Drying tail gas abnormity management method and system

Abstract

The invention provides a drying tail gas abnormity management method and system, and relates to the technical field of abnormal data processing, and the method comprises the steps: obtaining a monitoring data source, loading a sludge drying treatment process source, then carrying out expected abnormity monitoring, constructing an abnormity processing decision channel according to an abnormity processing decision learning function, and carrying out abnormal data processing. Generating a drying tail gas abnormity processing decision, and finally performing drying tail gas abnormity management. The technical problems that in the prior art, due to the fact that a tail gas monitoring system and an exception treatment system are independent of each other, deep data analysis and intelligent decision support are lacked, the discovery and treatment efficiency of the exception condition is low, and the overall effect of sludge drying treatment is further affected can be solved; tail gas abnormity management is effectively carried out according to the drying tail gas abnormity processing decision, normal operation of the drying process is guaranteed, and the influence on the environment is reduced.

Description

Technical Field

[0001] The present application relates to the technical field of abnormal data processing, and in particular to a method and system for managing abnormalities of dried exhaust gas. Background Art

[0002] During the industrial drying process, the tail gas may contain a large amount of harmful substances, posing a threat to the environment and human health. Therefore, effectively managing the abnormality of drying tail gas and reducing the emission of harmful substances are urgent needs for environmental protection.

[0003] At present, in the existing sludge drying process, the tail gas monitoring and abnormality handling systems are often independent of each other, lacking close integration and intelligent decision-making support. Traditional tail gas monitoring systems can usually only provide basic data collection functions, but cannot conduct in-depth analysis of data, and it is even more difficult to make fast and accurate abnormality handling decisions based on real-time data. At the same time, these systems often ignore the inherent connection between the sludge drying process and the tail gas monitoring data, resulting in large blind spots and delays in the identification and handling of abnormal situations.

[0004] In summary, the existing technology has technical problems such as the independence of exhaust gas monitoring and abnormality handling systems, lack of in-depth data analysis and intelligent decision-making support, resulting in low efficiency in the discovery and handling of abnormal situations, which further affects the overall effect of sludge drying treatment. Summary of the invention

[0005] The purpose of this application is to provide a method and system for managing abnormalities in drying exhaust gas, so as to solve the technical problems in the prior art that the exhaust gas monitoring and abnormality handling systems are independent of each other, lack of in-depth data analysis and intelligent decision-making support, resulting in low efficiency in the discovery and handling of abnormal situations, which further affects the overall effect of sludge drying treatment.

[0006] In view of the above problems, the present application provides a method and system for abnormal management of dry exhaust gas.

[0007] In the first aspect, the present application provides a method for managing dry exhaust gas abnormalities, which is implemented through a dry exhaust gas abnormality management system, wherein the method includes: obtaining a dry exhaust gas monitoring data source according to a dry exhaust gas monitoring end; loading a sludge drying process source corresponding to the dry exhaust gas monitoring data source, wherein the sludge drying process source includes a sludge component data set and a sludge drying process data set; performing expected abnormality detection on the dry exhaust gas monitoring data source according to the sludge drying process source to obtain a dry exhaust gas abnormality detection vector; constructing a dry exhaust gas abnormality processing decision channel according to a dry exhaust gas abnormality processing decision learning function; inputting the dry exhaust gas abnormality detection vector into the dry exhaust gas abnormality processing channel to generate a dry exhaust gas abnormality processing decision; and performing dry exhaust gas abnormality management according to the dry exhaust gas abnormality processing decision.

[0008] In the second aspect, the present application also provides a dry exhaust gas abnormality management system for executing a dry exhaust gas abnormality management method as described in the first aspect, wherein the system includes: a monitoring data source acquisition module, the monitoring data source acquisition module is used to obtain a dry exhaust gas monitoring data source according to a dry exhaust gas monitoring end; a treatment process source loading module, the treatment process source loading module is used to load a sludge drying treatment process source corresponding to the dry exhaust gas monitoring data source, wherein the sludge drying treatment process source includes a sludge component data set and a sludge drying treatment process data set; an abnormality detection module, the abnormality detection module is used to load a sludge component data set according to the The sludge drying process source performs expected anomaly detection on the drying exhaust gas monitoring data source to obtain a drying exhaust gas anomaly detection vector; an anomaly handling decision channel construction module, the anomaly handling decision channel construction module is used to construct a drying exhaust gas anomaly handling decision channel according to a drying exhaust gas anomaly handling decision learning function; an anomaly handling decision acquisition module, the anomaly handling decision acquisition module is used to input the drying exhaust gas anomaly detection vector into the drying exhaust gas anomaly handling channel to generate a drying exhaust gas anomaly handling decision; an anomaly management module, the anomaly management module is used to perform drying exhaust gas anomaly management according to the drying exhaust gas anomaly handling decision.

