20, classification and recognition model construction module 30, temperature loss cause association diagram construction module 40, temperature loss cause optimal solution acquisition module 50, early warning module 60. DETAILED DESCRIPTION

[0010] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below.

[0011] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings. The described embodiments should not be regarded as limiting the present application. All other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of this application.

[0012] In the following description, reference is made to "some embodiments", which describe a subset of all possible embodiments, but it is understood that "some embodiments" may be the same subset or different subsets of all possible embodiments, and may be combined with each other without conflict, and the terms "first\second" involved are merely to distinguish similar objects and do not represent a specific ordering of objects. The terms "including" and "having" and any variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or modules that are not clearly listed or inherent to these processes, methods, products, or devices. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those generally understood by technicians in the technical field of this application. The terms used herein are for the purpose of describing the embodiments of the present application only.

[0013] The embodiment of the present application provides an intelligent temperature loss early warning monitoring device for underground steam pipe network based on a self-organizing network, such as Figure 1 As shown, the device comprises: A temperature loss data acquisition system establishment module 10 is used to determine the protection pipeline diagram of the underground steam pipeline network, analyze the key node positions of the protection pipeline diagram, and establish a temperature loss data acquisition system for the key nodes of the pipeline. Specifically, the temperature loss data acquisition system establishment module needs to collect relevant information about the underground steam pipeline network and conduct field surveys, then draw a protection pipeline diagram, clarify the key node types, analyze pipeline characteristics and consider environmental factors in combination with the protection pipeline diagram, so as to determine the key node positions, select suitable sensors to be installed at the key nodes, and establish a data transmission network to transmit the collected data in real time to the data processing center for processing and storage, so as to provide an accurate data basis for the intelligent temperature loss early warning monitoring of the underground steam pipeline network.

[0014] In a possible implementation, the temperature loss data acquisition system establishment module 10 further includes: a data acquisition unit, which is used to collect real-time data and historical data of the key nodes. Specifically, the key node positions in the underground steam pipe network are determined. The key nodes are usually the parts where temperature loss problems are prone to occur, such as pipe connections, elbows, etc., and corresponding data acquisition equipment, such as temperature sensors, pressure sensors, etc., are installed to monitor various parameters of the key nodes in real time, and collect historical data of the key nodes from a database or storage device. The historical data may include temperature and pressure changes over a period of time in the past, as well as records of temperature loss events that have occurred.

[0015] A data filtering unit, the data filtering unit is used to obtain the time characteristics of the damage process of the key node based on the historical data of the key node, and filter the real-time data based on the time characteristics of the damage process. Specifically, the historical data of the key nodes are analyzed to understand the time pattern and development process of the temperature loss. For example, the changing trend of the temperature loss within a period of time after the occurrence and the time range in which the signs of temperature loss are shown are determined. The real-time collected data is filtered according to the time characteristics of the damage process obtained from the historical data. If the currently collected data reflects a drastic change in a short period of time, and it is judged based on the historical data that this situation is inconsistent with the normal development process of temperature loss, the system will regard it as abnormal data and filter it. Some abnormal data caused by sudden interference or erroneous measurement can be excluded, thereby improving the accuracy and reliability of the data.

[0016] A transaction logic chain construction unit, the transaction logic chain construction unit is used to construct a transaction logic chain of temperature loss and damage of the underground steam pipeline network, and filter the real-time data based on the transaction logic chain. Specifically, the construction of the transaction logic chain of temperature loss and damage of the underground steam pipeline network includes collecting the path information of steam flow, the operating status data of each key node, and maintenance records. The path of steam flow can help determine the direction in which the temperature loss may spread and the scope of influence; the operating status data of each key node can reflect the overall operation of the pipeline network; the maintenance records can provide information about the past maintenance of the pipeline network and potential problems that may exist. The real-time collected data is compared with the transaction logic chain. If a key node suddenly has abnormal temperature loss without any operation, and according to the transaction logic chain, it is judged that this situation does not conform to the normal operation logic, the system can judge that this data may be a false alarm, and then filter it. In this way, the consistency and coherence of the data can be ensured, and the quality of the data can be improved.

[0017] A real-time data output unit is used to output the real-time data of the key nodes. Specifically, after data collection, filtering, and comparison with the transaction logic chain, the screened real-time data of the key nodes is output, and the data can be transmitted to the monitoring center, data analysis system or other related equipment for further processing and analysis. The output data can be presented in the form of charts, reports, etc., which is convenient for management personnel to timely understand the operating status and temperature loss of the underground steam pipeline network, and provide a basis for taking corresponding preventive and maintenance measures.

[0018] In a possible implementation, the transaction logic chain construction unit further includes: a structural model construction subunit, and the structural model construction subunit is used to construct the structural model of the underground steam pipe network. Specifically, various information of the underground steam pipe network is collected to construct the structural model, including determining the direction of the pipeline through engineering drawings, field surveys, etc., clarifying the location distribution of each node in space, recording the pipe diameter size and material type used in different parts, and the connection method between pipes, such as welding, flange connection, etc., collating the historical maintenance records of each node, including the time, reason, measures taken, etc. of maintenance, integrating the information, and forming a structural model that comprehensively describes the physical structure and maintenance history of the underground steam pipe network.

