Method and system for monitoring abnormity of intelligent heat supply system based on Internet of Things

Abstract

The invention discloses an intelligent heat supply system abnormity monitoring method and system based on the Internet of Things, and relates to the technical field of abnormity monitoring, and the method comprises the steps: collecting multi-dimensional sensor data of a target user in real time through the Internet of Things, extracting key features representing the operation state of a system from the data, and constructing a dynamic judgment interval in combination with the real-time heat use state of the user; and comparing with a normal range in a background database to identify abnormity. And for abnormal events, further extracting multi-dimensional information such as inlet water temperature, indoor and outdoor temperature difference, dynamic heat dissipation coefficient change rate and spatial correlation abnormality, and inputting the information into the pre-trained multi-classification model for deep analysis to obtain an abnormality diagnosis result containing prediction reasons, reason probability and suggestions. Thus, hidden faults can be captured, targeted processing suggestions can be provided, the abnormal recognition accuracy and the fault processing efficiency are remarkably improved, powerful support is provided for fine management and active operation and maintenance of a heat supply system, and the heat supply efficiency and the user experience are comprehensively improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of anomaly monitoring, and more particularly, to an anomaly monitoring method and system for a smart heating system based on the Internet of Things. BACKGROUND

[0002] Under the background of current global energy structure transformation and increasing demand for fine management, improving energy utilization efficiency has become a core issue throughout industrial production and livelihood services. Especially in heavy industrial cities with oil exploration and processing as the economic backbone, the energy consumption is huge, and the efficiency optimization of the entire energy industry chain is crucial. In the oil industry, from exploration and mining to refining and transportation, complex monitoring systems based on the Internet of Things are widely used. Through real-time and high-precision monitoring and anomaly diagnosis of widely distributed pipeline networks, pressure vessels and fluid states, production safety is ensured, energy loss is reduced, and intelligent operation and maintenance are realized. This mature monitoring concept and technical system in harsh industrial environments provides valuable reference for the modernization and upgrading of other key infrastructure in cities, especially for city central heating systems that also rely on large-scale pipe network systems and are related to energy terminal utilization efficiency and livelihood quality. Reducing the application of industrial-level intelligent monitoring solutions in the field of civil heating to address the high energy consumption, low efficiency and lagging fault response of traditional heating modes has become an inevitable trend to promote smart city construction and achieve energy saving and emission reduction targets.

[0003] Although introducing the Internet of Things technology into the heating system to build a smart heating system has become an industry consensus, most existing solutions remain at the level of remote meter reading and other primary applications, failing to realize the data analysis potential. The anomaly monitoring logic generally follows a simple fixed threshold alarm, i.e., setting fixed upper and lower limits for a single parameter such as temperature and pressure. This static approach ignores the dynamic and correlation characteristics of the heating system, making it prone to false positives due to normal operation fluctuations and unable to detect slow-growing hidden faults such as pipe blockage and valve leakage, leading to energy waste and problem accumulation. In addition, existing systems only provide phenomenon-level information even when an alarm is triggered, failing to diagnose the root cause of the fault, and ultimately still relying on manual troubleshooting, which is inefficient and cannot achieve proactive maintenance.

[0004] Therefore, there is an urgent need for an optimized anomaly monitoring method and system for a smart heating system based on the Internet of Things. SUMMARY

[0005] To solve the above technical problems, the present application is proposed.

[0006] According to an aspect of the present application, a method for monitoring anomalies of a smart heating system based on Internet of Things is provided, which comprises: obtaining real-time original sensor data of a target user, the real-time original sensor data comprising inlet water temperature, return water temperature, flow rate, indoor temperature and outdoor temperature; extracting core features from the real-time original sensor data to obtain a real-time dynamic heat dissipation coefficient; extracting a lower limit of the dynamic heat dissipation coefficient and an upper limit of the dynamic heat dissipation coefficient of the target user from a background database; determining whether the real-time dynamic heat dissipation coefficient is within the range of the lower limit of the dynamic heat dissipation coefficient and the upper limit of the dynamic heat dissipation coefficient, and if not, calculating an anomaly score and generating an anomaly event; in response to receiving the anomaly event, inputting the real-time original sensor data into a pre-trained multi-classification model to obtain an anomaly diagnosis result, the anomaly diagnosis result comprising a predicted cause, a cause probability and a suggestion; based on the anomaly diagnosis result, customizing and displaying alarm information; wherein the core features are extracted from the real-time original sensor data to obtain the real-time dynamic heat dissipation coefficient according to the following formula: ; wherein, is the heat capacity rate, is the specific heat capacity of water, is the flow rate, is the inlet water temperature, is the return water temperature, is the indoor-outdoor temperature difference, is the indoor temperature, is the outdoor temperature, and is the dynamic heat dissipation coefficient.

[0007] According to another aspect of the present application, a smart heating system anomaly monitoring system based on Internet of Things is provided, comprising: an original sensor data acquisition module configured to acquire real-time original sensor data of a target user, the real-time original sensor data comprising inlet water temperature, return water temperature, flow rate, indoor temperature and outdoor temperature; a core feature extraction module configured to extract core features from the real-time original sensor data to obtain a real-time dynamic heat dissipation coefficient; a dynamic heat dissipation coefficient retrieval module configured to extract a dynamic heat dissipation coefficient lower limit and a dynamic heat dissipation coefficient upper limit of the target user from a background database; an anomaly judgment module configured to judge whether the real-time dynamic heat dissipation coefficient is within the range of the dynamic heat dissipation coefficient lower limit and the dynamic heat dissipation coefficient upper limit, and if not, to calculate an anomaly score and generate an anomaly event; an anomaly diagnosis module configured to, in response to receiving the anomaly event, input the real-time original sensor data into a pre-trained multi-classification model to obtain an anomaly diagnosis result, the anomaly diagnosis result comprising a predicted cause, a cause probability and a suggestion; and an alarm information generation module configured to customize and display alarm information based on the anomaly diagnosis result; wherein the core features are extracted from the real-time original sensor data to obtain the real-time dynamic heat dissipation coefficient according to the following formula: ; wherein, is the heat capacity rate, is the specific heat capacity of water, is the flow rate, is the inlet water temperature, is the return water temperature, is the indoor-outdoor temperature difference, is the indoor temperature, is the outdoor temperature, and is the dynamic heat dissipation coefficient.

[0008] Compared with the prior art, the smart heating system anomaly monitoring method and system based on Internet of Things provided by the present application can capture hidden faults and provide targeted processing suggestions, significantly improve the accuracy of anomaly identification and the efficiency of fault handling, and provide strong support for the fine management and active operation and maintenance of the heating system, and comprehensively improve the heating efficiency and user experience. BRIEF DESCRIPTION OF DRAWINGS

[0009] The above and other objects, features and advantages of the present application will become more apparent from the following detailed description when taken in conjunction with the accompanying drawings in which:

[0010] Figure 1 Flow chart of the method for monitoring abnormality of the smart heating system based on Internet of Things according to the embodiment of the present application.

