Generator temperature signal intelligent monitoring and leakage early warning method and system

Abstract

The invention relates to the technical field of generator state monitoring, in particular to a generator temperature signal intelligent monitoring and leakage early warning method and system, and the method comprises the following steps: extracting a disturbance moment and temperature sensing response difference value, calculating time difference change by combining numbers, sorting heat flux difference values, judging the direction, and screening a pressure difference and temperature rise consistent section. And overlapping the humidity sequence and the number overlapping area, and generating a monitoring and leakage early warning identification graph. According to the method, a time chain is constructed by extracting a temperature sensing node response sequence, a temperature disturbance starting point area is positioned, a heat diffusion path is restored in combination with a multi-cycle heat flux change direction, a cooling abnormal section is screened by using a pressure difference and temperature rise synchronous difference value, and a leakage risk number is identified by overlapping a humidity change jump feature. Accurate positioning of an abnormal source, dynamic tracking of a heat conduction path and multi-parameter linkage identification of a cooling system state are achieved, and the time sequence relevance of a monitoring chain, the continuity of path identification, the spatial accuracy of abnormal judgment and the stability and reliability of early warning response are enhanced.

Description

Technical Field

[0001] This invention relates to the field of generator condition monitoring technology, and in particular to a method and system for intelligent monitoring and early warning of generator temperature signals and leakage. Background Technology

[0002] The field of generator condition monitoring technology involves core aspects such as real-time perception, data acquisition, analysis, and anomaly identification of the operating status of power generation equipment in power systems. It covers the measurement and diagnostic analysis of multiple physical quantities, including temperature, voltage, current, vibration, noise, and lubricating oil quality. The technical system mainly includes sensor deployment, signal acquisition, electrical protection, information identification, and trend assessment. It is widely used to improve the reliability and safety of power equipment operation, prevent the escalation of faults, and assist in operation and maintenance decisions, making it an important component of intelligent and automated operation management of power equipment. Traditional intelligent monitoring and leakage early warning methods for generator temperature signals involve using temperature sensors such as thermistors or thermocouples to measure the temperature of key components such as stator windings, rotors, and bearings during generator operation status monitoring. The acquisition module converts analog signals into digital signals, and temperature anomaly identification and preliminary early warning are achieved based on set temperature thresholds or difference judgment standards. Simultaneously, possible cooling system leaks are indirectly inferred through cooling medium pressure detection or conductor insulation resistance changes. The entire process is typically based on a fixed wiring system and compares temperature rise data with a preset model, relying heavily on experience-based settings and single-parameter judgment methods.

[0003] Existing technologies rely on point-based temperature measurement, depending solely on set thresholds or single differences for state identification. They lack systematic analysis of the temperature change sequence, time gradient, and correlation between data sequences, making it impossible to continuously track the dynamic process of heat diffusion. The lag in the response of measurement points is not effectively captured, resulting in ambiguity in the identification of abnormal temperature sources. At the same time, the judgment of cooling system anomalies is highly dependent on changes in indirect parameters, lacking a multi-dimensional correlation analysis mechanism with heat transfer paths, periodic trends, and humidity changes. The data structure is simple and lacks means to identify the continuity and directionality of cooling state changes, leading to identification biases in leakage behavior, slow response speed, incomplete spatial coverage, and monitoring results that are difficult to meet the requirements of high-precision positioning and trend early warning. Summary of the Invention

[0004] To achieve the above objectives, the present invention adopts the following technical solution: a method for intelligent monitoring and leakage early warning of generator temperature signals, comprising the following steps: S1: Extract the temperature disturbance injection time at the end of the generator stator cooling channel, call the temperature sensing node data arranged along the axial direction, analyze the temperature curve slope change points, record the corresponding time, calculate the time difference with the disturbance time and match the node number to generate a temperature sensing response sequence diagram. S2: Based on the temperature response sequence diagram, extract continuous time difference data, calculate the change in time difference between adjacent nodes, compare it with the node axial structure number, and generate an axial heat conduction starting point region diagram. S3: In the axial heat conduction starting point region map, call the heat flux sensor data of the nodes in the region, extract the periodic heat flux data, extract the heat flux difference between adjacent periods, determine the positive and negative directions and sort them to generate a heat flux conduction path map. S4: Based on the heat flux conduction path diagram, extract the periodic pressure difference and temperature rise data measured by the pressure sensor and temperature sensor, construct a periodic pairing sequence and calculate the difference, filter the data segments with the same direction of pressure difference and temperature rise change, and generate a cooling state abnormal area diagram. S5: Cross-match the numbers of the abnormal cooling state area map and the axial heat conduction starting point area map, call the humidity monitoring data collected by the humidity sensor connected to the numbered node, and generate a monitoring and leakage early warning identification map.