[0009] One or more technical solutions provided in this application have at least the following technical effects or advantages: According to the drying tail gas monitoring end, a drying tail gas monitoring data source is obtained; the sludge drying treatment process source corresponding to the drying tail gas monitoring data source is loaded, wherein the sludge drying treatment process source includes a sludge component data set and a sludge drying treatment process data set; the drying tail gas monitoring data source is subjected to expected abnormality detection according to the sludge drying treatment process source to obtain a drying tail gas abnormality detection vector; according to the drying tail gas abnormality processing decision learning function, a drying tail gas abnormality processing decision channel is constructed; the drying tail gas abnormality detection vector is input into the drying tail gas abnormality processing channel to generate a drying tail gas abnormality processing decision; and the drying tail gas abnormality is managed according to the drying tail gas abnormality processing decision, which effectively solves the technical problem that the prior art is independent of the tail gas monitoring and abnormality processing systems, lacks in-depth data analysis and intelligent decision support, resulting in low efficiency in the discovery and processing of abnormal situations, further affecting the overall effect of sludge drying treatment, and effectively manages the tail gas abnormality according to the drying tail gas abnormality processing decision, ensures the normal operation of the drying process, and reduces the impact on the environment.

[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 according to the contents of the specification, and 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 specifically cited below. It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present application, nor is it intended to limit the scope of the present application. Other features of the present application will become easy to understand through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] In order to more clearly illustrate the technical solutions in the present application or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only exemplary, and for ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0012] Figure 1 A schematic diagram of a process for abnormal management of dry exhaust gas in this application; Figure 2 This is a structural schematic diagram of a dry exhaust gas abnormal management system for the present application.

[0013] Description of reference numerals: Monitoring data source acquisition module 11, processing process source loading module 12, anomaly detection module 13, anomaly processing decision channel construction module 14, anomaly processing decision acquisition module 15, anomaly management module 16. DETAILED DESCRIPTION

[0014] The present application provides a method and system for managing abnormalities in drying exhaust gas, thereby solving the technical problem in the prior art that the exhaust gas monitoring and abnormality handling systems are independent of each other, lack in-depth data analysis and intelligent decision-making support, resulting in low efficiency in discovering and handling abnormal situations, which further affects the overall effect of sludge drying treatment. The application effectively manages the abnormalities in exhaust gas based on the abnormal handling decisions for drying exhaust gas, ensures the normal operation of the drying process, and reduces the impact on the environment.

[0015] Below, the technical solutions in the present application will be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments of the present application. It should be understood that the present application is not limited to the example embodiments described herein. Based on the embodiments of the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present application. It should also be noted that, for the convenience of description, only the parts related to the present application are shown in the accompanying drawings, rather than all of them.

[0016] Embodiment 1 Please see attached Figure 1 The present application provides a method for managing abnormalities of dry exhaust gas, wherein the method is applied to a management system for abnormalities of dry exhaust gas, and the method specifically comprises the following steps: S1: According to the exhaust gas drying monitoring terminal, a exhaust gas drying monitoring data source is obtained.

[0017] Specifically, the exhaust gas monitoring terminal is a monitoring device installed at the exhaust gas outlet of the drying equipment, which is used to collect and transmit exhaust gas data in real time. The data source of the exhaust gas monitoring terminal is the raw data collected by the exhaust gas monitoring terminal, which reflects the real-time situation of exhaust gas emissions.

[0018] S2: Loading the sludge drying treatment process source corresponding to the drying tail gas monitoring data source, wherein the sludge drying treatment process source includes a sludge component data set and a sludge drying treatment process data set.

[0019] Specifically, the sludge composition dataset contains information on various components in the sludge, such as heavy metal content, trace highly toxic organic matter, such as PCBs, AOX, etc., and the types and quantities of pathogenic microorganisms such as bacteria, virions and parasite eggs. By analyzing the sludge composition, the drying process can be optimized, the emission of harmful substances can be reduced, and the drying efficiency can be improved. The sludge drying process dataset includes all aspects of sludge drying treatment, including specific operating parameters and conditions for pretreatment, heating, evaporation, drying and other steps. The key process parameters such as temperature, pressure, time, etc. under different sludge types and different treatment stages, as well as the corresponding tail gas emissions are recorded.

[0020] S3: performing expected anomaly detection on the drying exhaust gas monitoring data source according to the sludge drying treatment process source to obtain a drying exhaust gas anomaly detection vector.

[0021] Specifically, data preprocessing is performed on the drying tail gas monitoring data source and the sludge drying process source, including data cleaning, standardization and normalization. Key process characteristics related to drying tail gas emissions, such as temperature, humidity, sludge composition, etc., are extracted from the sludge drying process source. At the same time, key indicators of tail gas emissions, such as the concentration of various pollutants, are extracted from the drying tail gas monitoring data source. Using historical data and statistical methods, an expected model of tail gas emissions based on the characteristics of the sludge drying process under normal operating conditions is established. The normal range of various indicators in the tail gas under given process conditions can be predicted. The real-time monitored drying tail gas data is compared with the expected model to identify data points with large deviations from the model prediction value, which are regarded as abnormal points. Anomaly detection can use statistical methods such as Z-score, IQR, etc. or machine learning algorithms such as isolation forest, one-class support vector machine, etc. According to the results of anomaly detection, an anomaly detection vector is generated for each monitoring indicator.

[0022] S4: Construct a dry exhaust gas abnormality processing decision channel according to the dry exhaust gas abnormality processing decision learning function.