[0011] Figure 2 Data flow chart of the method for monitoring abnormality of the smart heating system based on Internet of Things according to the embodiment of the present application.

[0012] Figure 3 Flow chart of sub-step S3 of the method for monitoring abnormality of the smart heating system based on Internet of Things according to the embodiment of the present application.

[0013] Figure 4 Flow chart of sub-step S5 of the method for monitoring abnormality of the smart heating system based on Internet of Things according to the embodiment of the present application.

[0014] Figure 5 Block diagram of the system for monitoring abnormality of the smart heating system based on Internet of Things according to the embodiment of the present application. DETAILED DESCRIPTION

[0015] Embodiments of the present disclosure will be described more fully hereinafter with reference to the accompanying drawings. While several embodiments of the present disclosure are shown in the drawings, it is understood that the present disclosure can be embodied in various forms and should not be construed as being limited to the embodiments set forth herein; rather, these embodiments are provided so that this disclosure will be thorough and complete, and fully convey the scope of the present disclosure to those skilled in the art. It should be understood that the drawings and embodiments are only for illustrative purposes and should not be construed as limiting the scope of protection of the present disclosure.

[0016] In view of the above background art, the present application proposes a method for monitoring abnormality of a smart heating system based on Internet of Things. Figure 1 Flow chart of the method for monitoring abnormality of the smart heating system based on Internet of Things according to the embodiment of the present application. Figure 2 Data flow chart of the method for monitoring abnormality of the smart heating system based on Internet of Things according to the embodiment of the present application. Figure 1 and Figure 2As shown, the Internet of Things-based intelligent heating system anomaly monitoring method comprises the steps of: S1, acquiring real-time original sensor data of a target user, the real-time original sensor data including inlet water temperature, return water temperature, flow rate, indoor temperature and outdoor temperature; S2, extracting core features from the real-time original sensor data to obtain a real-time dynamic heat dissipation coefficient; S3, extracting a dynamic heat dissipation coefficient lower limit and a dynamic heat dissipation coefficient upper limit of the target user from a background database; S4, determining whether the real-time dynamic heat dissipation coefficient is within the range of the dynamic heat dissipation coefficient lower limit and the dynamic heat dissipation coefficient upper limit, and if not, calculating an anomaly score and generating an anomaly event; S5, in response to receiving the anomaly event, inputting the real-time original sensor data into a pre-trained multi-classification model to obtain an anomaly diagnosis result, the anomaly diagnosis result including a predicted cause, a cause probability and a suggestion; and S6, based on the anomaly diagnosis result, customizing and displaying alarm information.

[0017] In the above Internet of Things-based intelligent heating system anomaly monitoring method, the step S1, acquiring real-time original sensor data of a target user, the real-time original sensor data including inlet water temperature, return water temperature, flow rate, indoor temperature and outdoor temperature. It can be understood that, due to the dynamic complexity of the running state of the heating system and the heat transfer process, the present application acquires multi-dimensional sensor data in real time to obtain real-time original sensor data that can truly reflect the current running state of the heating system and the user's heating environment, wherein the inlet water temperature and the return water temperature reflect the heat loss of the heating medium, the flow rate data reflect the heat delivery efficiency, the indoor temperature is directly related to the user's actual heating experience, and the outdoor temperature affects the building heat dissipation degree. This provides a reliable data source for subsequent in-depth analysis and avoids misjudgment caused by data loss or lag.

[0018] In particular, in one possible embodiment, the implementation process of step S1 is as follows: first, temperature sensors are installed at the water inlet end and the water return end of the target user's heating pipeline to collect real-time water inlet temperature and water return temperature data. Second, a flow sensor is installed on the heating pipeline to monitor the flow of hot water through the pipeline, and an indoor temperature sensor is deployed at a suitable location in the user's room, and an outdoor temperature sensor is installed in an open outdoor area to obtain indoor and outdoor temperature data, respectively. All sensors are connected to an edge computing gateway through an Internet of Things communication module. The sensors collect data at intervals of 1 minute and send the raw data to the edge computing gateway through encrypted transmission. Then, the edge computing gateway performs preliminary verification on the received sensor data, such as removing outliers that are obviously beyond the physical range, and then transmits the verified real-time raw sensor data to the background database for storage through the 4G / 5G network. The background system reads the latest sensor data of the target user from the database through a timing task to form a real-time raw sensor data set for subsequent feature extraction and anomaly analysis module calls.

[0019] In the above-mentioned Internet of Things-based intelligent heating system anomaly monitoring method, step S2 extracts core features from the real-time raw sensor data to obtain a real-time dynamic heat dissipation coefficient. It should be understood that since the heating system operating state is affected by multiple parameters such as water inlet temperature, water return temperature, flow, and indoor and outdoor temperature difference, a single raw data cannot directly represent the dynamic characteristics of system heat dissipation. The dynamic heat dissipation coefficient can be constructed by fusing the above-mentioned parameters to reflect the actual heat transfer efficiency of the system. Therefore, the present application converts multi-dimensional raw data into a key indicator representing the heat dissipation efficiency of the system by extracting a real-time dynamic heat dissipation coefficient from real-time raw sensor data, thereby realizing the quantification of heating system operation anomalies and providing more accurate feature input for anomaly judgment.

[0020] Specifically, step S2 includes extracting core features from the real-time raw sensor data to obtain a real-time dynamic heat dissipation coefficient using the following formula: ; wherein, is the heat capacity, is the specific heat capacity of water, is the flow, is the water inlet temperature, is the water return temperature, is the indoor and outdoor temperature difference, is the indoor temperature, is the outdoor temperature, and is the dynamic heat dissipation coefficient.

[0021] In the above-mentioned IoT-based intelligent heating system anomaly monitoring method, step S3 extracts the dynamic heat dissipation coefficient lower limit and the dynamic heat dissipation coefficient upper limit of the target user from the background database. Specifically, by extracting the dynamic heat dissipation coefficient lower limit and upper limit of the target user, based on the statistical characteristics of the user historical heat consumption data, such as the mean and standard deviation obtained by Gaussian distribution fitting, a dynamic threshold interval is constructed, so that the system can compare the real-time calculated dynamic heat dissipation coefficient with the personalized normal range, thereby accurately identifying the abnormal state deviating from the actual operation rule of the user, avoiding false judgments caused by fixed thresholds, and laying a foundation for subsequent in-depth diagnosis and accurate processing of abnormal events.