[0005] As a further embodiment of the present invention, the temperature response sequence diagram includes a sequence of temperature response node numbers, a time interval between each node response, and the time corresponding to the slope change point; the axial heat conduction starting point region diagram includes the heat conduction starting number position, the region with the smallest time difference, and the concentrated region of temperature rise abrupt change; the heat flux conduction path diagram includes a heat flux change direction indicator, a node number chain path, and a periodic heat flux difference; the cooling state abnormal region diagram includes a pressure difference and temperature rise synchronous change section, an abnormal continuous number segment, and a cooling path number matching sequence; and the monitoring and leakage early warning identification diagram includes a number overlap matching interval, a humidity change jump number segment, and an early warning risk identification indicator.

[0006] As a further aspect of the present invention, the point where the slope of the temperature curve changes refers to the location where the rate of temperature change of the temperature sensing node changes abruptly over time.

[0007] As a further aspect of the present invention, the humidity change sequence of the overlapping numbered areas refers to the set of humidity sensing data changes extracted in chronological order in the area where the cooling anomaly area and the heat conduction starting point are numbered.

[0008] As a further aspect of the present invention, the specific steps of S1 are as follows: S101: Obtain the temperature disturbance injection time at the end of the generator stator cooling channel, the temperature sensing node data arranged along the axial direction, perform differential calculation on each set of temperature time series, monitor the slope change rate at time points, make judgments based on the change characteristics in adjacent time periods, record the timestamps corresponding to the abrupt change points, and generate a set of temperature sensing node slope change time points. S102: Call the set of time points for slope change of the temperature sensing node, retrieve the temperature disturbance injection time, calculate the time difference between the change time point and the injection time, and index the time difference with the node number to obtain the temperature sensing node response delay number mapping set. S103: Based on the temperature sensing node response delay number mapping set, aggregate the response delay values ​​in order of node number to construct the corresponding time series, convert it into an image generation input format, and establish a temperature sensing response sequence diagram.

[0009] As a further aspect of the present invention, the specific steps of S2 are as follows: S201: Based on the temperature response sequence diagram, extract the continuous data column formed by the response time difference, perform the difference calculation between adjacent time differences in the order of node number, use numerical operation between sequential adjacent items, construct the difference sequence, and then arrange it in a structured manner according to the node index order to generate the sequence of changes in adjacent time differences. S202: Call the adjacent time difference change sequence and retrieve the node layout axial logic structure. Compare the position index in the difference sequence with the node number in the axial structure, record the position number of the corresponding difference, and obtain the response difference node number. S203: Based on the response difference node number, call the node temperature change data, identify the temperature rise point and mark it as the temperature rise number, perform cross-retrieval on the coordinate range of the two numbers, extract the node set corresponding to the intersection interval, construct the spatial identification range of the set in the axial structure, and establish an axial heat conduction starting point region map.

[0010] As a further aspect of the present invention, the specific steps of S3 are as follows: S301: Based on the region located in the axial heat conduction starting point region map, call the multi-cycle heat flux data of the corresponding sensing nodes in the region, extract the heat flux value of each node in two adjacent cycles, subtract the cycle values, construct the heat flux change difference column of the nodes, and generate the heat flux change value sequence between nodes. S302: Call the sequence of heat flux change values ​​between nodes, perform positive and negative direction judgment on each heat flux difference, mark the conduction direction attribute according to the difference sign, and rearrange the numerical sequence with direction attribute according to the node number order to obtain the node heat flux direction identification sequence. S303: Based on the node heat flux direction identification sequence, the heat flux directions are linked in a chain using the number as an index to construct a continuous path structure, and a heat flux conduction path diagram is established using the path structure as an input configuration item.

[0011] As a further aspect of the present invention, the specific steps of S4 are as follows: S401: Based on the node sequence in the heat flux conduction path diagram, extract the corresponding cooling path number, and call the periodic pressure difference data and temperature rise data of the node corresponding to the number to construct a pairing sequence of pressure difference value and temperature rise value in two adjacent cycles. Rearrange according to the number index structure to generate a cooling path periodic pairing value sequence. S402: Call the cooling path periodic pairing value sequence, perform numerical difference calculation on the pressure difference value and temperature rise value under adjacent periods, make consistency judgment based on the sign direction of the pressure difference change and temperature rise change, filter the data index segments with the same direction, and obtain the synchronous change number segment of the cooling path. S403: Based on the synchronous change of the numbering segment of the cooling path, extract the continuity feature of the numbering, perform numbering aggregation processing on the continuous numbering segment, map the aggregation result to the spatial distribution position of the node in the path map, and generate a cooling state abnormal area map.