[0019] An event association subunit is used to obtain the temperature loss characteristics of the underground steam pipeline network and the corresponding key nodes and times, analyze the key nodes and times to associate the temperature loss characteristics with actual temperature loss events, and construct a multi-stage transaction logic chain. Specifically, the temperature damage characteristics that may occur in the underground steam pipeline network are determined, such as pipeline aging, which is manifested by brittleness of pipeline materials and color changes; temperature and pressure fluctuations may cause uneven expansion and contraction of the pipeline; in terms of environmental factors, changes in soil moisture may cause pipeline corrosion, and geological activities may cause pipeline deformation; improper maintenance may result in poor sealing, installation errors, etc. When a temperature loss event is detected, the corresponding key nodes and time are determined, the location and characteristics of the key nodes and various factors at the time nodes are analyzed, and the temperature loss damage characteristics are associated with the actual temperature loss event. Based on the association relationship, a multi-stage transaction logic chain is constructed. For example, if the soil moisture in a certain area is high for a long time, the pipeline at the corresponding node may first show signs of corrosion, manifested as rust on the pipeline surface. This is the initial stage of temperature loss; over time, corrosion worsens and the pipeline wall becomes thinner, which may lead to increased temperature fluctuations and enter the mid-term stage; if it continues to develop, fine cracks may appear in the pipeline, and the pressure begins to be abnormal, further deteriorating; ultimately, the crack expansion causes the pipeline to rupture, with reduced flow or increased leakage, which is the final damage stage.

[0020] A logic chain adjustment subunit, the logic chain adjustment subunit is used to dynamically adjust the multi-stage transaction logic chain based on the real-time data and historical data. Specifically, the real-time data of the underground steam pipeline network is continuously collected, including the changes in parameters such as temperature, pressure, and flow, and the development process and processing results of similar temperature loss events in the historical data are referred to. If the real-time data shows that the development of temperature loss does not conform to the preset transaction logic chain, for example, the performance characteristics of a certain stage appear too early or too late, or new abnormal situations occur, it is necessary to dynamically adjust the multi-stage transaction logic chain. The adjustment includes modifying the performance characteristic description of a certain stage, adjusting the transition conditions between stages, adding or deleting certain stages, etc. By continuously adjusting according to real-time data and historical data, the transaction logic chain can more accurately reflect the actual development of temperature loss in the underground steam pipeline network, and improve the system's ability to judge and predict temperature loss.

[0021] The temperature loss vocabulary building module 20 is used to obtain temperature loss warning historical data, obtain text feature vectors and structured fields based on the temperature loss warning historical data, and the text feature vectors are obtained through unstructured fields. The text feature vectors and the structured fields are analyzed to establish a finite set temperature loss vocabulary. Specifically, the scope of the temperature loss warning historical data is clarified. The data not only includes specific data when the temperature loss occurs, such as abnormal temperature change values, pressure fluctuations, etc., but also includes the processing flow of the corresponding temperature loss event. It is collected from the monitoring system database, maintenance record archives and related management information systems of the underground steam pipeline network. The collected historical data is preliminarily sorted and classified for subsequent analysis. For example, it is classified according to factors such as the time, location, and severity of the temperature loss, to prepare for the subsequent extraction of feature vectors and structured fields. For the unstructured fields in the temperature loss warning historical data, such as descriptive text in the processing flow, notes in the maintenance records, etc., The word / sentence vector embedding method is used to extract text feature vectors, convert natural language text into feature vectors of fixed length, which can effectively represent the meaning of the text, convert words in unstructured fields into vector representations, and then obtain the feature vectors of sentences by combining and processing the word vectors in sentences. Effective information is extracted from unstructured fields such as alarm loads, and new valuable features are added for alarm evaluation. Structured fields are extracted from temperature loss warning historical data, such as clear data fields such as the time, location, and degree of temperature loss. Structured fields can be directly used for data analysis and model building. The acquired text feature vectors and structured fields are comprehensively analyzed. Comprehensive analysis, using statistical analysis, cluster analysis and other methods to find out the key features and patterns related to temperature loss. For example, by calculating the similarity between text feature vectors, similar texts are classified into one category, so as to find the potential connection between different temperature loss events. Combined with the time, place and other information in the structured field, the occurrence pattern and trend of temperature loss events are analyzed. In the analysis process, machine learning algorithms such as principal component analysis can also be introduced to reduce the dimension of feature vectors, remove redundant information, and improve analysis efficiency and accuracy. According to the analysis results, a finite set temperature loss vocabulary is established, which contains keywords and phrases related to temperature loss, as well as their corresponding feature vectors and structures. Structured field information, for example, the vocabulary may include keywords such as "pipeline rupture", "temperature anomaly", "pressure fluctuation", and their corresponding text feature vectors and structured fields, such as time, location, severity, etc. After the vocabulary is established, it can be continuously optimized and updated. With the occurrence and processing of new temperature loss events, new keywords and feature information are continuously added to improve the accuracy and completeness of the vocabulary. The temperature loss vocabulary construction module can make full use of the unstructured fields and structured fields in the temperature loss warning historical data, extract valuable feature information, and establish a finite set temperature loss vocabulary to provide strong support for intelligent temperature loss warning monitoring of underground steam pipelines.