[0022] In particular, in one specific embodiment, Figure 3 The flowchart of step S3 of the IoT-based intelligent heating system anomaly monitoring method according to the embodiments of the present application is shown in FIG. 3. Figure 3 As shown in FIG. 3, step S3 includes: S31, extracting the historical data of the dynamic heat dissipation coefficient of the target user from the background database; S32, processing the historical data of the dynamic heat dissipation coefficient using a Gaussian distribution fitting method to obtain a dynamic heat dissipation coefficient mean and a dynamic heat dissipation coefficient standard deviation; S33, subtracting the dynamic heat dissipation coefficient standard deviation from the dynamic heat dissipation coefficient mean to obtain a dynamic heat dissipation coefficient lower limit; and S34, adding the dynamic heat dissipation coefficient standard deviation to the dynamic heat dissipation coefficient mean to obtain a dynamic heat dissipation coefficient upper limit.

[0023] Specifically, step S31 extracts the historical data of the dynamic heat dissipation coefficient of the target user from the background database. It should be understood that due to the time-varying and individual difference of the user's heat consumption behavior, outdoor environment temperature and other factors, the real-time dynamic heat dissipation coefficient at a single time point cannot reflect the normal operation rule and fluctuation range. Based on this, the present application extracts the historical data of the dynamic heat dissipation coefficient of the target user, such as the time series data in the past month, and analyzes the central tendency and fluctuation amplitude of the daily operation based on the statistical characteristics of the data, and then constructs a dynamic normal range that adapts to the individual characteristics of the user.

[0024] In particular, in one possible embodiment, the implementation process of step S31 is as follows: first, a query request is sent to the database, and the query condition is set to extract the historical data of the dynamic heat dissipation coefficient of the target user in the past month, with a time granularity of 1 minute per record. After receiving the query instruction, the database quickly locates the corresponding record of the target user through indexing, and filters out all dynamic heat dissipation coefficient data with a time stamp within the current time and 30 days ago. After completing data extraction, the data is processed for deduplication and format unification through the extraction, transformation and loading process, forming an ordered historical data list of the dynamic heat dissipation coefficient, and storing it in a temporary cache area for subsequent processing and analysis.

[0025] Specifically, the step S32 uses a Gaussian distribution fitting method to process the historical data of the dynamic heat dissipation coefficient to obtain a dynamic heat dissipation coefficient average value and a dynamic heat dissipation coefficient standard deviation. It can be understood that the Gaussian distribution fitting method is a method of statistically modeling data based on normal distribution characteristics, which can determine the probability distribution parameters of a set of data through mathematical fitting means, so that the fitted theoretical distribution is as close as possible to the actual data distribution. The Gaussian distribution fitting method is used to model the historical data of the dynamic heat dissipation coefficient in the present application. The calculated dynamic heat dissipation coefficient average value reflects the central tendency of normal operation, and the dynamic heat dissipation coefficient standard deviation reflects the data fluctuation range, providing a quantitative basis for constructing the dynamic lower and upper limits, so that the normal range can adapt to the natural fluctuations of the data, thereby providing statistical support for setting a reasonable dynamic threshold in the subsequent process, and improving the accuracy of anomaly recognition.

[0026] In particular, in one possible embodiment, the implementation process of the step S32 is as follows: after obtaining the historical data list of the dynamic heat dissipation coefficient of the target user, a statistical analysis library is used to perform Gaussian distribution fitting on the data. Specifically, first, the historical data is imported into an array structure, for example, an array containing 30 days x 24 hours x 60 minutes = 43200 data. Then, the maximum likelihood estimation method is used to fit the array to calculate the optimal mean and standard deviation parameters. For example, assuming that the mean of the historical data is 250 W / ℃ and the standard deviation is 15 W / ℃, the fitted Gaussian distribution parameters are μ = 250 and σ = 15. Finally, the fitted average value and standard deviation are stored in the configuration table of the corresponding user in the database for subsequent calculation of the dynamic upper and lower limits.

[0027] Specifically, the steps S33 and S34, the dynamic heat dissipation coefficient average value is subtracted by the dynamic heat dissipation coefficient standard deviation to obtain a dynamic heat dissipation coefficient normal lower limit, and the dynamic heat dissipation coefficient average value is added to the dynamic heat dissipation coefficient standard deviation to obtain a dynamic heat dissipation coefficient normal upper limit. That is, by using the average value and the standard deviation of the dynamic heat dissipation coefficient, a personalized normal operation range is constructed, so that the system can compare the real-time dynamic heat dissipation coefficient with the range to accurately identify abnormal states that deviate from the actual operation rules of the user, and provide a dynamic and user-specific threshold standard for anomaly judgment. The dynamic heat dissipation coefficient normal upper limit generated in this way can adaptively reflect the reasonable fluctuation interval of the user's daily heat use mode and environmental changes, significantly improving the pertinence and accuracy of anomaly recognition.

[0028] In the above-mentioned abnormality monitoring method for the smart heating system based on the Internet of Things, in step S4, it is determined whether the real-time dynamic heat dissipation coefficient is within the range of the lower limit of the normal dynamic heat dissipation coefficient and the upper limit of the normal dynamic heat dissipation coefficient. If not, an abnormality score is calculated and an abnormality event is generated. Specifically, it is determined whether the real-time dynamic heat dissipation coefficient is within the range of the lower limit of the normal dynamic heat dissipation coefficient and the upper limit of the normal dynamic heat dissipation coefficient, which can accurately identify the abnormal state deviating from the real operation rule of the user. When the real-time value exceeds the range, it indicates that the system operation state has deviated from the normal fluctuation range, and there is an abnormal risk.

[0029] Specifically, in step S4, if the real-time dynamic heat dissipation coefficient is less than the lower limit of the normal dynamic heat dissipation coefficient, the abnormality score is calculated according to the following formula: If the real-time dynamic heat dissipation coefficient is greater than the upper limit of the normal dynamic heat dissipation coefficient, the abnormality score is calculated according to the following formula: Wherein, is the abnormality score, is the real-time dynamic heat dissipation coefficient, is the lower limit of the normal dynamic heat dissipation coefficient, and is the upper limit of the normal dynamic heat dissipation coefficient.

[0030] In the above-mentioned abnormality monitoring method for the smart heating system based on the Internet of Things, in step S5, in response to receiving the abnormality event, the real-time original sensor data is input into the pre-trained multi-classification model to obtain an abnormality diagnosis result, which includes a predicted reason, a reason probability and a suggestion. Specifically, the real-time original sensor data is input into the multi-classification model to extract features and construct an abnormal state vector, and then the classification ability of the model is used to obtain an abnormality diagnosis result including a predicted reason, a reason probability and a processing suggestion, so as to accurately identify different abnormal types such as pipe blockage, valve leakage and radiator failure, and provide accurate root analysis and disposal guidance for abnormal events.