[0023] Specifically, the dry exhaust gas anomaly processing decision learning function can output the corresponding processing decision based on the input dry exhaust gas anomaly detection vector. It can be a machine learning model such as a rule-based system, a decision tree, a neural network, a random forest, etc. Collect historical data, including the dry exhaust gas anomaly detection vector and the corresponding optimal processing decision. This data will be used to train the learning function so that it can learn the mapping relationship from anomaly detection to processing decision. Use the prepared data set to train the learning function. During the training process, cross-validation, grid search and other techniques are used to optimize the model performance. Use an independent validation data set to evaluate the performance of the learning function. This is done by calculating indicators such as accuracy, recall, and F1 score. If the performance is not good, it is necessary to adjust the model parameters or try different models. After the learning function is trained and verified that its performance meets the requirements, it is deployed as part of the dry exhaust gas anomaly processing decision channel. This channel will receive the real-time dry exhaust gas anomaly detection vector as input and output the corresponding processing decision through the learning function.

[0024] S5: Inputting the dry exhaust gas abnormality detection vector into the dry exhaust gas abnormality processing channel to generate a dry exhaust gas abnormality processing decision.

[0025] Specifically, ensure that the dry exhaust anomaly detection vector is accurate and up-to-date, containing key indicator information extracted from the exhaust monitoring data source and detected for anomalies. These indicators include the degree of abnormality of parameters such as pollutant concentration, temperature, and pressure. Input the prepared dry exhaust anomaly detection vector into the dry exhaust anomaly processing channel. This channel is built based on a previously trained learning function and is able to receive anomaly detection vectors as input. The learning function will analyze and calculate based on the input anomaly detection vector. By comparing the input data with the patterns in the training data, the learning function will determine the most appropriate processing decision. This decision includes adjusting operating parameters, initiating emergency response procedures, notifying operators, etc.

[0026] S6: Performing dry exhaust gas abnormality management according to the dry exhaust gas abnormality handling decision.

[0027] Specifically, when the drying exhaust gas abnormality processing channel generates processing decisions, these decisions are received in a timely manner and accurately understood. Processing decisions include specific operational suggestions, the range of adjustment parameters, emergency procedures that need to be initiated, etc. According to the received processing decisions, a detailed implementation plan is formulated. The plan includes the implementation steps, timetable, required resources, and possible risks and countermeasures. According to the implementation plan, the operating parameters of the drying equipment are adjusted, equipment parts are replaced or cleaned, and the efficiency of the exhaust gas treatment device is increased.

[0028] Furthermore, step S3 of the present application also includes: Data enhancement optimization is performed according to the drying exhaust gas monitoring data source to generate a drying exhaust gas monitoring vector; drying exhaust gas health prediction is performed according to the sludge drying treatment process source to construct a drying exhaust gas health vector; the drying exhaust gas health vector is used as the drying exhaust gas expected constraint, and the drying exhaust gas monitoring vector is subjected to deviation analysis to generate the drying exhaust gas abnormality detection vector.

[0029] Specifically, the data source of the exhaust gas drying monitoring data is preprocessed, including data cleaning, denoising, and standardization. Data enhancement techniques such as oversampling, undersampling, and data synthesis are used to balance data of different categories, especially abnormal data. Time series analysis technology is introduced to capture the characteristics of exhaust gas data changing over time. Key features such as the concentration, temperature, and humidity of various pollutants are extracted from the optimized data source. These features are combined into a multi-dimensional vector, namely the exhaust gas drying monitoring vector. This vector will represent the exhaust gas emission state at a certain point in time. Data from the sludge drying process source, especially process parameters and sludge composition information related to exhaust gas emissions, are used. By establishing a prediction model such as regression analysis and neural network, the healthy state of exhaust gas emissions under normal process conditions is predicted. This healthy state refers to the ideal range that various indicators in the exhaust gas should reach under normal operation. The various indicators of the predicted exhaust gas emission health state are combined into a multi-dimensional vector, namely the exhaust gas drying health vector. This vector will serve as the expected constraint for subsequent deviation analysis and anomaly detection. The exhaust gas drying monitoring vector generated in real time is compared with the exhaust gas drying health vector. Calculate the deviation between the two, and use various distance measurement methods, such as Euclidean distance, Mahalanobis distance, etc., or similarity calculation methods, such as cosine similarity, etc. to achieve this. According to the results of the deviation analysis, determine whether the exhaust emissions are abnormal. If the deviation exceeds the preset threshold, the exhaust emissions at that time point are considered abnormal. Combine the abnormal judgment results at each time point into a vector, namely the dry exhaust abnormality detection vector. Each element in this vector indicates whether the exhaust emissions at the corresponding time point are abnormal.

[0030] Furthermore, the present application also includes: Loading a dried exhaust gas health record library; aligning and associating the dried exhaust gas health record library according to the sludge drying process source to establish a dried exhaust gas health alignment association space; classifying dried exhaust gas health characteristic attributes according to the dried exhaust gas health alignment association space to construct multiple dried exhaust gas health characteristic domains; performing central tendency evaluation based on the multiple dried exhaust gas health characteristic domains to obtain multiple dried exhaust gas health characteristic baseline values; generating the dried exhaust gas health vector based on the multiple dried exhaust gas health characteristic baseline values.