[0012] As a further aspect of the present invention, the specific steps of S5 are as follows: S501: Based on the node numbers in the abnormal cooling state area map and the axial heat conduction starting point area map, perform a bidirectional cross-matching operation, sequentially determine the equality of number values ​​in the number sequences of the two maps, extract the overlapping number intervals and perform index aggregation processing to obtain the number segments that overlap between the abnormality and the heat source. S502: Call the numbered segment where the anomaly and heat source coincide, extract the humidity monitoring sequence of the corresponding node, construct a trend change curve for the humidity data under each node number according to the periodic sequence, and perform sign judgment on the change direction between adjacent points, record the position number of the change direction jump, and generate a humidity change jump number sequence. S503: Based on the humidity change jump number sequence, combined with the node index range in the overlapping numbering segment of the anomaly and heat source, select the continuous numbering segment that appears in both, perform overlay analysis on the numbering segment and map it to the graphic structure, set the identification style parameters, and generate the monitoring and leakage early warning identification map.

[0013] The generator temperature signal intelligent monitoring and leakage early warning system includes: The temperature sensing response identification module is used to achieve S1: extract the temperature disturbance injection time at the end of the generator stator cooling channel, call the temperature sensing node data arranged along the axial direction, analyze the temperature curve slope change points, record the corresponding time, calculate the time difference with the disturbance time and match the node number, and generate a temperature sensing response sequence diagram. The heat conduction starting point positioning module is used to achieve S2: based on the temperature sensing response sequence diagram, extract continuous time difference data, calculate the change in time difference between adjacent nodes, compare it with the node axial structure number, and generate an axial heat conduction starting point region map. The heat flux path analysis module is used to implement S3: In the axial heat conduction starting point region map, it calls the multi-cycle heat flux data of the corresponding sensing node, extracts the heat flux difference between adjacent cycles, determines the positive and negative directions and sorts them to generate a heat flux conduction path map; The cooling anomaly identification module is used to implement S4: Based on the heat flux conduction path diagram, extract the periodic pressure difference and temperature rise data measured by the pressure sensor and temperature sensor, construct a periodic pairing sequence and calculate the difference, filter the data segments with the same direction of pressure difference and temperature rise change, and generate a cooling state anomaly area diagram. The leakage early warning analysis module is used to implement S5: cross-match the numbering of the abnormal cooling state area map with the axial heat conduction starting point area map, call the humidity monitoring data collected by the humidity sensor connected to the numbered node, and generate a monitoring and leakage early warning identification map.

[0014] Compared with the prior art, the advantages and positive effects of the present invention are as follows: In this invention, a time chain is constructed by extracting the response sequence of temperature sensing nodes to locate the starting region of temperature disturbance. The heat diffusion path is reconstructed by combining the direction of multi-cycle heat flux change. The cooling anomaly section is screened by using the synchronous difference between pressure difference and temperature rise. The leakage risk number is identified by superimposing the abrupt change feature of humidity change. This enables accurate location of the anomaly source, dynamic tracking of the heat conduction path, and multi-parameter linkage identification of the cooling system status. It enhances the temporal correlation of the monitoring chain, the continuity of path identification, the spatial accuracy of anomaly judgment, and the stability and reliability of early warning response. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a schematic diagram of the steps of the present invention; Figure 2 This is a detailed schematic diagram of S1 of the present invention; Figure 3 This is a detailed schematic diagram of S2 of the present invention; Figure 4 This is a detailed schematic diagram of S3 of the present invention; Figure 5 This is a detailed schematic diagram of S4 of the present invention; Figure 6 This is a detailed schematic diagram of S5 of the present invention; Figure 7 This is a system module diagram of the present invention. Detailed Implementation

[0017] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0018] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0019] In the embodiments of this invention, the terms "image" and "picture" may sometimes be used interchangeably. It should be noted that, without emphasizing the difference between them, they convey the same meaning. Similarly, the terms "of," "corresponding (relevant)," and "corresponding" may sometimes be used interchangeably. It should be noted that, without emphasizing the difference between them, they convey the same meaning.

[0020] In this embodiment of the invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0021] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0022] Please see Figure 1 This invention provides a method for intelligent monitoring and leakage early warning of generator temperature signals, comprising the following steps: S1: Obtain the temperature disturbance injection time at the end of the generator stator cooling channel, call the temperature sensing node data arranged along the axial direction, extract the time corresponding to the slope change point in each group of temperature curves in sequence, calculate the difference between the time and the injection time and match the node number, connect them in the order of the number to build a time sequence chain, and generate a temperature sensing response sequence diagram. S2: Based on the temperature response sequence diagram, extract the continuous data column formed by the response time difference, perform the subtraction operation of the change in adjacent time difference according to the node arrangement order, compare the result with the axial logic structure of the node layout, locate the intersection area of ​​the shortest difference number and the temperature rise change number, and generate the axial heat conduction starting point area map. S3: Based on the region located by the axial heat conduction starting point region map, call the multi-cycle heat flux data of the corresponding sensor node, extract the heat flux change of the same numbered node between adjacent cycles, identify the positive and negative directions of the difference and sort it according to the number sequence, construct the numbered chain directional path, and generate the heat flux conduction path map. S4: Based on the node sequence in the heat flux conduction path diagram, extract the pressure difference and temperature rise data corresponding to the cooling path number, construct a periodic pairing sequence and calculate the difference of the pairing value during adjacent periods, screen out the data segments with the same direction of pressure difference and temperature rise change and continuous numbering distribution, and generate a cooling state abnormal area map. S5: Based on the node numbers in the abnormal cooling state area map and the axial heat conduction starting point area map, perform a cross-number matching operation to obtain the overlapping number intervals and extract the humidity change sequence. Overlay the humidity change curves to generate a monitoring and leak early warning identification map by showing number segments with abrupt changes in the direction of change.