[0031] In particular, in one specific embodiment, Figure 4 is a flowchart of step S5 of the abnormality monitoring method for the smart heating system based on the Internet of Things according to the embodiments of the present application. As Figure 4As shown, the step S5 comprises: S51, extracting the water inlet temperature, the water return temperature, the flow rate, the indoor temperature and the outdoor temperature from the real-time raw sensor data; S52, calculating the indoor-outdoor temperature difference based on the indoor temperature and the outdoor temperature; S53, calculating the change rate of the dynamic heat dissipation coefficient; S54, calculating the spatial correlation anomaly degree of the target user; S55, arranging the water inlet temperature, the water return temperature, the flow rate, the indoor temperature, the outdoor temperature, the indoor-outdoor temperature difference, the change rate of the dynamic heat dissipation coefficient and the spatial correlation anomaly degree into an abnormal state vector; and S56, inputting the abnormal state vector into the pre-trained multi-classification model to obtain the abnormal diagnosis result.

[0032] Specifically, the step S51 extracts the water inlet temperature, the water return temperature, the flow rate, the indoor temperature and the outdoor temperature from the real-time raw sensor data. It should be understood that the water inlet temperature, the water return temperature, the flow rate, the indoor temperature and the outdoor temperature are basic parameters reflecting the operation state of the heating system and the user's heating environment. The application converts the raw data collected by the sensor into structured features that can be used for deep analysis by extracting the above-mentioned raw parameters from the real-time raw sensor data, thereby providing basic data for subsequent construction of an abnormal state vector, so that the multi-classification model can predict the abnormal cause based on a complete feature set.

[0033] In particular, in a possible embodiment, the implementation process of the step S51 is as follows: in response to receiving the abnormal event, the real-time raw sensor data record of the target user is read from the real-time data cache area, which is in the JSON format and contains a timestamp and parameter values. For example, the record at 10:00 on June 27, 2025 is read: the water inlet temperature is 45°C, the water return temperature is 38°C, the flow rate Fr=200 kg / h, the indoor temperature is 20°C, and the outdoor temperature is 5°C. Then, by parsing the JSON field, the parameter values are stored in temporary variables, and the timestamp is recorded synchronously. After the extraction is completed, the data is subjected to validity check, such as checking whether the temperature is within the sensor range of -20°C~100°C and whether the flow rate is positive. If the check is passed, the subsequent processing flow is entered, otherwise the data is marked as invalid and reacquired.

[0034] Specifically, the steps S52 and S53 calculate the indoor-outdoor temperature difference based on the indoor temperature and the outdoor temperature, and calculate the change rate of the dynamic heat dissipation coefficient. It should be understood that by calculating the indoor-outdoor temperature difference, the indoor-outdoor temperature is converted into a quantitative index reflecting the building heat load, and at the same time, the change rate of the dynamic heat dissipation coefficient is combined, so as to provide the multi-classification model with multi-dimensional features containing spatial heat environment and time trend, so that the model can distinguish between transient environmental fluctuations and gradual faults, and accurately identify hidden abnormalities, such as slow decline of radiator efficiency.

[0035] In particular, in one possible embodiment, the steps S52 and S53 are implemented as follows: first, the current indoor temperature and outdoor temperature are extracted from real-time sensor data , and a temperature difference calculation is performed .

[0036] Before calculation, the temperature data needs to be checked for validity, such as checking whether it is within the range of -20°C to 100°C, and the DHC historical data needs to be checked for timestamp consistency, ensuring that the interval between the previous and subsequent data is 1 hour. The temperature difference and change rate results are stored in a temporary feature table, with fields including user ID, timestamp, temperature difference value, and change rate, waiting for subsequent processing.

[0037] Specifically, the step S54 calculates the spatial correlation abnormality degree of the target user. It should be understood that the spatial correlation abnormality degree is a quantitative indicator for measuring the correlation degree of the target user's abnormal event in the spatial dimension. By counting the distribution density of the same type of abnormal event of other users in the same unit building, it can be determined whether the abnormality is an individual fault or a systematic problem, thereby significantly reducing the troubleshooting range, improving positioning efficiency, and reducing manual troubleshooting costs.

[0038] In particular, in one specific embodiment, the step S54 includes: querying the unit building ID in which the target user is located based on the user ID of the target user; querying all user IDs under the unit building ID except the target user; counting the total number of user IDs with the same abnormal event as the target user as the total number of abnormal neighbors; and dividing the total number of abnormal neighbors by the total number of users to obtain the spatial correlation abnormality degree. It should be understood that users in the same unit building share the same riser and have a direct hydraulic relationship. When a systematic problem occurs, such as a valve misoperation or a main pipeline failure, it will affect multiple users in the same unit within a short period of time, resulting in a concentration of similar abnormal events. Therefore, by querying the unit building ID in which the target user is located and the user IDs of other users in the unit building ID, the total number of abnormal neighbors of similar abnormal events is counted, the distribution density of abnormal events in the spatial dimension is quantified, and the spatial correlation abnormality degree indicator is generated, which can determine whether the abnormality has spatial correlation, thereby distinguishing between individual device faults and systematic faults, providing more targeted troubleshooting direction for maintenance personnel, reducing the manual troubleshooting range, and improving processing efficiency.

[0039] In particular, in one possible embodiment, the step S54 is implemented as follows: for example, the user ID of the target user is "U-00123", first, the unit building ID corresponding to the user is queried from the user table as "B-3 unit". Then, all user IDs under "B-3 unit" are queried, and after excluding the target user "U-00123", the other user IDs in the unit are obtained as "U-00124" to "U-00139", a total of 16 users. Then, a query is performed in the abnormal event history record to filter out records with the unit building ID as "B-3 unit", the user ID not equal to "U-00123", the event timestamp within the past 2 hours, and the same abnormal type as the target user. In one possible embodiment, the user IDs that meet the conditions are "U-00125", "U-00127", "U-00130", and "U-00135", a total of 4. Finally, the total number of abnormal neighbors 4 is divided by the total number of users in the unit 16 to obtain a spatial correlation anomaly degree of 0.25. This value indicates that 25% of the users in the unit have the same type of abnormality in the same time period, and the system judges that there may be a standpipe flow abnormality or a local pipe blockage, etc. Systematic problem, providing key spatial features for subsequent multi-classification model diagnosis.