[0031] Specifically, the dry tail gas health record library is loaded from the database. This record library contains the health data of the historical tail gas emissions, that is, the data collected under normal operating conditions. These data include key indicators such as the concentration, temperature, and pressure of various pollutants in the tail gas, as well as the corresponding sludge drying process parameters. The data in the sludge drying process source are used to align and associate with the data in the dry tail gas health record library. Through this alignment and association, a dry tail gas health alignment association space is established. This space links the process conditions and the tail gas emission status to form a multi-dimensional data space. Based on the dry tail gas health alignment association space, the health characteristic attributes of the tail gas emissions are classified. This is achieved through methods such as cluster analysis and principal component analysis, with the aim of classifying related features into one category. Based on the classification results, multiple dry tail gas health feature domains are constructed. Each feature domain represents a set of interrelated features that have a significant impact on the health status of the tail gas. For each dry tail gas health feature domain, a central tendency evaluation is performed. This includes calculating the mean, median, or other statistics of the feature to reflect the central tendency of the data in the feature domain. Through the central tendency evaluation, the benchmark values ​​of each characteristic domain can be obtained. These benchmark values ​​represent the expected or typical values ​​of each characteristic domain under normal process conditions. Finally, the benchmark values ​​of each characteristic domain are combined to generate the dry exhaust health vector. This vector is a multi-dimensional data representation that captures the health status of exhaust emissions under normal operating conditions.

[0032] Furthermore, the present application also includes: According to the dried exhaust gas health record library, a first dried exhaust gas health record group is extracted, wherein the first dried exhaust gas health record group includes a first historical sludge drying treatment process source, and first historical healthy dried exhaust gas monitoring data corresponding to the first historical sludge drying treatment process source; according to the sludge drying treatment process source, a registration association analysis is performed on the first historical sludge drying treatment process source to generate a first registration association coefficient; if the first registration association coefficient is greater than / equal to a preset registration association coefficient, the first historical healthy dried exhaust gas data is added to the dried exhaust gas health registration association space; according to the preset registration association coefficient, the dried exhaust gas health record library is continued to be registered and associated in combination with the sludge drying treatment process source to generate the dried exhaust gas health registration association space.

[0033] Specifically, the first drying tail gas health record group is extracted from the drying tail gas health record library. This record group contains the first historical sludge drying treatment process source and the corresponding first historical healthy drying tail gas monitoring data. The first historical sludge drying treatment process source includes process parameters such as temperature, pressure, sludge composition, and treatment time. The first historical healthy drying tail gas monitoring data includes key indicators such as pollutant concentration, temperature, and humidity in the tail gas. The current sludge drying treatment process source and the first historical sludge drying treatment process source are used for registration association analysis. This is achieved by calculating the correlation, distance measurement, or other similarity indicators between the two. The result of the analysis will generate a first registration association coefficient, which indicates the degree of similarity between the current process source and the first historical process source. The first registration association coefficient is compared with the preset registration association coefficient. The preset registration association coefficient is a threshold used to determine whether the two process sources are similar enough. If the first registration association coefficient is greater than or equal to the preset registration association coefficient, it means that the current process conditions are highly similar to the first historical process conditions. In this case, the first historical healthy drying tail gas data is added to the drying tail gas health registration association space. These data will provide a reference for subsequent exhaust health status prediction and anomaly detection. Repeat the above steps, use the preset registration correlation coefficient as the standard, and continue to perform registration correlation analysis on other historical records in the dry exhaust health record library. By continuously adding qualified historical healthy exhaust data to the registration correlation space, a complete and representative dry exhaust health registration correlation space is finally constructed.

[0034] Furthermore, step S4 of the present application also includes: Load the dry exhaust gas abnormality detection vector record set and the dry exhaust gas abnormality processing decision record set; based on the dry exhaust gas abnormality processing decision learning function, train H dry exhaust gas abnormality processing decision sub-models according to the dry exhaust gas abnormality detection vector record set and the dry exhaust gas abnormality processing decision record set, wherein H is a positive integer greater than 1; collect the output data sets of the H dry exhaust gas abnormality processing decision sub-models to generate a decision integration model construction data set; based on the dry exhaust gas abnormality processing decision learning function, train the dry exhaust gas abnormality processing decision integration model according to the decision integration model construction data set and the dry exhaust gas abnormality processing decision record set; merge the H dry exhaust gas abnormality processing decision sub-models as parallel independent nodes to generate a dry exhaust gas abnormality processing decision parsing layer; merge the dry exhaust gas abnormality processing decision parsing layer with the input layer of the dry exhaust gas abnormality processing decision integration model to generate the dry exhaust gas abnormality processing decision channel.