[0023] The temperature response sequence diagram includes the temperature response node number sequence, the response time interval of each node, and the time corresponding to the slope change point. The axial heat conduction starting point region diagram includes the heat conduction starting number position, the region with the smallest time difference, and the concentrated area of ​​temperature rise abrupt change. The heat flux conduction path diagram includes the heat flux change direction indicator, the node number chain path, and the heat flux difference during the cycle. The cooling state abnormal region diagram includes the pressure difference and temperature rise synchronous change section, the abnormal continuous number segment, and the cooling path number matching sequence. The monitoring and leakage early warning identification diagram includes the number overlap matching interval, the humidity change jump number segment, and the early warning risk identification indicator.

[0024] Please see Figure 2 The specific steps of S1 are as follows: S101: Obtain the temperature disturbance injection time at the end of the generator stator cooling channel, the temperature sensing node data arranged along the axial direction, perform differential calculation on each set of temperature time series, monitor the slope change rate at time points, make judgments based on the change characteristics in adjacent time periods, record the timestamps corresponding to the abrupt change points, and generate a set of temperature sensing node slope change time points. By controlling and setting specific time points to introduce temperature disturbances into the coolant, such as by briefly turning on the heater, this time point is recorded as the injection reference. Multiple temperature sensing nodes are evenly distributed along the axial direction on the cooling channel wall or inner cavity, ensuring that the node number corresponds one-to-one with its physical location. Continuous temperature data of these temperature sensing nodes are acquired during the stable periods before and after the disturbance injection. A data sampling frequency of once per second is recommended, with a total acquisition time of 10 minutes, thus obtaining a complete time series data set for each node. Subsequently, each set of temperature time series data is subjected to hourly differential processing, that is, the temperature difference between the current time point and its previous time point is calculated sequentially according to the time series, forming... The differential sequence, primarily based on the rate of temperature change, is further differentially processed by calculating the difference between the differential values ​​at each time point to obtain the magnitude of the rate of temperature change. A threshold value for the rate of temperature change is set, which can be determined by combining experimental experience and thermal inertia characteristics to 0.2 degrees Celsius per second. All differential change values ​​are iterated over, and when the rate of change at a certain time point exceeds the threshold value, this moment is recorded as the response change time of the current node. Finally, the process of identifying the response jump points of all nodes is completed, and the identified jump times of each node are organized into a record set according to the node number. Each item in this set represents the jump response time point of a node.

[0025] S102: Call the set of time points for slope change of temperature sensing nodes, retrieve the temperature disturbance injection time, calculate the time difference between the change time point and the injection time, and index the time difference with the node number to obtain the temperature sensing node response delay number mapping set. Obtain the set temperature disturbance injection reference time. Subtract the injection reference time from the mutation response time of each node to obtain the response delay time of each node relative to the disturbance. Then, using the node number as an index, map the delay time corresponding to each node into key-value pairs to form the delay number mapping set for this step. During this process, the response time difference needs to be filtered for validity. Records with negative values ​​or greater than the set maximum delay tolerance need to be removed. This tolerance value can be set to 30 seconds based on the actual thermal response. If the response delay of a node is -3 seconds or 35 seconds, it means that the node was not injected before the disturbance or the response was too slow, and it will not participate in subsequent processing. Organize the mapping records of valid node numbers and their response delays to ensure that the numbers are ordered and the record format is uniform, so that the set data can be identified and processed by the next step. Finally, a complete temperature sensing node response delay number mapping set is formed. This mapping set will serve as the data basis for generating temperature sensing response pattern images.