[0040] Specifically, in the step S55, the water inlet temperature, the water return temperature, the flow, the indoor temperature, the outdoor temperature, the indoor-outdoor temperature difference, the change rate of the dynamic heat dissipation coefficient, and the spatial correlation anomaly degree are arranged into an abnormal state vector. Specifically, by sequentially arranging the multi-dimensional features into an abnormal state vector, a standardized feature matrix that meets the input requirements of the pre-trained multi-classification model is constructed, so that the model can realize deep semantic understanding of abnormal events based on basic parameters such as water inlet temperature and water return temperature, combined with derived indexes such as indoor-outdoor temperature difference, change rate of dynamic heat dissipation coefficient, and spatial correlation features, and provide multi-dimensional data support for abnormal root cause analysis. It should be understood that the abnormal state vector integrates thermodynamic parameters, environmental parameters, dynamic change features, and spatial propagation features of the heating system operation. After inputting the multi-classification model, the recognition ability of hidden faults can be significantly improved, for example, when the water inlet temperature is normal but the water return temperature is low, the flow is reduced, the dynamic heat dissipation coefficient is decreased, and the spatial correlation anomaly degree is high, the model can accurately judge that the standpipe is locally blocked, rather than a single user equipment fault, thereby providing accurate fault positioning basis for operation and maintenance.

[0041] In particular, in one possible embodiment, the implementation process of step S55 is as follows: first, the water inlet temperature is 45°C, the return water temperature is 38°C, the flow rate is 200 kg / h, the indoor temperature is 20°C, and the outdoor temperature is 5°C. Second, the indoor-outdoor temperature difference is calculated as 20°C-5°C=15°C, and assuming that the current DHC value is 108.9 W / °C and the DHC value one hour ago is 340 W / °C, the change rate of the dynamic heat dissipation coefficient is calculated as (108.9-340) / 340=-0.68. Next, the spatial correlation anomaly degree is obtained as 0.25, and as in the previous example, 4 of the 16 units in the unit building have the same anomaly. Finally, the above water inlet temperature, return water temperature, flow rate, indoor temperature, outdoor temperature, indoor-outdoor temperature difference, change rate of dynamic heat dissipation coefficient, and spatial correlation anomaly degree are sequentially arranged as the anomaly state vector [45, 38, 200, 20, 5, 15, -0.68, 0.25].

[0042] Specifically, in step S56, the anomaly state vector is input into the pre-trained multi-classification model to obtain the anomaly diagnosis result. Specifically, after the anomaly state vector is input into the model, the model can quickly calculate the probability values of each possible cause based on the learned knowledge, such as the probability distribution of “water inlet temperature normal but return water temperature low + flow rate decreased + dynamic heat dissipation coefficient decreased + high spatial correlation anomaly degree” corresponding to “standpipe valve half open”, and output the predicted probability of different abnormal causes and the corresponding processing suggestions, and preferentially recommend the high-probability cause and processing measures, thereby providing precise fault positioning basis for the operation and maintenance personnel.

[0043] In particular, in one possible embodiment, the step S56 is implemented as follows: first, the abnormal state vector is input into the pre-trained multi-classification model, which is an encoder-multi-classification head architecture. The encoder part is responsible for extracting high-level and robust feature representations from the 8-dimensional input features, which can be composed of multiple layers of fully connected neural networks, each followed by a ReLU activation function to enhance non-linear representation capabilities. The multi-classification head part receives the features output by the encoder and maps them to the probability distribution of the 5 abnormal reasons, such as half-open standpipe valve, slight pipe blockage, radiator dust, user end valve not fully open, and heat source temperature fluctuation. During the forward propagation of the model, the input features are first processed by the encoder to learn and abstract features layer by layer, extracting discriminative intermediate feature representations. Then, these intermediate features are fed into the multi-classification head, which passes through the internal fully connected layers and Softmax activation function, finally outputting the probability distribution of each abnormal reason. For example, the output result is: half-open standpipe valve probability 0.85, slight pipe blockage probability 0.10, radiator dust probability 0.03, user end valve not fully open probability 0.01, and heat source temperature fluctuation probability 0.01. Taking the half-open standpipe valve corresponding to the maximum probability value as the predicted reason, the abnormal diagnosis result is output.

[0044] It is worth noting that when the abnormal state vector is input into the multi-classification model for classification, since the abnormal event state has been determined, the overall prior state of the classification mapping is determined, and in the process of determining the prior state, the dynamic heat dissipation coefficient is used as the core index for state determination, so the change rate of the dynamic heat dissipation coefficient becomes the focused prior value, on the other hand, since the spatial correlation anomaly degree of the target user will significantly affect the nature of the abnormal diagnosis result, the spatial correlation anomaly degree of the target user can also be classified as a focused prior value, thus the improvement of the multi-classification task is a multi-classification optimization based on partial prior value focusing under the overall prior state certainty.

[0045] Specifically, in another possible embodiment, inputting the abnormal state vector into the pre-trained multi-classification model to obtain the abnormal diagnosis result includes: first, inputting the abnormal state vector into the encoder of the pre-trained multi-classification model to obtain an abnormal state latent encoding vector. It should be understood that the encoder of the pre-trained multi-classification model consists of a multi-layer fully connected neural network, with each layer followed by a ReLU activation function to enhance nonlinear expressive power. Its core responsibility is to extract high-level, robust feature representations from the original 8-dimensional input features (including inlet water temperature, return water temperature, flow rate, indoor temperature, outdoor temperature, indoor-outdoor temperature difference, rate of change of dynamic heat dissipation coefficient, and spatial correlation anomaly degree). Specifically, although the sensor data and derived features of the heating system are abundant, they may contain redundancy and complex nonlinear relationships, making direct use for diagnosis potentially inefficient. Transforming them into a more compact and discriminative latent encoding vector through the encoder can effectively filter out noise and capture the essential characteristics of abnormal states, such as deep patterns of overall system heat loss or local circulation obstacles.

[0046] Then, the hidden encoding vector of the abnormal state Local correlation aggregation is performed to obtain the local correlation aggregation encoding vector of the abnormal state, i.e., one-dimensional convolution is performed to obtain... ,in, This refers to one-dimensional convolution, where the kernel length is greater than a predetermined threshold, such as greater than five. In other words, many abnormal phenomena in heating systems, such as partial pipe blockage or radiator dust accumulation, often exhibit specific local correlation patterns in sensor data. For example, there may be specific linkages between inlet water temperature, return water temperature, and flow rate, or a correlation between the indoor / outdoor temperature difference and the rate of change of the heat dissipation coefficient. Through local correlation aggregation (e.g., using one-dimensional convolution operations with sufficiently long kernels), local focusing physical behavior can be simulated during weighted aggregation based on mid-range neighborhoods to fully capture the correlation patterns of some focusing prior feature values. This allows for the discovery of deep connections between local features, improving the perception of specific physical behaviors (such as increased local thermal resistance) and facilitating local classification decisions.