[0035] Specifically, the dry exhaust anomaly detection vector record set contains records of exhaust anomalies detected in history, and each record is an anomaly detection vector that captures various features of exhaust anomalies. The dry exhaust anomaly processing decision record set records the processing decisions and their results taken in history for exhaust anomalies. Use the dry exhaust anomaly processing decision learning function, which is an algorithm for learning how to make anomaly processing decisions from data. Based on the two loaded record sets, train H dry exhaust anomaly processing decision sub-models. Each sub-model will try to learn from the data how to make the best processing decision based on the anomaly detection vector. H is a positive integer greater than 1, indicating the number of sub-models. Multiple sub-models can increase the diversity and robustness of decisions. Collect the output data sets of these H sub-models for the same set of anomaly detection vectors. These outputs reflect the decision differences of different sub-models for the same input. Combine these output data sets with the corresponding actual processing decision records to generate a decision integration model construction data set. This data set will be used to train a higher-level integration model. The dry exhaust gas abnormality handling decision learning function is used again, but this time it is trained based on the decision integration model to build a data set and a dry exhaust gas abnormality handling decision record set. The H sub-models are merged as parallel independent nodes to generate a dry exhaust gas abnormality handling decision parsing layer. This layer can process input data in parallel and output the decision results of each sub-model. Finally, the output of the dry exhaust gas abnormality handling decision parsing layer is used as the input of the dry exhaust gas abnormality handling decision integration model. In this way, the integrated model can comprehensively consider the decisions of each sub-model to make a more accurate and comprehensive final decision.

[0036] Furthermore, the present application also includes: According to H dry exhaust gas abnormality processing decision learners, extract the h-th dry exhaust gas abnormality processing decision learner, wherein h belongs to H and h is a positive integer; take the dry exhaust gas abnormality detection vector record set as input data, take the dry exhaust gas abnormality processing decision record set as output data, train and test the h-th dry exhaust gas abnormality processing decision learner, and obtain the h-th dry exhaust gas abnormality processing decision maker, as well as the h-th dry exhaust gas abnormality processing decision accuracy coefficient and the h-th dry exhaust gas abnormality processing decision error coefficient corresponding to the h-th dry exhaust gas abnormality processing decision maker; the h-th dry exhaust gas abnormality processing decision accuracy coefficient and the h-th dry exhaust gas abnormality processing decision error coefficient are calculated. h drying exhaust gas abnormality handling decision error coefficient is input into the drying exhaust gas abnormality handling decision learning function to obtain the h-th drying exhaust gas abnormality handling decision learning index; it is determined whether the h-th drying exhaust gas abnormality handling decision learning index is greater than / equal to the preset drying exhaust gas abnormality handling decision learning index; if the h-th drying exhaust gas abnormality handling decision learning index is greater than / equal to the preset drying exhaust gas abnormality handling decision learning index, the h-th drying exhaust gas abnormality handling decision device is set to the h-th drying exhaust gas abnormality handling decision sub-model; the h-th drying exhaust gas abnormality handling decision sub-model is added to the H drying exhaust gas abnormality handling decision sub-models.

[0037] Specifically, from H dry exhaust gas abnormality processing decision learners, the hth learner is selected for operation, where h is a positive integer value in H. The hth dry exhaust gas abnormality processing decision learner is trained and tested using the dry exhaust gas abnormality detection vector record set as input data and the dry exhaust gas abnormality processing decision record set as output data, i.e., label. Through this process, the trained hth dry exhaust gas abnormality processing decision maker is obtained. In the testing phase, the performance of the hth dry exhaust gas abnormality processing decision maker is evaluated, and the hth dry exhaust gas abnormality processing decision accuracy coefficient and the hth dry exhaust gas abnormality processing decision error coefficient are calculated. These coefficients reflect the accuracy and error rate of the decision maker when handling abnormal situations. The hth dry exhaust gas abnormality processing decision accuracy coefficient and the hth dry exhaust gas abnormality processing decision error coefficient are input into the dry exhaust gas abnormality processing decision learning function, and the hth dry exhaust gas abnormality processing decision learning index is calculated. This learning index is a comprehensive indicator used to evaluate the overall performance of the decision maker. Determine whether the h-th dry exhaust gas abnormality handling decision learning index is greater than or equal to the preset dry exhaust gas abnormality handling decision learning index. If the h-th dry exhaust gas abnormality handling decision learning index is greater than or equal to the preset value, it means that the performance of the decision maker has reached the expected standard. In this case, the h-th dry exhaust gas abnormality handling decision maker is set to the h-th dry exhaust gas abnormality handling decision sub-model. Finally, the set h-th dry exhaust gas abnormality handling decision sub-model is added to the set of H dry exhaust gas abnormality handling decision sub-models.

[0038] Furthermore, the present application also includes: The dry exhaust gas abnormality processing decision learning function is: ; in, Characterize the learning index of decision-making for abnormal handling of dry exhaust gas, Characterize the decision learning factor of the predetermined dry exhaust abnormality processing, and, greater than 1, Characterizes the accuracy coefficient of the decision-making on abnormal treatment of dry exhaust gas, Characterizes the error coefficient of decision-making for abnormal handling of dry exhaust gas.

[0039] Specifically, this function is used to calculate the learning index of the dry exhaust abnormality handling decision, which is a comprehensive evaluation index based on the dry exhaust abnormality handling decision accuracy coefficient and the dry exhaust abnormality handling decision error coefficient. Through this function, a comprehensive learning index can be obtained. , which takes into account both the accuracy and error rate of the decision. A high learning index means that the decision maker performs well in making exception handling decisions, while a low learning index may indicate that further training or optimization is needed.