[0026] S103: Based on the temperature sensing node response delay number mapping set, aggregate the response delay values ​​in order of node number to construct the corresponding time series, and convert it into the image generation input format to establish a temperature sensing response sequence diagram; First, all node numbers are sorted according to their physical deployment order. Then, their corresponding response delay values ​​are extracted sequentially and formed into a continuous delay data sequence. This sequence should correspond one-to-one with the actual arrangement of the nodes, completing a one-dimensional mapping of the node response time in spatial distribution. Next, the response delay sequence is converted into a standard data format for image input. During the conversion process, the original delay values ​​are first normalized to ensure they are all within a standard range, such as setting the maximum value to 1, the minimum value to 0, and the intermediate values ​​to be linearly interpolated proportionally. After processing, each node number corresponds to a normalized response delay value. Subsequently, these numbers and normalized values ​​are paired to form coordinate pairs, creating a standard image input matrix. This matrix represents the node sequence on the horizontal axis and the normalized response time on the vertical axis, used to generate a thermal response behavior image. The image can be used in subsequent temperature distribution analysis or feature extraction steps to ensure that the original response pattern can be accurately reflected in the image, ultimately completing the construction of the temperature response sequence map.

[0027] Please see Figure 3 The specific steps of S2 are as follows: S201: Based on the temperature response sequence diagram, extract the continuous data column composed of response time differences, perform the difference calculation between adjacent time differences in the order of node numbers, use numerical operations between sequentially adjacent items, construct the difference sequence, and then arrange it in a structured manner according to the node index order to generate the sequence of changes in adjacent time differences. First, the standardized response delay time data recorded by each temperature sensing node during the disturbance response process is extracted. This data is arranged in ascending order of node number, forming a complete sequence of response time difference data. For example, the response time differences corresponding to nodes numbered T1 to T8 are 3s, 4s, 6s, 9s, 13s, 14s, 15s, and 16s, respectively. Next, the difference in response time difference is calculated sequentially between adjacent nodes, that is, each term starting from the second term is subtracted from the previous term to obtain a sequence of changes in adjacent response time differences. After performing the above calculation, the difference sequence is 1s, 2s, 3s, 4s, 1s, 1s, 1s. Based on this difference sequence, the original node number index is used to... In the corresponding order, the difference is assigned to the next node in the adjacent node pair, i.e., node T2 corresponds to a difference of 1s, T3 corresponds to 2s, and so on. The obtained difference results are rearranged into a structured sequence according to the node number to ensure that the difference data is consistent with the physical node structure. After the difference sequence is completed, each difference value range needs to be checked to confirm whether there is a large jump. If a certain difference exceeds 5s, it needs to be marked as an abnormal difference for subsequent analysis of the possible location of the node thermal response abnormal area. Each operation in this process needs to be recorded. The difference data needs to form a stable mapping relationship with the node number. The final generated sequence of adjacent time difference changes is a standardized data set with a clear physical correspondence.

[0028] S202: Call the sequence of changes in adjacent time differences, retrieve the axial logic structure of the node layout, compare the position index in the difference sequence with the node number in the axial structure, record the position number of the corresponding difference, and obtain the response difference node number. Simultaneously, the logical sequence structure of the temperature sensing nodes along the axial direction in the actual stator structure is retrieved. This structure records the node numbers and layout coordinates in physical position order. During the comparison operation, the position index corresponding to each item in the difference sequence is extracted first, and then the corresponding node number in the axial structure is read. The two are compared one by one. During the comparison process, it is checked whether the position number of each item in the difference sequence is the same as a certain number in the axial structure. If a match is successful, the node number corresponding to the current difference is recorded as the response difference node number. For example, if the node number corresponding to the 4th item in the difference sequence is T5, and this number also exists in the axial structure, then T5 is recorded as the response difference node number. 5. Add the response difference node number set. The entire process needs to be completed item by item. Record the comparison result after each comparison operation and establish a number mapping table of successful comparisons. The recorded content includes information such as difference index, corresponding node number, and node physical coordinates. If the position number in the difference sequence cannot match any deployment node number during the comparison process, skip the item and do not record it. To avoid misjudgment, a strict restriction on number matching can be set, that is, only completely identical numbers can be considered as a successful match. Through the above operations, the number mapping from the time difference change sequence to the actual deployment node is completed, and finally the response difference node number set is obtained.

[0029] S203: Based on the response difference node number, call the node temperature change data, identify the temperature rise point and mark it as the temperature rise number, perform cross-retrieval on the coordinate range of the two numbers, extract the node set corresponding to the intersection interval, construct the spatial identification range of the set in the axial structure, and establish an axial heat conduction starting point region map. Further extract the set of node numbers for temperature rise abrupt changes recorded in the previous steps. Both sets must be confirmed to have the same numbering format and correspond one-to-one with their positions in the axial node structure. During the cross-referencing operation, first, sort the two sets in ascending order of node numbers. For each number in set A (response difference node numbers), perform a matching search in set B (temperature rise abrupt change numbers). If a matching number is found, add that node number to the intersection interval node set. Continue this process until all matching is complete, obtaining all node numbers that appear in both sets, forming the intersection node set. Then, the physical coordinates of these intersection nodes in the axial layout structure are read, and the spatial coverage of the set is determined by the minimum and maximum coordinate values. For example, the axial coordinates of T4, T5, and T6 are 420mm, 460mm, and 500mm, respectively. The spatial range of the intersection interval is 420mm to 500mm. This coordinate range is defined as the axial heat conduction starting area. At the same time, the boundary node number and coordinate information of the starting area are recorded, and a unified structural identification record is formed for subsequent location comparison with other thermal response characteristic sections. Finally, an axial heat conduction starting area map is established.