[0022] In a possible implementation, the warm loss vocabulary building module 20 further includes: an unstructured field acquisition unit, which is used to acquire unstructured fields based on the warm loss warning historical data. Specifically, a comprehensive review of the warm loss warning historical data is carried out. The historical data contains both structured data, such as clear numerical information like time, location, and warm loss degree, and unstructured descriptive text content, such as the text description in a fault report and the remarks in a maintenance record. Unstructured fields are screened out from the historical data. The fields are usually in the form of natural language text, containing rich detailed information about warm loss events, but lacking a clear structure and format.

[0023] A text feature vector acquisition unit, which is used to process the unstructured fields based on natural language processing technology to obtain the text feature vectors. Specifically, natural language processing technology is used to process the unstructured fields. First, text cleaning is performed to remove some irrelevant punctuation marks, special characters, etc. Word embedding techniques, such as Word2Vec and GloVe, are adopted to convert the words in the unstructured fields into vector representations. The vectors can capture the semantic information of the words, and in the vector space, words with similar semantics will have similar vector representations. For sentence-level unstructured fields, the feature vectors of the sentences can be obtained by combining and processing the word vectors in the sentences. For example, methods such as average pooling and max pooling are used to combine the word vectors in the sentence into a fixed-length vector as the feature vector representation of the sentence.

[0024] A vector preprocessing unit, which is used to preprocess the text feature vectors. Specifically, operations for removing noise words are carried out. Common stop words, such as "de", "shi", "zai", etc., have a high frequency of occurrence in the text but contribute less to expressing the core meaning of the text. They are removed from the text feature vectors to reduce noise interference. Operations for unifying the word forms are carried out, including stemming and lemmatization. Stemming is to remove the affixes of the words to obtain the stem part of the words, and lemmatization is to restore the words to their basic forms to ensure that words in different forms can be correctly recognized and processed. Standardization processing is carried out, such as logarithmic scaling and normalization. Logarithmic scaling can reduce the values of the feature vectors with large numerical values to a certain range to avoid the influence of overly large numerical values on subsequent analysis. Normalization can normalize the values of the feature vectors to a specific interval, such as [0,1] or [-1,1], so that different feature vectors are comparable.

[0025] A cluster analysis unit is used to perform cluster analysis on the preprocessed text feature vectors to obtain multiple groups of text feature vectors. Specifically, a cluster analysis algorithm is used to group the preprocessed text feature vectors. In the clustering process, the text feature vectors are grouped according to the semantic similarity of the words, word frequency co-occurrence and other information. Feature vectors with high semantic similarity will be classified into one category, that is, if two words are semantically similar, the distance between their corresponding feature vectors in the vector space will also be close; word frequency co-occurrence means that if two words often appear in the same text context, they may also have similar semantics and thus be classified into one category, and a preliminary label is given to each category to facilitate the subsequent identification and analysis of different vector groups. For example, the labels can be named according to the words or topics that appear more frequently in the group of vectors.

[0026] A vector grouping unit, the vector grouping unit is used to establish a finite set temperature loss vocabulary based on the grouping of the multiple sets of text feature vectors. Specifically, based on the results of the grouping of the multiple sets of text feature vectors, a finite set temperature loss vocabulary is established, the vocabulary contains various temperature loss related keywords and their corresponding feature vectors, the keywords are preliminarily screened, and represent the common types of temperature loss in underground steam pipe networks and their performance characteristics, for example, including keywords such as "pipeline rupture", "temperature anomaly", "pressure fluctuation", and their corresponding feature vectors after cluster analysis, in order to enhance the accuracy of the vocabulary, expert knowledge or existing literature can also be used to proofread and correct the keywords in the vocabulary, experts can evaluate the accuracy and representativeness of the keywords based on their experience and expertise, and make modification suggestions; relevant research results in existing literature can also be used as a reference to further improve the content of the vocabulary.