[0040] In summary, the method for managing abnormal exhaust gas drying provided by the present application has the following technical effects: According to the drying tail gas monitoring end, a drying tail gas monitoring data source is obtained; the sludge drying treatment process source corresponding to the drying tail gas monitoring data source is loaded, wherein the sludge drying treatment process source includes a sludge component data set and a sludge drying treatment process data set; the drying tail gas monitoring data source is subjected to expected abnormality detection according to the sludge drying treatment process source to obtain a drying tail gas abnormality detection vector; according to the drying tail gas abnormality processing decision learning function, a drying tail gas abnormality processing decision channel is constructed; the drying tail gas abnormality detection vector is input into the drying tail gas abnormality processing channel to generate a drying tail gas abnormality processing decision; and the drying tail gas abnormality is managed according to the drying tail gas abnormality processing decision, which effectively solves the technical problem that the prior art is independent of the tail gas monitoring and abnormality processing systems, lacks in-depth data analysis and intelligent decision support, resulting in low efficiency in the discovery and processing of abnormal situations, further affecting the overall effect of sludge drying treatment, and effectively manages the tail gas abnormality according to the drying tail gas abnormality processing decision, ensures the normal operation of the drying process, and reduces the impact on the environment.

[0041] Embodiment 2 Based on the same inventive concept as the method for managing abnormal dry exhaust gas in the aforementioned embodiment, the present application also provides an abnormal dry exhaust gas management system, see Attachment 1. Figure 2 , the system comprising: The monitoring data source acquisition module 11 is used to obtain a dried exhaust gas monitoring data source according to a dried exhaust gas monitoring terminal.

[0042] The processing source loading module 12 is used to load the sludge drying processing source corresponding to the drying tail gas monitoring data source, wherein the sludge drying processing source includes a sludge component data set and a sludge drying processing data set.

[0043] The anomaly detection module 13 is used to perform expected anomaly detection on the drying tail gas monitoring data source according to the sludge drying treatment process source to obtain a drying tail gas anomaly detection vector.

[0044] The abnormality handling decision channel construction module 14 is used to construct a dry exhaust gas abnormality handling decision channel according to the dry exhaust gas abnormality handling decision learning function.

[0045] The abnormality handling decision acquisition module 15 is used to input the dry exhaust gas abnormality detection vector into the dry exhaust gas abnormality handling channel to generate a dry exhaust gas abnormality handling decision.

[0046] The abnormality management module 16 is used to perform abnormality management of the dry exhaust gas according to the abnormality handling decision of the dry exhaust gas.

[0047] Furthermore, the abnormality detection module 13 in the system is also used for: Data enhancement optimization is performed according to the drying exhaust gas monitoring data source to generate a drying exhaust gas monitoring vector; drying exhaust gas health prediction is performed according to the sludge drying treatment process source to construct a drying exhaust gas health vector; the drying exhaust gas health vector is used as the drying exhaust gas expected constraint, and the drying exhaust gas monitoring vector is subjected to deviation analysis to generate the drying exhaust gas abnormality detection vector.

[0048] Furthermore, the system further comprises a characteristic reference value acquisition module, wherein the characteristic reference value acquisition module is used to: Loading a dried exhaust gas health record library; aligning and associating the dried exhaust gas health record library according to the sludge drying process source to establish a dried exhaust gas health alignment association space; classifying dried exhaust gas health characteristic attributes according to the dried exhaust gas health alignment association space to construct multiple dried exhaust gas health characteristic domains; performing central tendency evaluation based on the multiple dried exhaust gas health characteristic domains to obtain multiple dried exhaust gas health characteristic baseline values; generating the dried exhaust gas health vector based on the multiple dried exhaust gas health characteristic baseline values.

[0049] Furthermore, the system further comprises a registration association space generation module, wherein the registration association space generation module is used for: According to the dried exhaust gas health record library, a first dried exhaust gas health record group is extracted, wherein the first dried exhaust gas health record group includes a first historical sludge drying treatment process source, and first historical healthy dried exhaust gas monitoring data corresponding to the first historical sludge drying treatment process source; according to the sludge drying treatment process source, a registration association analysis is performed on the first historical sludge drying treatment process source to generate a first registration association coefficient; if the first registration association coefficient is greater than / equal to a preset registration association coefficient, the first historical healthy dried exhaust gas data is added to the dried exhaust gas health registration association space; according to the preset registration association coefficient, the dried exhaust gas health record library is continued to be registered and associated in combination with the sludge drying treatment process source to generate the dried exhaust gas health registration association space.

[0050] Furthermore, the exception handling decision channel construction module 14 in the system is also used for: Load the dry exhaust gas abnormality detection vector record set and the dry exhaust gas abnormality processing decision record set; based on the dry exhaust gas abnormality processing decision learning function, train H dry exhaust gas abnormality processing decision sub-models according to the dry exhaust gas abnormality detection vector record set and the dry exhaust gas abnormality processing decision record set, wherein H is a positive integer greater than 1; collect the output data sets of the H dry exhaust gas abnormality processing decision sub-models to generate a decision integration model construction data set; based on the dry exhaust gas abnormality processing decision learning function, train the dry exhaust gas abnormality processing decision integration model according to the decision integration model construction data set and the dry exhaust gas abnormality processing decision record set; merge the H dry exhaust gas abnormality processing decision sub-models as parallel independent nodes to generate a dry exhaust gas abnormality processing decision parsing layer; merge the dry exhaust gas abnormality processing decision parsing layer with the input layer of the dry exhaust gas abnormality processing decision integration model to generate the dry exhaust gas abnormality processing decision channel.