[0030] Please see Figure 4 The specific steps of S3 are as follows: S301: Based on the region located by the axial heat conduction starting point region map, call the multi-cycle heat flux data of the corresponding sensing nodes in the region, extract the heat flux value of each node in two adjacent cycles, subtract the cycle values, construct the heat flux change difference column of the nodes, and generate the heat flux change value sequence between nodes. Determine the specific temperature sensing node numbers within the area. For example, if the axial coordinates of the heat conduction initiation zone are 420mm to 500mm, the corresponding node numbers are T4, T5, T6, and T7. Retrieve heat flux data collected from these nodes over multiple operating cycles. The cycle can be set to 30 seconds. Extract the heat flux readings for each node in two consecutive cycles. For example, if the heat flux values ​​for node T4 in cycles n and n+1 are 620W / m² and 580W / m², subtract the values ​​from 620 (580), resulting in a change of -40W / m². This indicates a decrease in heat flux at this node between the two cycles. Then... The same numerical subtraction operation is then performed on nodes T5, T6, T7, etc. Assuming that the heat flux of T5 in periods n and n+1 are 640W / m² and 670W / m² respectively, its heat flux change is +30W / m². The heat flux difference of all nodes is calculated in this way, and an independent heat flux change difference column is constructed for each node, forming an ordered data set composed of node heat flux change values. For the convenience of subsequent processing, these difference data are paired one by one with their corresponding node numbers to form a standard data structure mapping table, such as {T4: -40, T5: +30, T6: -20, T7: +10}. Finally, the data is organized into a sequence of heat flux change values ​​between nodes.

[0031] S302: Call the sequence of heat flux change values ​​between nodes, perform positive and negative direction judgment on each heat flux difference, mark the conduction direction attribute according to the sign of the difference, and rearrange the numerical sequence with direction attribute according to the node number order to obtain the node heat flux direction identification sequence. For each node, the heat flux difference is used to determine its direction. This determination is based on the sign of the value: a positive value is marked as "positive," indicating heat flux is conducted from upstream to downstream; a negative value is marked as "negative," indicating heat is flowing backward or trapped. For example, a heat flux difference of -40 W / m² for T4 is marked as negative, while a difference of +30 W / m² for T5 is marked as positive. After this determination, the direction attribute of each node is combined with its node number to form a sequence pair with direction information. This sequence of heat flux changes with direction attributes is then reordered according to the node number. During the sorting process, the order is ensured to be consistent with the physical layout structure to reflect the actual heat flow conduction path, thereby forming a logically continuous direction identification sequence. For example, the final direction identification sequence is {T4: reverse, T5: forward, T6: reverse, T7: forward}. The judgment of all heat flux change directions must be completed under numerical precision control. For changes close to zero, a judgment tolerance threshold must be set. For example, heat flux changes within ±5W / m² are considered invalid direction changes and are not included in the judgment logic. The direction marks are uniformly encoded according to logical symbols to facilitate subsequent path construction and visualization processing, and finally the generation of the node heat flux direction identification sequence is completed.

[0032] S303: Based on the node heat flux direction identification sequence, the heat flux directions are linked in a chain using the number as the index to construct a continuous path structure, and the heat flux conduction path diagram is established using the path structure as the input configuration item. First, extract the node number and its direction label as input. Then, traverse the nodes sequentially using the node number as the index, chaining the heat flux directions between the current node and its adjacent nodes. This chaining process must follow the actual physical node sequence. For example, if nodes are numbered T4 to T7, and T4 is in reverse direction and T5 is in forward direction, the connection path is represented as T4→T5, and the path direction is marked as "reverse-forward". This process continues. If a node's direction is empty or its change value is in an invalid range, skip that connection segment to maintain path continuity. After completing the directional connections between all valid nodes, the path... The path structure is uniformly represented as a set of directional paths. Each segment in the path set represents a directional heat flux transfer segment, such as {T4→T5: reverse-forward, T5→T6: forward-reverse, T6→T7: reverse-forward}. Finally, this set of directional paths is imported into the heat flux conduction path graph construction module as an input configuration item. In the graph, the path graph is constructed with node numbers as nodes and directional connections as edges. The conduction direction attribute is marked on each edge. At the same time, the original physical location information of the nodes is preserved in the graph, so that the image not only presents the direction of the heat conduction path, but also has spatial correspondence capability, and finally the heat flux conduction path graph is established.