[0047] Next, the local correlation aggregation encoding vector of the abnormal state is subjected to propagation and diffusion based on focusing prior to obtain the a priori propagation and diffusion encoding vector of the abnormal state, that is, for eigenvalues and Perform the following operation, which is represented as: ;in, and These respectively represent the first and second parts of the local correlation aggregation encoding vector of the abnormal state. The and the first 1 eigenvalue, The first element representing the prior propagation diffusion encoding vector of the abnormal state Several eigenvalues. It should be understood that in the diagnosis of heating system anomalies, certain indicators are considered highly focused prior information, having a decisive impact on the diagnostic results. For example, a significant rate of change in the dynamic heat dissipation coefficient directly points to the overall thermal efficiency of the system, while a high degree of spatial correlation anomaly (multiple users in the same building exhibiting the same anomaly) strongly suggests a systemic fault such as a riser or main pipe, rather than a problem with a single user's equipment. These focused priors require stronger diagnostic influence. By using differential signal gradient propagation, these focused prior signals are diffused and strengthened in the locally correlated aggregated encoding vector, allowing this key information to more profoundly influence the representation of other features. Thus, based on a locally preset focused pattern, the differential gradient propagation of local focused prior eigenvalues ​​through an extension process amplifies the weight and influence of these key prior information in the overall feature representation, guiding the model to focus more on these highly diagnostic features, such as distinguishing between individual radiator faults and building riser problems.

[0048] Next, the overall prior comparison factor between the anomalous state prior propagation diffusion encoding vector and the anomalous state latent encoding vector is calculated, and is expressed as: ,in, This represents the prior propagation diffusion encoding vector of the abnormal state. The hidden encoding vector representing the abnormal state Norm 1 This represents vector subtraction. The L2 norm of a vector. This represents the overall prior comparison factor. It should be understood that while the anomalous state prior propagation and diffusion encoding vector and the anomalous state latent encoding vector can enhance specific information, for complex heating systems, overemphasizing local or single priors while detaching from the overall context may lead to misjudgment. Therefore, it is necessary to compare these processed vectors with the original, more general latent encoding vector to ensure that the enhanced feature representation is consistent with the overall anomalous state of the heating system. In this way, by using the overall prior state as a deterministic reference and focusing on the multidimensional response of the prior through comparison matching, while still using the prior state as an expected reference to provide a response reference point, the prior stability of the overall classification decision is promoted. This provides a quantified overall prior comparison factor that reflects the difference or consistency between the feature representation after prior propagation and diffusion and the original latent representation, providing important guidance for subsequent feature fusion and ensuring the stability and consistency of the diagnostic process.

[0049] Then, the abnormal state prior propagation diffusion encoding vector and the abnormal state local correlation aggregation encoding vector are fused based on the overall prior comparison factor to obtain an optimized abnormal state implicit encoding vector, denoted as: wherein, represents point multiplication by position, that is, the eigenvalue of each position of is added to , represents vector addition, represents the abnormal state local correlation aggregation encoding vector, represents the optimized abnormal state implicit encoding vector. It should be understood that the heating system anomaly is a multi-faceted problem that needs to be considered comprehensively. The abnormal state has been encoded and enhanced from different aspects (original feature extraction, local correlation, and focused prior propagation) in the foregoing. Now, it is necessary to effectively integrate the information at different levels to form a final feature representation that contains comprehensive information and highlights key priors. By taking the overall prior comparison factor as a weight or modulation factor, the vector after prior propagation diffusion and the vector after local correlation aggregation can be intelligently fused, so that the final optimized vector can take into account both local details and global prior guidance, thereby more comprehensively reflecting the abnormal condition of the heating system. In this way, the classification decision is promoted by focusing on partial prior values and determining the overall prior state, thereby improving the classification effect of the multi-classification model.

[0050] Finally, the optimized abnormal state implicit encoding vector is input into the multi-classification head of the pre-trained multi-classification model to obtain the abnormal diagnosis result. It should be understood that the multi-classification head part receives the optimized abnormal state implicit encoding vector and maps it to the probability distribution of the pre-set 5 types of heating system abnormal reasons, such as half-open stand pipe valve, slight pipe blockage, radiator dust, user end valve not fully open, and heat source temperature fluctuation. Its internal usually contains a fully connected layer and a Softmax activation function, which is responsible for calculating the probability of each abnormal reason.

[0051] In the above-mentioned intelligent heating system anomaly monitoring method based on Internet of Things, the step S6 is to customize and display alarm information based on the abnormal diagnosis result. Specifically, by structurally analyzing and visually presenting the abnormal diagnosis result, the abstract model output is converted into standardized alarm information containing fault type, confidence, and disposal suggestion, and different abnormal levels are combined to adopt a differentiated display strategy, such as color identification and priority sorting, to provide precise fault handling guidance for operation and maintenance personnel and improve the automation and intelligent level of heating system operation and maintenance.

[0052] In particular, in one possible embodiment, the implementation of step S6 is as follows: after the multi-classification model outputs the abnormal diagnosis result for the target user U-00123, the abnormal diagnosis result is first structured and parsed. The result shows that the predicted cause is "standpipe valve half open" with a probability of 85%, the secondary cause is "slight pipe blockage" with a probability of 10%, and the suggestion of "checking the standpipe valve opening degree of unit B-3" is attached. At the same time, the key parameters in the abnormal state vector include the inlet water temperature of 45°C, the return water temperature of 38°C, the flow of 200 kg / h, the indoor temperature of 20°C, the outdoor temperature of 5°C, the indoor and outdoor temperature difference of 15°C, the dynamic heat dissipation coefficient change rate of -0.68, and the spatial correlation abnormality degree of 0.25. The key information such as the main cause prediction, the probability value, the correlation suggestion, the secondary cause, the abnormal score, and the spatial correlation abnormality degree is extracted to provide data support for subsequent alarm customization.

[0053] In determining the alarm level and display strategy, according to the conditions that the main cause probability is 85% and the abnormal score is 0.32, the alarm is determined as "emergency alarm". The visual identifier of red background, valve icon and exclamation mark is adopted, and the highest priority is set to make it displayed on the top of the home page of the operation and maintenance platform to remind the operation and maintenance personnel. This differentiated display strategy ensures that the emergency failure can attract attention at the first time, and avoids the processing delay caused by the confusion of information levels.