[0027] The classification and recognition model construction module 30 is used to establish a self-supervised alarm type classification and recognition model based on the finite set temperature loss vocabulary, obtain an alarm type sequence, associate each alarm type in the alarm type sequence with multiple temperature loss causes, and evaluate the hazard level of each alarm type. Specifically, the content and structure of the finite set temperature loss vocabulary are clarified, including various temperature loss-related keywords and their corresponding feature vectors, representing the common types of temperature loss in underground steam pipelines and their performance characteristics, providing basic data for establishing a self-supervised alarm type classification and recognition model, further organizing and analyzing the finite set temperature loss vocabulary, understanding the meaning and characteristics of the temperature loss represented by different keywords, considering the frequency and importance of keywords in actual temperature loss events, preparing for the establishment and training of subsequent models, selecting appropriate machine learning algorithms and model architectures to establish a self-supervised alarm type classification and recognition model, and using neural network models in deep learning. According to the characteristic vectors in the temperature loss vocabulary and the characteristics of the alarm type, the input layer, hidden layer and output layer of the model are designed. The input layer receives the keyword characteristic vector in the temperature loss vocabulary, the hidden layer abstracts and extracts the features through a multi-layer neural network structure, and the output layer outputs the classification result of the alarm type. In the process of model establishment, the self-supervised learning method can be used, that is, learning with unlabeled data. For example, the keywords in the temperature loss vocabulary can be randomly masked or replaced, and the model can predict the masked or replaced parts, so as to learn the semantic relationship between the keywords and the characteristics of the alarm type, and input the actual temperature loss data into the established self-supervised alarm. In the type classification recognition model, the data can be the temperature loss data monitored in real time or part of the historical temperature loss data. The model processes and analyzes the input data and outputs an alarm type sequence. The sequence represents the alarm types at different time points or different key nodes. For example, different alarm types such as "temperature abnormality alarm", "pressure fluctuation alarm" and "pipeline leakage alarm" are arranged in chronological order to form an alarm type sequence. Different alarm types are analyzed in depth to determine the various temperature loss causes that may correspond to each alarm type. Through the analysis of historical temperature loss data, the introduction of expert knowledge and the understanding of the physical structure and operation principle of the underground steam pipeline network, the model can be used to identify the temperature loss data of the underground steam pipeline network. For example, if an "abnormal temperature alarm" occurs, possible causes of temperature loss include damage to the pipeline insulation layer, abnormal steam flow, changes in external ambient temperature, etc. If a "pressure fluctuation alarm" occurs, possible causes include pipeline blockage, valve failure, steam generator instability, etc. Each alarm type is associated with multiple temperature loss causes, and a mapping relationship table is established so that in actual applications, the possible cause of temperature loss can be quickly determined according to the alarm type. The hazard level of each alarm type is evaluated based on factors such as the severity of the temperature loss, the impact on the operation of the underground steam pipeline network, and the possible consequences. The evaluation is carried out using a combination of quantitative and qualitative methods.Quantitative assessment can consider specific indicators such as the temperature change amplitude, pressure change value, leakage volume caused by temperature loss; qualitative assessment can consider the impact of temperature loss on the safety, reliability, economy and other aspects of the pipeline network, and divide the hazard level of the alarm type into different levels, such as high, medium and low. In practical applications, the priority of processing can be determined according to the hazard level of the alarm type, and the alarm type with a high hazard level can be processed first to minimize the impact of temperature loss on the underground steam pipeline network.

[0028] In a possible implementation, the classification and recognition model building module 30 further includes: a structured field acquisition unit, wherein the structured field is used to acquire a structured field based on the temperature loss warning historical data. Specifically, the temperature loss warning historical data is sorted and analyzed. The historical data contains various information, including both unstructured text descriptions and structured data fields. Structured fields are extracted from the historical data. The fields are usually data with clear formats and meanings, such as the time, location, and specific values ​​of the temperature loss degree. The structured fields can be directly used for data analysis and model building to provide basic data for subsequent steps.

[0029] An initial self-supervised alarm type classification and recognition model construction unit is used to establish an initial self-supervised alarm type classification and recognition model. Specifically, a suitable machine learning algorithm and model architecture are selected to construct an initial self-supervised alarm type classification and recognition model, and a neural network model in deep learning is considered, such as a convolutional neural network (CNN), a recurrent neural network (RNN), or a long short-term memory network (LSTM). According to the task requirements of alarm type classification and recognition, the input layer, hidden layer, and output layer of the model are designed. The input layer can receive the text comprehensive feature vector generated in the subsequent steps, and the hidden layer abstracts and extracts these features through a multi-layer neural network structure. The output layer outputs the classification result of the alarm type.

[0030] A text comprehensive feature vector acquisition unit is used to acquire a text comprehensive feature vector based on the structured field and the text feature vector. Specifically, the previously acquired structured field and text feature vector are fused, and a variety of methods are used for fusion, such as inputting the structured field into the model as an additional feature, and processing it together with the text feature vector through a neural network, and splicing the structured field and the text feature vector into a longer feature vector by feature splicing. Through the fusion operation, a text comprehensive feature vector is obtained. The feature vector contains both the clear information of the structured data and the semantic features of the unstructured text, which can more comprehensively describe the characteristics of the temperature loss alarm and provide richer information for the classification and identification of the alarm type.

[0031] A model training unit is used to construct positive sample pairs and negative sample pairs based on the text comprehensive feature vector, input the initial self-supervised alarm type classification and recognition model, perform training, and obtain the self-supervised alarm type classification and recognition model. Specifically, positive sample pairs and negative sample pairs are constructed based on the comprehensive feature vector of the text. According to the method of self-supervised learning, virtual alarm pairs can be generated through data enhancement technology. For example, some temperature loss alarm data are subjected to data masking, rotation, cropping and other operations to generate a set of enhanced data of similar alarms, and they are marked as positive sample pairs, while irrelevant alarm pairs are marked as negative sample pairs, such as alarm pairs from different time and place or with different temperature loss characteristics. The positive sample pairs and negative sample pairs are input into the initial self-supervised alarm type classification and recognition model for training. The model continuously adjusts the parameters so that the similarity score of the positive sample pairs is higher than that of the negative sample pairs. During the training process, the model learns the intrinsic characteristics of the alarm type and gradually improves the classification accuracy of the alarm type. After multiple iterative training, when the performance of the model on the verification set reaches a certain standard, the training is stopped to obtain the final self-supervised alarm type classification and recognition model. The model can be used to accurately classify and identify the alarm types of new temperature loss alarm data.