[0033] Please see Figure 5 The specific steps of S4 are as follows: S401: Based on the node sequence in the heat flux conduction path diagram, extract the corresponding cooling path number, and call the periodic pressure difference data and temperature rise data of the node corresponding to the number to construct a pairing sequence of pressure difference value and temperature rise value in two adjacent cycles. Rearrange according to the number index structure to generate a cooling path periodic pairing value sequence. Extract the cooling path number corresponding to each node one by one. For example, the nodes in the diagram are T3 to T8, and the corresponding cooling path numbers are P1, P2, P3, P4, P5, and P6, respectively. Retrieve the operating cycle data for each node corresponding to each number, and extract the differential pressure data and temperature rise data within the cycle from the historical data platform. Set the cycle unit to 30 seconds and select two consecutive cycles as comparison cycles, such as cycle n and cycle n+1. For node T5 corresponding to number P3, the differential pressure value is 0.185 MPa and the temperature rise value is 10.8 degrees Celsius in cycle n, and the differential pressure value is 0.213 MPa and the temperature rise value is 11.4 degrees Celsius in cycle n+1. Record these two sets of data separately. The data is recorded and paired in cyclic order. Similarly, for path numbers P1, P2, P4, P5, and P6, the cyclic pressure difference and temperature rise values ​​of their respective nodes are retrieved to form two-cycle paired data groups corresponding to each path number. For example, P1 is (0.150MPa, 9.6℃) - (0.162MPa, 9.9℃), P2 is (0.173MPa, 10.1℃) - (0.191MPa, 10.7℃), and so on. After all the paired data is completed, the data is rearranged in ascending order of path number so that the data structure of P1 to P6 remains consistent with the path topology. The number and its corresponding cyclic data paired value are combined into a structured sequence, which finally forms the cooling path cyclic paired value sequence.

[0034] S402: Call the cooling path cycle pairing value sequence, perform numerical difference calculation on the pressure difference value and temperature rise value under adjacent cycles, make consistency judgment based on the sign direction of the pressure difference change and temperature rise change, filter the data index segments with the same direction, and obtain the cooling path synchronous change number segment. For each path number, the two periodic data items are processed sequentially. The pressure difference value and temperature rise value of period n+1 are subtracted from the corresponding values ​​of period n, respectively, to obtain the pressure difference change and temperature rise change for each path number. For example, for path number P3, the pressure difference change is 0.213 minus 0.185, and the temperature rise change is 11.4 minus 10.8, resulting in a pressure difference change of +0.028 MPa and a temperature rise change of +0.6℃. Then, the positive and negative directions of the changes are determined. If the signs of the pressure difference change and temperature rise change are the same, it indicates that there is a consistent change in direction for that path. The numbering record is the synchronous change path number. If the pressure difference change is positive and the temperature rise change is negative, or vice versa, it is considered that the direction is inconsistent and is not recorded. This judgment process is repeated for all paths P1 to P6. For cases where the change value is less than a certain error threshold, such as a pressure difference change of less than 0.005MPa or a temperature rise change of less than 0.1℃, they need to be removed before judgment to avoid small changes affecting the synchronous judgment result. Finally, based on the change direction marking results of all paths, the path numbers with consistent directions are selected to form a set, which constitutes the synchronous change numbering segment of the cooling path.