[0051] Furthermore, the system further comprises an exception handling sub-model acquisition module, wherein the exception handling sub-model acquisition module is used to: According to H dry exhaust gas abnormality processing decision learners, extract the h-th dry exhaust gas abnormality processing decision learner, wherein h belongs to H and h is a positive integer; take the dry exhaust gas abnormality detection vector record set as input data, take the dry exhaust gas abnormality processing decision record set as output data, train and test the h-th dry exhaust gas abnormality processing decision learner, and obtain the h-th dry exhaust gas abnormality processing decision maker, as well as the h-th dry exhaust gas abnormality processing decision accuracy coefficient and the h-th dry exhaust gas abnormality processing decision error coefficient corresponding to the h-th dry exhaust gas abnormality processing decision maker; the h-th dry exhaust gas abnormality processing decision accuracy coefficient and the h-th dry exhaust gas abnormality processing decision error coefficient are calculated. h drying exhaust gas abnormality handling decision error coefficient is input into the drying exhaust gas abnormality handling decision learning function to obtain the h-th drying exhaust gas abnormality handling decision learning index; it is determined whether the h-th drying exhaust gas abnormality handling decision learning index is greater than / equal to the preset drying exhaust gas abnormality handling decision learning index; if the h-th drying exhaust gas abnormality handling decision learning index is greater than / equal to the preset drying exhaust gas abnormality handling decision learning index, the h-th drying exhaust gas abnormality handling decision device is set to the h-th drying exhaust gas abnormality handling decision sub-model; the h-th drying exhaust gas abnormality handling decision sub-model is added to the H drying exhaust gas abnormality handling decision sub-models.

[0052] Furthermore, the present application also includes a learning function construction module, wherein the learning function construction module is used to: The dry exhaust gas abnormality processing decision learning function is: ; in, Characterize the learning index of decision-making for abnormal handling of dry exhaust gas, Characterize the decision learning factor of the predetermined dry exhaust abnormality processing, and, greater than 1, Characterizes the accuracy coefficient of the decision-making on abnormal treatment of dry exhaust gas, Characterizes the error coefficient of decision-making for abnormal handling of dry exhaust gas.

[0053] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. Figure 1 The method and specific examples of abnormal management of dry exhaust gas in the first embodiment are also applicable to the abnormal management system of dry exhaust gas in the present embodiment. Through the above detailed description of the abnormal management method of dry exhaust gas, those skilled in the art can clearly know the abnormal management system of dry exhaust gas in the present embodiment, so for the sake of brevity of the specification, it will not be described in detail here. For the system disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part description.

[0054] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

[0055] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application belong to the scope of the present application and its equivalent technology, the present application is also intended to include these modifications and variations.

Claims

1. A method for managing abnormalities of dry exhaust gas, characterized in that: The method comprises: According to the dry exhaust gas monitoring terminal, a dry exhaust gas monitoring data source is obtained; Loading the sludge drying treatment process source corresponding to the drying tail gas monitoring data source, wherein the sludge drying treatment process source includes a sludge component data set and a sludge drying treatment process data set; Performing expected anomaly detection on the drying tail gas monitoring data source according to the sludge drying treatment process source to obtain a drying tail gas anomaly detection vector; According to the dry exhaust gas abnormal processing decision learning function, a dry exhaust gas abnormal processing decision channel is constructed; Inputting the dry exhaust gas abnormality detection vector into the dry exhaust gas abnormality processing channel to generate a dry exhaust gas abnormality processing decision; Abnormal management of dry exhaust gas is performed according to the abnormal handling decision of dry exhaust gas.

2. The method according to claim 1, characterized in that According to the sludge drying treatment process source, the drying tail gas monitoring data source is subjected to expected abnormality detection to obtain a drying tail gas abnormality detection vector, including: Performing data enhancement optimization according to the dry exhaust gas monitoring data source to generate a dry exhaust gas monitoring vector; Perform a drying tail gas health prediction based on the sludge drying treatment process source and construct a drying tail gas health vector; The dry exhaust gas healthy vector is used as the dry exhaust gas expected constraint, and the dry exhaust gas monitoring vector is subjected to deviation analysis to generate the dry exhaust gas abnormality detection vector.

3. The method according to claim 2, characterized in that According to the sludge drying treatment process source, the drying tail gas health prediction is performed to construct the drying tail gas health vector, including: Load the dry exhaust health record library; According to the sludge drying treatment process source, the dried tail gas health record library is registered and associated to establish a dried tail gas health registration association space; Classifying dry exhaust gas health characteristic attributes according to the dry exhaust gas health registration association space, and constructing multiple dry exhaust gas health characteristic domains; Performing a central tendency evaluation based on the plurality of dry exhaust gas health characteristic domains to obtain a plurality of dry exhaust gas health characteristic benchmark values; The dry exhaust gas health vector is generated according to the multiple dry exhaust gas health characteristic reference values.