[0032] The temperature loss cause association diagram construction module 40 is used to obtain the alarm type sequence, the hazard level, and the multiple temperature loss causes, calculate the alarm type similarity, and establish an alarm type temperature loss cause association diagram. Specifically, the alarm type sequence, the hazard level of each alarm type, and the various temperature loss causes related to it are collected from the previous modules. The alarm type sequence is obtained through the classification and recognition model construction module, which records the alarm types that appear at different time points or different key nodes. The hazard level is determined after evaluating the degree of impact caused by each alarm type. The various temperature loss causes are obtained by in-depth analysis of the alarm type, associating historical data and professional knowledge to obtain the possible causes of the alarm type. The cosine similarity calculation method is used to evaluate the similarity between different alarm types. For example, for alarm types based on text features, their similarity can be determined by calculating the cosine similarity between text feature vectors. If the text descriptions of two alarm types are semantically close, then their similarity will be higher. In addition to text features, other factors can also be combined, such as the time interval between alarms, the positional relationship of key nodes, etc., to comprehensively calculate the similarity of alarm types. If two alarm types are close in time to each other, their similarity will be higher. If the alarm types are generated or involve adjacent key nodes, they may also have certain similarities. According to the calculated alarm type similarity and the obtained hazard level and various temperature loss causes, an alarm type temperature loss cause association diagram is constructed. The association diagram can use a graphical method to display different alarm types, temperature loss causes and the relationship between them. The nodes in the association diagram can represent the alarm type and the temperature loss cause. For the alarm type node, its name, hazard level and other information can be marked; for the temperature loss cause node, the specific temperature loss situation can be described in detail, such as pipeline aging, temperature and pressure fluctuations, etc. The edge represents the association relationship between the alarm type and the temperature loss cause. If an alarm type is highly likely to be associated with a certain temperature loss cause, then they are connected by an edge in the association diagram, and different weights can be assigned to the edge according to the size of the similarity to indicate the strength of the association. By establishing the association diagram, the relationship between different alarm types and temperature loss causes can be intuitively seen, providing strong support for quickly determining the temperature loss cause and taking corresponding treatment measures.

[0033] In one possible implementation, Figure 2As shown, the temperature loss cause association graph construction module 40 further includes: an association graph structure acquisition unit, the association graph structure acquisition unit is used to obtain the preliminary association relationship between the alarm type and the temperature loss cause based on the alarm type sequence and the temperature loss cause, and establish the basic structure of the alarm type temperature loss cause association graph based on the preliminary association relationship. Specifically, analyze the alarm type sequence, understand the various alarm types that appear at different time points, sort out the temperature loss causes, and clarify the various factors that may cause temperature loss. Through the analysis of historical data, the introduction of expert knowledge, and logical reasoning, the preliminary association relationship between the alarm type and the temperature loss cause is established. For example, if a certain alarm type appears multiple times in the historical data with a specific temperature loss cause, it can be preliminarily considered that there is an association between them. Based on the preliminary association relationship, the basic structure of the alarm type temperature loss cause association graph is constructed. The basic structure can be represented in a graphical manner, in which the alarm type and the temperature loss cause are used as nodes, and the preliminary association relationship is used as an edge connecting the nodes.

[0034] A weight adjustment unit, the weight adjustment unit is used to dynamically adjust the weight value of the preliminary association relationship based on the similarity of the alarm types. Specifically, the similarity between different alarm types is calculated, such as using algorithms such as cosine similarity, and their similarity is measured based on information such as the characteristic vector of the alarm type. According to the similarity of the alarm types, the weight value of the preliminary association relationship is dynamically adjusted. If the similarity of two alarm types is high, and they are initially associated with the same temperature loss cause, then the weight of the association relationship between this temperature loss cause and the two alarm types can be appropriately increased. Conversely, if the similarity of the alarm types is low, the weight of the corresponding association relationship may be reduced, so that the association diagram more accurately reflects the actual association strength between different alarm types and temperature loss causes.

[0035] A multi-level association relationship construction unit, the multi-level association relationship construction unit is used to construct a multi-level association relationship based on the association type and the preliminary association relationship. Specifically, the preliminary association relationship is further analyzed to determine different association types. The association type may include direct association, indirect association, etc. For example, some alarm types may be directly caused by specific temperature loss causes, which is a direct association; while some alarm types may be associated with the temperature loss cause through an intermediate link, which is an indirect association. According to the association type and the preliminary association relationship, a multi-level association relationship is constructed. In the association graph, nodes and edges of different levels can be used to represent multi-level association relationships. For example, direct associations can be represented at a lower level, and indirect associations can be represented at a higher level through the connection of intermediate nodes, thereby forming a more complex and comprehensive association network.

[0036] A temperature loss cause association diagram construction unit, the temperature loss cause association diagram construction unit is used to establish the alarm type temperature loss cause association diagram based on the multi-level association relationship. Specifically, using the constructed multi-level association relationship, a complete alarm type temperature loss cause association diagram is finally established. The association diagram will clearly display different alarm types, temperature loss causes and the complex association relationship between them, and the association diagram will be visualized so that users can more intuitively understand and analyze the relationship between alarm types and temperature loss causes. The association diagram is continuously updated and optimized, and with the addition of new data and knowledge, the accuracy and practicality of the association diagram are further improved.