[0054] Subsequently, the system generates customized alarm content according to the preset template. The content takes "emergency alarm: B-3 unit user U-00123 heat supply abnormality" as the title, and presents the predicted cause "standpipe valve half open (confidence 85%), abnormal characteristics "inlet water temperature 45°C, return water temperature 38°C, flow 200 kg / h, dynamic heat dissipation coefficient 108.9 W / °C, 25% of users in the unit have the same abnormality", processing suggestion "immediately check the standpipe valve opening degree of unit B-3, and suggest opening the valve to restore normal heat flow", and auxiliary information "the secondary cause is slight pipe blockage, and whether there is debris deposition in the pipe can be checked synchronously". The template-based content generation method not only ensures the integrity of the information, but also facilitates the operation and maintenance personnel to quickly obtain the key instructions.

[0055] In terms of multi-terminal adaptive display, the operation and maintenance management platform presents the complete alarm content in a card-style pop-up window, accompanied by a nearly 1-hour DHC value change curve and a unit user abnormal distribution heat map. The left curve intuitively shows the deviation of the real-time DHC value from the normal interval, and the right heat map marks the abnormal user location in red. Mouse hovering can view detailed data, providing visual fault analysis assistance for operation and maintenance personnel. The mobile APP pushes a simplified version of the notification, with the title "

Emergency

[0056] After the alarm information is generated, the system stores it in the background database and performs multi-dimensional association. The association content includes target user ID, diagnosis time, abnormal state vector, and other basic information, as well as automatically retrieving the unit's historical processing records, such as the case of a half-open standpipe valve due to rust three months ago, prompting operation and maintenance personnel to focus on checking the valve rust condition. When the operation and maintenance personnel click the "handled" button, they need to enter the actual processing result, such as "the valve is not fully open, has been adjusted to fully open state", and through automatic comparison of the predicted cause and the actual cause, the data is included in the model training set to continuously optimize the subsequent diagnosis accuracy.

[0057] In the alarm closed-loop management link, the system regularly checks the alarm processing status. If it is not processed for more than 30 minutes, it is automatically upgraded and pushed to the operation and maintenance supervisor. The processing status displays the estimated completion time, the handled status records the processing time, the processor, and pushes the recovery notification to the user-side applet. For example, after the operation and maintenance personnel complete the valve adjustment at 13:00, the system pushes the "heating has returned to normal" notification to the user at 13:05, and marks the alarm status as green "solved" on the operation and maintenance platform. Through this whole-process closed-loop management, it ensures that every alarm can be followed up and processed in time, avoiding the omission of abnormal events, thereby effectively improving the stability and reliability of the heating system.

[0058] In summary, the abnormality monitoring method of the smart heating system based on the Internet of Things according to the embodiments of the present application is illustrated, which collects multi-dimensional sensor data of a target user in real time through the Internet of Things, extracts key features representing the system operation state therefrom, and constructs a dynamic judgment interval in combination with the real-time heat use state of the user, and identifies abnormalities by comparing with the normal range in the background database. For abnormal events, further extract multi-dimensional information such as inlet water temperature, indoor and outdoor temperature difference, dynamic heat dissipation coefficient change rate, and spatial correlation anomaly degree, and input into a pre-trained multi-classification model for deep analysis to obtain abnormal diagnosis results including predicted reasons, reason probability and suggestions. In this way, hidden faults can be captured and targeted treatment suggestions can be provided, significantly improving the accuracy of abnormality identification and fault handling efficiency, providing strong support for fine management and active operation and maintenance of the heating system, and comprehensively improving the heating efficiency and user experience.

[0059] Figure 5 The block diagram of the abnormality monitoring system of the smart heating system based on the Internet of Things according to the embodiments of the present application is shown. As shown in Figure 5 The abnormality monitoring system of the smart heating system based on the Internet of Things according to the embodiments of the present application 100 includes: an original sensor data acquisition module 110 for acquiring real-time original sensor data of a target user, the real-time original sensor data including inlet water temperature, return water temperature, flow, indoor temperature and outdoor temperature; a core feature extraction module 120 for extracting core features from the real-time original sensor data to obtain a real-time dynamic heat dissipation coefficient; a dynamic heat dissipation coefficient retrieval module 130 for extracting a dynamic heat dissipation coefficient normal lower limit and a dynamic heat dissipation coefficient normal upper limit of the target user from a background database; an abnormality judgment module 140 for judging whether the real-time dynamic heat dissipation coefficient is within the range of the dynamic heat dissipation coefficient normal lower limit and the dynamic heat dissipation coefficient normal upper limit, if not, calculating an abnormality score and generating an abnormal event; an abnormality diagnosis module 150 for inputting the real-time original sensor data into a pre-trained multi-classification model to obtain an abnormality diagnosis result in response to receiving the abnormal event, the abnormality diagnosis result including predicted reasons, reason probability and suggestions; an alarm information generation module 160 for customizing and displaying alarm information based on the abnormality diagnosis result.

[0060] As described above, the IOT-based intelligent heating system anomaly monitoring system 100 according to the embodiments of the present application can be implemented in various wireless terminals. In one possible implementation, the IOT-based intelligent heating system anomaly monitoring system 100 according to the embodiments of the present application can be integrated into a wireless terminal as a software module and / or a hardware module. For example, the IOT-based intelligent heating system anomaly monitoring system 100 can be a software module in the operating system of the wireless terminal, or can be an application program developed for the wireless terminal; of course, the IOT-based intelligent heating system anomaly monitoring system 100 can also be one of the many hardware modules of the wireless terminal.

[0061] Alternatively, in another example, the IOT-based intelligent heating system anomaly monitoring system 100 and the wireless terminal can also be separate devices, and the IOT-based intelligent heating system anomaly monitoring system 100 can be connected to the wireless terminal through a wired and / or wireless network, and transmit interactive information in an agreed data format.

[0062] Here, those skilled in the art can understand that the specific operations of each step in the above IOT-based intelligent heating system anomaly monitoring system have been described in detail above with reference to the description of the IOT-based intelligent heating system anomaly monitoring method of Figures 1 to 4 , and therefore the repeated description thereof will be omitted.