4. The method according to claim 3, characterized in that According to the sludge drying treatment process source, the drying tail gas health record library is registered and associated to establish a drying tail gas health registration association space, including: Extracting a first drying exhaust gas health record group according to the drying exhaust gas health record library, wherein the first drying exhaust gas health record group includes a first historical sludge drying treatment process source and first historical healthy drying exhaust gas monitoring data corresponding to the first historical sludge drying treatment process source; Performing a registration correlation analysis on the first historical sludge drying process source according to the sludge drying process source to generate a first registration correlation coefficient; If the first registration correlation coefficient is greater than / equal to a preset registration correlation coefficient, adding the first historical healthy dried exhaust gas data to the dried exhaust gas healthy registration correlation space; According to the preset registration correlation coefficient, the dried exhaust gas health record library is further registered and associated in combination with the sludge drying process source to generate the dried exhaust gas health registration correlation space.

5. The method according to claim 1, characterized in that According to the dry exhaust gas abnormal processing decision learning function, a dry exhaust gas abnormal processing decision channel is constructed, including: Load the dry exhaust gas anomaly detection vector record set and the dry exhaust gas anomaly processing decision record set; Based on the dry exhaust gas abnormality processing decision learning function, according to the dry exhaust gas abnormality detection vector record set and the dry exhaust gas abnormality processing decision record set, H dry exhaust gas abnormality processing decision sub-models are trained, wherein H is a positive integer greater than 1; Collecting the output data sets of the H dry exhaust gas abnormality processing decision sub-models to generate a decision integration model construction data set; Based on the dry exhaust gas abnormality processing decision learning function, constructing a data set and a dry exhaust gas abnormality processing decision record set according to the decision integration model, and training the dry exhaust gas abnormality processing decision integration model; The H dry exhaust gas abnormality processing decision sub-models are merged as parallel independent nodes to generate a dry exhaust gas abnormality processing decision parsing layer; The dry exhaust gas abnormality processing decision parsing layer is merged with the input layer of the dry exhaust gas abnormality processing decision integration model to generate the dry exhaust gas abnormality processing decision channel.

6. The method according to claim 5, characterized in that Based on the dry exhaust gas abnormality processing decision learning function, according to the dry exhaust gas abnormality detection vector record set and the dry exhaust gas abnormality processing decision record set, H dry exhaust gas abnormality processing decision sub-models are trained, including: According to H dry exhaust gas abnormal processing decision learners, extracting the h-th dry exhaust gas abnormal processing decision learner, where h belongs to H and h is a positive integer; Taking the dry exhaust gas abnormality detection vector record set as input data and the dry exhaust gas abnormality processing decision record set as output data, the h-th dry exhaust gas abnormality processing decision learner is trained and tested to obtain the h-th dry exhaust gas abnormality processing decision maker, and the h-th dry exhaust gas abnormality processing decision accuracy coefficient and the h-th dry exhaust gas abnormality processing decision error coefficient corresponding to the h-th dry exhaust gas abnormality processing decision maker; Inputting the h-th dry exhaust gas abnormal processing decision accuracy coefficient and the h-th dry exhaust gas abnormal processing decision error coefficient into the dry exhaust gas abnormal processing decision learning function to obtain the h-th dry exhaust gas abnormal processing decision learning index; Determining whether the hth dry exhaust gas abnormal processing decision learning index is greater than / equal to a preset dry exhaust gas abnormal processing decision learning index; If the h-th dry exhaust gas abnormality processing decision learning index is greater than / equal to the preset dry exhaust gas abnormality processing decision learning index, setting the h-th dry exhaust gas abnormality processing decision maker to the h-th dry exhaust gas abnormality processing decision sub-model; The hth dry exhaust gas abnormality processing decision sub-model is added to the H dry exhaust gas abnormality processing decision sub-models.

7. The method according to claim 1, characterized in that The dry exhaust gas abnormality processing decision learning function is: ； in, Characterize the learning index of decision-making for abnormal handling of dry exhaust gas, Characterize the decision learning factor of the predetermined dry exhaust abnormality processing, and, greater than 1, Characterizes the accuracy coefficient of the decision-making on abnormal treatment of dry exhaust gas, Characterizes the error coefficient of decision-making for abnormal handling of dry exhaust gas.

8. A system for managing abnormal exhaust gas from drying, characterized in that: The method comprises: A monitoring data source acquisition module, the monitoring data source acquisition module is used to obtain a dry exhaust gas monitoring data source according to a dry exhaust gas monitoring terminal; A processing process source loading module, the processing process source loading module is used to load the sludge drying processing process source corresponding to the drying tail gas monitoring data source, wherein the sludge drying processing process source includes a sludge component data set and a sludge drying processing process data set; An anomaly detection module, the anomaly detection module is used to perform expected anomaly detection on the drying tail gas monitoring data source according to the sludge drying treatment process source, and obtain a drying tail gas anomaly detection vector; An abnormality handling decision channel construction module, wherein the abnormality handling decision channel construction module is used to construct a dry exhaust gas abnormality handling decision channel according to a dry exhaust gas abnormality handling decision learning function; An abnormality handling decision acquisition module, the abnormality handling decision acquisition module is used to input the dry exhaust gas abnormality detection vector into the dry exhaust gas abnormality handling channel to generate a dry exhaust gas abnormality handling decision; An abnormality management module is used to perform abnormal management of dry exhaust gas according to the dry exhaust gas abnormality processing decision.

Images