[0037] In a possible implementation, the association graph structure acquisition unit further includes: an association graph composition unit, the association graph composition unit is used for the alarm type temperature loss cause association graph including association graph nodes and association graph edges, the association graph nodes are the alarm type and the temperature loss cause, and the association graph edges are the preliminary association relationship between the alarm type and the temperature loss cause. Specifically, for the association graph nodes, they are determined as alarm types and temperature loss causes. On the one hand, by analyzing various data during the operation of the underground steam pipeline network, including historical alarm records, monitoring data, etc., different alarm types are identified. For example, there may be abnormal temperature alarms, which are manifested as temperatures at key nodes exceeding the normal range; pressure fluctuation alarms, that is, unstable pressure in the pipeline, etc. On the other hand, in-depth research on various possible causes of temperature loss involves analysis of pipeline material properties, such as pipeline aging may be due to long-term use leading to material performance degradation; external environmental factors, such as corrosion caused by changes in soil moisture, the impact of geological activities on pipelines, etc.; it may also be improper maintenance, such as improper installation, untimely maintenance, etc. The association graph edge represents the preliminary correlation between the alarm type and the temperature loss cause. The correlation is determined in a variety of ways. One is the mining of historical data. If a certain alarm type frequently appears at the same time as a specific temperature loss cause in past cases, it can be preliminarily considered that there is a correlation between them. Second, combined with expert knowledge, industry experts judge which temperature loss causes may cause certain alarm types based on experience. Third, logical reasoning is performed to analyze the operating principles and structural characteristics of the underground steam pipeline network, and infer the potential connection between the alarm type and the temperature loss cause. After determining this preliminary correlation, the corresponding alarm type node and the temperature loss cause node are connected using the association graph edge, thereby constructing a complete alarm type temperature loss cause correlation graph, so as to more intuitively analyze and understand the complex relationship between the alarm type and the temperature loss cause, and provide strong support for temperature loss warning and processing of underground steam pipeline networks.

[0038] The module 50 for obtaining the optimal solution to the temperature loss cause is used to analyze the reported alarm information of multiple key nodes based on the self-supervised alarm type classification and identification model, determine multiple alarm types, input the multiple alarm types into the alarm type temperature loss cause association diagram, and obtain the optimal solution to the temperature loss cause. Specifically, the alarm information uploaded by multiple key nodes is obtained, and the key nodes are the parts in the underground steam pipe network that are determined to be prone to temperature loss problems after analysis, such as pipe connections, elbows, etc. The reported alarm information includes data such as temperature anomaly, pressure fluctuation, flow change, etc. The self-supervised alarm type classification and recognition model is used to analyze these reported alarm information. The model is trained and can determine the corresponding alarm type based on the input alarm information characteristics. For example, the model can determine whether it is a temperature anomaly alarm, a pressure fluctuation alarm, or other types of alarms based on the amplitude of temperature change, the frequency of pressure fluctuations, and other characteristics. Through the analysis of the model, multiple alarm types are determined to provide a basis for the subsequent determination of the cause of temperature loss. Once multiple alarm types are determined, these alarm types are input into the alarm type temperature loss cause association diagram. The association diagram is constructed through the previous steps, and it shows the correlation between different alarm types and temperature loss. The relationship between the causes. The association diagram analyzes and infers the input alarm type through its internal logic and association relationships. For example, if the input alarm type includes temperature abnormality alarm and pressure fluctuation alarm, the association diagram will conduct a comprehensive analysis based on the correlation strength between these two alarm types and different temperature loss causes. After analyzing the association diagram, the optimal solution to the temperature loss cause is finally obtained. The optimal solution is the most likely combination of temperature loss causes after comprehensively considering multiple alarm types. For example, it may be concluded that the cause of temperature loss is the combined effect of pipeline aging and external environmental influences. The optimal solution to the temperature loss cause can provide accurate guidance for pipeline network managers so that they can take targeted maintenance and preventive measures to reduce the impact of temperature loss on underground steam pipelines.

[0039] The early warning module 60 is used to issue an early warning based on the optimal solution of the temperature loss cause and the key node position. Specifically, the early warning module first receives the optimal solution of the temperature loss cause from the temperature loss cause optimal solution acquisition module, and determines the key node position where the temperature loss occurs. Then, according to the optimal solution of the temperature loss cause, the possible impact and urgency are analyzed, while considering the importance of the key node position, a suitable early warning method is selected, such as sound and light alarms, SMS notifications, system pop-ups, etc., to issue accurate and complete early warning information, including the cause of temperature loss, the degree of harm, the key node position and recommended measures. After the early warning is issued, the temperature loss situation and key node changes are continuously monitored. If the situation deteriorates or there is new alarm information, the early warning content is updated in time and the early warning method and scope are adjusted to ensure the safe operation of the underground steam pipeline network.