Claims

1. An abnormality monitoring method for an Internet of Things (IoT)-based intelligent heating system, characterized by comprising: The method comprises: obtaining real-time raw sensor data of a target user, the real-time raw sensor data comprising inlet water temperature, return water temperature, flow rate, indoor temperature, and outdoor temperature; extracting core features from the real-time raw sensor data to obtain a real-time dynamic heat dissipation coefficient; extracting a dynamic heat dissipation coefficient lower limit and a dynamic heat dissipation coefficient upper limit of the target user from a background database; determining whether the real-time dynamic heat dissipation coefficient is within the range of the dynamic heat dissipation coefficient lower limit and the dynamic heat dissipation coefficient upper limit, and if not, calculating an anomaly score and generating an anomaly event; In response to receiving the abnormal event, inputting the real-time raw sensor data into a pre-trained completed multi-classification model to obtain an abnormal diagnosis result, the abnormal diagnosis result including a predicted cause, a cause probability and a suggestion; based on the abnormal diagnosis result, customizing and displaying alarm information; wherein core features are extracted from the real-time raw sensor data to obtain a real-time dynamic heat dissipation coefficient according to the following formula: ; wherein, is a heat capacity rate, is a specific heat capacity of water, is a flow rate, is an inlet water temperature, is a return water temperature, is an indoor-outdoor temperature difference, is an indoor temperature, is an outdoor temperature, and is a dynamic heat dissipation coefficient. 2.The IoT-based intelligent heat supply system anomaly monitoring method of claim 1, wherein, extracting a dynamic heat dissipation coefficient lower limit and a dynamic heat dissipation coefficient upper limit of the target user from a background database, comprising: extracting historical data of the dynamic heat dissipation coefficient of the target user from the background database; processing the historical data of the dynamic heat dissipation coefficient using a Gaussian distribution fitting method to obtain a dynamic heat dissipation coefficient mean and a dynamic heat dissipation coefficient standard deviation; subtracting the dynamic heat dissipation coefficient standard deviation from the dynamic heat dissipation coefficient mean to obtain a dynamic heat dissipation coefficient lower limit; adding the dynamic heat dissipation coefficient standard deviation to the dynamic heat dissipation coefficient mean to obtain a dynamic heat dissipation coefficient upper limit. 3.The IoT-based intelligent heat supply system anomaly monitoring method of claim 2, wherein, The historical data of the dynamic heat dissipation coefficient is the dynamic heat dissipation coefficient of the target user in the past month. 4.The IoT-based intelligent heat supply system anomaly monitoring method of claim 1, wherein, determining whether the real-time dynamic heat dissipation coefficient is within a range of a lower limit of the normal dynamic heat dissipation coefficient and an upper limit of the normal dynamic heat dissipation coefficient, and if not, calculating an abnormality score and generating an abnormality event, including: if the real-time dynamic heat dissipation coefficient is less than the lower limit of the normal dynamic heat dissipation coefficient, calculating the abnormality score in accordance with the following formula, wherein the formula is: ; if the real-time dynamic heat dissipation coefficient is greater than the upper limit of the normal dynamic heat dissipation coefficient, calculating the abnormality score in accordance with the following formula, wherein the formula is: ; wherein is the abnormality score, is the real-time dynamic heat dissipation coefficient, is the lower limit of the normal dynamic heat dissipation coefficient, and is the upper limit of the normal dynamic heat dissipation coefficient. 5.The IoT-based intelligent heating system anomaly monitoring method of claim 4, wherein, In response to receiving the anomaly event, inputting the real-time raw sensor data into a pre-trained multi-classification model to obtain an anomaly diagnosis result, the anomaly diagnosis result comprising a predicted cause, a cause probability, and a suggestion, comprising: extracting inlet water temperature, return water temperature, flow rate, indoor temperature, and outdoor temperature from the real-time raw sensor data; calculating an indoor-outdoor temperature difference based on the indoor temperature and the outdoor temperature; calculating a dynamic heat dissipation coefficient change rate; calculating a spatial correlation anomaly degree of the target user; arranging the inlet water temperature, return water temperature, flow rate, indoor temperature, outdoor temperature, indoor-outdoor temperature difference, dynamic heat dissipation coefficient change rate, and spatial correlation anomaly degree into an anomaly state vector; inputting the anomaly state vector into the pre-trained multi-classification model to obtain the anomaly diagnosis result. 6.The IoT-based intelligent heat supply system anomaly monitoring method of claim 5, wherein, The abnormal state vector is input into the pre-trained multi-classification model to obtain the abnormal diagnosis result, including: inputting the abnormal state vector into an encoder of the pre-trained multi-classification model to obtain an abnormal state implicit encoding vector; performing local correlation aggregation on the abnormal state implicit encoding vector to obtain an abnormal state local correlation aggregation encoding vector; performing focus prior-based propagation diffusion on the abnormal state local correlation aggregation encoding vector to obtain an abnormal state prior propagation diffusion encoding vector; calculating an overall prior comparison factor between the abnormal state prior propagation diffusion encoding vector and the abnormal state implicit encoding vector; fusing the abnormal state prior propagation diffusion encoding vector and the abnormal state local correlation aggregation encoding vector based on the overall prior comparison factor to obtain an optimized abnormal state implicit encoding vector; inputting the optimized abnormal state implicit encoding vector into a multi-classification head of the pre-trained multi-classification model to obtain the abnormal diagnosis result. 7.The IoT-based intelligent heating system anomaly monitoring method of claim 1, wherein, The spatial correlation abnormality degree of the target user is calculated, including: based on the user ID of the target user, querying the unit building ID where the user ID is located; querying all user IDs under the unit building ID except the target user; counting the total number of user IDs with the same abnormal event as the target user in all user IDs as the total number of abnormal neighbors; dividing the total number of abnormal neighbors by the total number of users to obtain the spatial correlation abnormality degree.

8. An Internet of Things-based intelligent heating system anomaly monitoring system, characterized in that, Comprise: An original sensor data acquisition module is configured to acquire real-time original sensor data of a target user, wherein the real-time original sensor data includes water inlet temperature, water return temperature, flow rate, indoor temperature, and outdoor temperature; A core feature extraction module is configured to extract core features from the real-time original sensor data to obtain a real-time dynamic heat dissipation coefficient; A dynamic heat dissipation coefficient retrieval module is configured to extract a dynamic heat dissipation coefficient normal lower limit and a dynamic heat dissipation coefficient normal upper limit of the target user from a background database; An abnormality judgment module is configured to judge whether the real-time dynamic heat dissipation coefficient is within the range of the dynamic heat dissipation coefficient normal lower limit and the dynamic heat dissipation coefficient normal upper limit, and if not, to calculate an abnormal score and generate an abnormal event; An abnormality diagnosis module is configured to input the real-time raw sensor data into a pre-trained multi-classification model to obtain an abnormality diagnosis result in response to receiving the abnormal event, the abnormality diagnosis result including a predicted cause, a cause probability, and a suggestion; an alarm information generation module is configured to customize and display alarm information based on the abnormality diagnosis result; wherein core features are extracted from the real-time raw sensor data to obtain a real-time dynamic heat dissipation coefficient according to the following formula: ; wherein, is a heat capacity rate, is a specific heat capacity of water, is a flow rate, is an inlet water temperature, is a return water temperature, is an indoor-outdoor temperature difference, is an indoor temperature, is an outdoor temperature, and is a dynamic heat dissipation coefficient.

Images