[0040] In a possible implementation, the early warning module 60 further includes: a temperature loss processing decision tree construction unit, which is used to establish a temperature loss processing decision tree based on historical early warning data and preset processing solutions. Specifically, the historical early warning data of the underground steam pipeline network is collected, and the data includes the alarm type, temperature loss cause, key node location, degree of harm and other information when the temperature loss event occurred in the past. Through the analysis of the historical data, the characteristics and processing methods of different types of temperature loss events can be understood, and the preset processing solution can be determined. The preset processing solution can be formulated by experts based on experience, or it can be obtained by summarizing historical successful processing cases. The solution includes maintenance methods, emergency measures, resource allocation, etc. for different temperature loss causes. Based on the historical early warning data and the preset processing solution, a temperature loss processing decision tree is established. Each node of the decision tree represents a decision point, and branches to different processing solutions according to different conditions. For example, the root node of the decision tree can be the severity of the temperature loss, which is divided into three branches of high, medium and low according to the severity. Each branch is further subdivided according to other factors such as the cause of the temperature loss and the location of the key node, and finally reaches a specific processing solution node.

[0041] A processing solution production unit, the processing solution production unit is used to input the entered warning information into the temperature loss processing decision tree to generate a processing solution. Specifically, when new warning information is generated, the warning information is entered into the processing solution production unit. The warning information may include the alarm type of the current temperature loss event, the cause of the temperature loss, the location of key nodes, the degree of hazard, etc. The processing solution production unit inputs the entered warning information into the temperature loss processing decision tree. The decision tree starts from the root node according to the characteristics of the warning information, and makes judgments according to the branch conditions of the decision tree, and gradually traverses downward until it reaches a specific processing solution node to generate a processing solution. This processing solution is determined based on the judgment results of the decision tree, including specific maintenance measures, emergency actions, resource requirements, etc.

[0042] A processing result data acquisition unit, the processing result data acquisition unit is used to execute the processing scheme, obtain the processing result data, and update the temperature loss processing decision tree according to the processing result data. Specifically, the processing scheme is executed. Relevant personnel take actions according to the generated processing scheme to repair and process the temperature loss. During the execution process, the processing process and result data are recorded, including the actual measures taken, processing time, resource consumption, temperature loss repair status, etc., to obtain the processing result data. After the processing is completed, the processing results are evaluated and analyzed to obtain the processing result data. These data can reflect the effectiveness and efficiency of the processing scheme, and the temperature loss processing decision tree is updated according to the processing result data. The processing result data is fed back to the temperature loss processing decision tree construction unit, and the decision tree is adjusted and optimized according to the actual processing results. If the processing scheme successfully solves the temperature loss problem, the weight of the scheme in the decision tree can be strengthened; if the processing scheme is not effective, the reasons can be analyzed, and the branch conditions and processing schemes of the decision tree can be adjusted to improve the accuracy and practicality of the decision tree.

[0043] The embodiment of the present application adopts the method of determining the underground steam pipeline protection pipeline diagram, analyzing the key node positions and establishing a temperature loss data acquisition system; obtaining temperature loss warning historical data, extracting text feature vectors and structured fields to establish a temperature loss vocabulary; establishing an alarm type classification and recognition model based on the temperature loss vocabulary, obtaining the alarm type sequence and associating the temperature loss cause, and evaluating the hazard level; calculating the alarm type similarity to establish an association graph; analyzing the key node alarm information to determine the alarm type, inputting the association graph to obtain the optimal solution for the temperature loss cause, and issuing an alarm based on the optimal solution and the key node position. By introducing unstructured data analysis, damage process time characteristics and transaction logic chain data filtering, and multi-node multi-dimensional alarm prediction mutual verification and other means, the accuracy and reliability of underground steam pipeline temperature loss warnings are improved, potential temperature loss problems are accurately discovered, false alarms are reduced, and the causes of temperature loss are quickly located, ensuring the safe and stable operation of the pipeline network.

[0044] Although the present application makes various references to certain modules in the system according to the embodiments of the present application, any number of different modules may be used and run on the user terminal and / or server, and the various units and modules included are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be achieved; in addition, the specific names of the functional units are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of the present invention.

[0045] The above specific implementation manner does not constitute a limitation to the protection scope of the present application. It should be understood by those skilled in the art that various modifications, combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present application should be included in the protection scope of the present application. In some cases, the actions or steps recorded in the present application can be performed in an order different from that in the embodiment and can still achieve the desired results. In addition, the process depicted in the accompanying drawings does not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

Claims

1. An intelligent temperature loss early warning monitoring device for underground steam pipe network based on a self-organizing network, characterized in that: include: A temperature loss data acquisition system establishment module, the temperature loss data acquisition system establishment module is used to determine the protection pipeline diagram of the underground steam pipeline network, perform key node position analysis on the protection pipeline diagram, and establish a temperature loss data acquisition system for key nodes of the pipeline; A temperature loss word list construction module, the temperature loss word list construction module is used to obtain temperature loss warning historical data, obtain text feature vectors and structured fields based on the temperature loss warning historical data, the text feature vectors are obtained through unstructured fields, the text feature vectors and the structured fields are analyzed, and a limited set temperature loss word list is established; A classification and recognition model building module, wherein the classification and recognition model building module is used to establish a self-supervised alarm type classification and recognition model based on the finite set temperature loss vocabulary, obtain an alarm type sequence, associate each alarm type in the alarm type sequence with multiple temperature loss causes, and evaluate the hazard level of each alarm type; A temperature loss cause association graph construction module, the temperature loss cause association graph construction module is used to obtain the alarm type sequence, the hazard level, the multiple temperature loss causes, calculate the alarm type similarity, and establish an alarm type temperature loss cause association graph; A module for obtaining an optimal solution to the cause of temperature loss, the module for obtaining an optimal solution to the cause of temperature loss is used to analyze the reported alarm information of the plurality of key nodes based on the self-supervised alarm type classification and identification model, determine a plurality of alarm types, input the plurality of alarm types into the alarm type temperature loss cause association diagram, and obtain an optimal solution to the cause of temperature loss; An early warning module is used to issue an early warning based on the optimal solution to the temperature loss cause and the key node position.

2. The underground steam pipe network intelligent temperature loss early warning monitoring device based on ad hoc network according to claim 1, characterized in that: The temperature loss vocabulary building module includes: A data collection unit, the data collection unit is used to collect real-time data and historical data of the key nodes; A data filtering unit, the data filtering unit is used to obtain the damage process time characteristics of the key node based on the historical data of the key node, and filter the real-time data based on the damage process time characteristics; A transaction logic chain construction unit, the transaction logic chain construction unit is used to construct a transaction logic chain of temperature loss damage of the underground steam pipe network, and filter the real-time data based on the transaction logic chain; A real-time data output unit, wherein the real-time data output unit is used to output the real-time data of the key node.

3. The intelligent temperature loss early warning monitoring device for underground steam pipe network based on ad hoc network according to claim 2, characterized in that: The transaction logic chain building unit includes: A structural model building subunit, the structural model building subunit is used to build a structural model of the underground steam pipe network; An event association subunit, the event association subunit is used to obtain the temperature damage characteristics of the underground steam pipe network and the corresponding key nodes and time, analyze the key nodes and time to associate the temperature damage characteristics with the actual temperature damage events, and construct a multi-stage transaction logic chain; A logic chain adjustment subunit, wherein the logic chain adjustment subunit is used to dynamically adjust the multi-stage transaction logic chain based on the real-time data and historical data.

4. The intelligent temperature loss early warning monitoring device for underground steam pipe network based on ad hoc network according to claim 1, characterized in that: The temperature loss vocabulary building module includes: An unstructured field acquisition unit, the unstructured field acquisition unit is used to acquire an unstructured field based on the temperature loss warning historical data; A text feature vector acquisition unit, the text feature vector acquisition unit is used to process the unstructured field based on natural language processing technology to obtain the text feature vector; A vector preprocessing unit, wherein the vector preprocessing unit is used to preprocess the text feature vector; A cluster analysis unit, the cluster analysis unit is used to perform cluster analysis on the pre-processed text feature vectors to obtain multiple groups of text feature vectors; A vector grouping unit is used to establish a finite set temperature loss vocabulary based on the plurality of groups of text feature vector groups.

5. The intelligent temperature loss early warning monitoring device for underground steam pipe network based on ad hoc network according to claim 1, characterized in that: The classification recognition model building module includes: A structured field acquisition unit, wherein the structured field is used to acquire a structured field based on the temperature loss warning historical data; An initial self-supervisory alarm type classification and identification model construction unit, wherein the initial self-supervisory alarm type classification and identification model construction unit is used to establish an initial self-supervisory alarm type classification and identification model; A text comprehensive feature vector acquisition unit, the text comprehensive feature vector acquisition unit is used to acquire a text comprehensive feature vector based on the structured field and the text feature vector; A model training unit is used to construct positive sample pairs and negative sample pairs based on the text comprehensive feature vector, input the initial self-supervised alarm type classification and recognition model, perform training, and obtain the self-supervised alarm type classification and recognition model.

6. The intelligent temperature loss early warning monitoring device for underground steam pipe network based on ad hoc network according to claim 1, characterized in that: The temperature loss cause association diagram construction module includes: A correlation graph structure acquisition unit, the correlation graph structure acquisition unit is used to acquire a preliminary correlation relationship between the alarm type and the temperature loss cause based on the alarm type sequence and the temperature loss cause, and establish a basic structure of an alarm type temperature loss cause correlation graph based on the preliminary correlation relationship; A weight adjustment unit, the weight adjustment unit being used to dynamically adjust a weight value of the preliminary association relationship based on the alarm type similarity; A multi-level association relationship construction unit, the multi-level association relationship construction unit is used to construct a multi-level association relationship based on the association type and the preliminary association relationship; A temperature loss cause association diagram construction unit, wherein the temperature loss cause association diagram construction unit is used to establish the alarm type temperature loss cause association diagram based on the multi-level association relationship.

7. The underground steam pipe network intelligent temperature loss early warning monitoring device based on ad hoc network according to claim 6, characterized in that: The alarm type and temperature loss cause association graph includes association graph nodes and association graph edges. The association graph nodes are the alarm types and the temperature loss causes, and the association graph edges are the preliminary association relationships between the alarm types and the temperature loss causes.

8. The underground steam pipe network intelligent temperature loss early warning monitoring device based on ad hoc network according to claim 1, characterized in that: The early warning module comprises: A temperature loss processing decision tree building unit, wherein the temperature loss processing decision tree building unit is used to establish a temperature loss processing decision tree based on historical warning data and a preset processing solution; A processing solution production unit, the processing solution production unit is used to input the input warning information into the temperature loss processing decision tree to generate a processing solution; A processing result data acquisition unit is used to execute the processing scheme, acquire processing result data, and update the temperature loss processing decision tree according to the processing result data.

Images