[0035] S403: Based on the synchronous change of the numbering segment of the cooling path, extract the continuity feature of the numbering, perform numbering aggregation processing on the continuous numbering segment, map the aggregation result to the spatial distribution position of the node in the path map, and generate a map of abnormal cooling status areas. After sorting all the numbers in the synchronization set in ascending order of value, each number is read one by one. Adjacent numbers are compared by difference. A difference of 1 indicates a continuous extension; a difference of 1 indicates a break, and a new segment start point is recorded. For example, if the synchronization numbers are P1, P2, P3, P4, P6, P7, and P9, the differences between P1 and P2 (1), P2 and P3 (1), and P3 and P4 (1) form the first continuous segment P1–P4. A difference of 2 between P4 and P6 indicates a break. A difference of 1 between P6 and P7 forms the second continuous segment P6–P7. A difference of 2 between P7 and P9 indicates a break. P9 remains a separate number for the third segment. After completing the continuous segment division, the corresponding node number is retrieved for each continuous segment, starting with the cooling path number. The node number mapping table is used to read the one-to-one correspondence, and then the axial installation position and radial hierarchical position of the nodes are read from the sensor node deployment table. In the example, P1, P2, P3, and P4 correspond to nodes T1, T2, T3, and T4, respectively. The axial position of T1 is set to 360mm, T2 to 390mm, T3 to 420mm, and T4 to 450mm. The radial hierarchical position of the four nodes belongs to the inner ring channel and is 25mm. After reading, the nodes in this segment are grouped into a spatial sequence according to their numbering order. The minimum axial value of 360mm and the maximum axial value of 450mm in this sequence are extracted to form an axial coverage interval. At the same time, the radial position set of the nodes within this interval is recorded at 25mm, completing the spatial boundary data processing for the first segment. The second segment, P6–P7, corresponds to node T6. For T7, read the mapping table to obtain the axial position of T6 (510mm) and T7 (540mm). The radial level is located at the radial position of 35mm in the middle ring channel. Extract the minimum axial value of 510mm and the maximum value of 540mm in the same way to form the second coverage interval. Record the radial position set of 35mm. For the third segment, P9 corresponds to node T9. Read its axial position of 600mm and radial position of 45mm. Since this segment has only one node, no interval stretching is performed; the point coordinates are directly retained. Establish a unified segment-node-coordinate record for the spatial data of all segments. The record entries are arranged according to segment order. The first segment record is {P1–P4, corresponding to T1, T2, T3, T4, corresponding to axial 360–450mm, corresponding to radial 25mm}. The second segment record... The first segment is recorded as {P6–P7, corresponding to T6 and T7, corresponding to axial 510–540mm and radial 35mm}. The third segment is recorded as {P9, corresponding to T9, corresponding to axial 600mm and radial 45mm}. Then, the axial interval and radial position recorded are written into the path diagram coordinate system. The first segment is continuously marked on the path diagram along the axial range of 360mm to 450mm, and T1 to T4 numbered markers are placed at the node coordinates within this range. The second segment is continuously marked along the axial range of 510mm to 540mm, and T6 and T7 numbered markers are placed in the middle of this range. The third segment has a T9 single-point marker placed at axial 600mm and radial 45mm. After the spatial marking on the path diagram is completed, the cooling state abnormal area diagram is output.

[0036] Please see Figure 6 The specific steps of S5 are as follows: S501: Based on the node numbers in the abnormal cooling state area map and the axial heat conduction starting point area map, perform a bidirectional cross-matching operation, sequentially judge the numbering sequence in the two maps to determine if the numbering values ​​are equal, extract the overlapping numbering intervals and perform index aggregation processing to obtain the numbering segments that overlap between the abnormality and the heat source. First, extract all node ID sets from the two graphical structures. Let N_A be the ID sequence extracted from the cooling anomaly region map (e.g., T4, T5, T6, T7, T8), and N_B be the ID sequence extracted from the axial heat conduction starting point region map (e.g., T3, T4, T5, T6). Then, perform a bidirectional cross-matching operation: using each ID in N_A as a reference, compare it item by item with all IDs in N_B to determine if there are any items with completely identical ID values. Node IDs T4, T5, and T6 that match successfully in both sets are marked. The matching process requires strict consistency of ID values; they cannot be identical. Fuzzy matching or similarity comparison is used to record all nodes with the same number value, forming a number overlap interval. Then, index aggregation processing is performed on the number overlap interval. First, the position index corresponding to these numbers in the original image is identified. For example, T4 has an index of 1 in the anomaly area image and an index of 2 in the starting point area image. T5 and T6 also obtain their respective index values. Then, these nodes are combined into a continuous number segment according to the index order. If T4 to T6 are identified as being within the continuous index interval, they are aggregated into the number segment T4–T6. Finally, this number segment is defined as the number segment that overlaps with the anomaly and heat source, which serves as the location basis for subsequent humidity monitoring overlay analysis.

[0037] S502: Call the number segment where the abnormality and heat source coincide, extract the humidity monitoring sequence of the corresponding node, construct a trend change curve for the humidity data under each node number according to the periodic sequence, and perform sign judgment of the change direction between adjacent points, record the position number of the change direction jump, and generate a humidity change jump number sequence. Retrieve all node numbers within the overlapping numbering segment of the anomaly and heat source, such as T4, T5, and T6. For each of these three node numbers, retrieve the corresponding humidity monitoring time series data, with a period unit of 30 seconds. Select at least five consecutive periods of humidity data for trend change judgment. For example, the humidity data for node T4 in periods 1 to 5 are 46%, 47%, 49%, 48%, and 46%, respectively. Arrange these in chronological order to form a humidity change trend curve. Determine the direction of the difference between the humidity values ​​of any two adjacent periods on this curve. If the humidity value of the later period is greater than that of the previous period, it is considered an increase; if it is less, it is considered a decrease, marked as "+" and "-" respectively. Taking 4 as an example, the period from 1 to 2 is +, 2 to 3 is +, 3 to 4 is -, and 4 to 5 is -. From the 3rd to the 4th period, there is a sign change from "+" to "-", so the number of this change point is recorded as T4. The same operation is repeated to judge T5 and T6. If the trend of T5 is +, +, +, -, -, then the change occurs in the 4th period and is recorded as T5 change. If T6 is a stable increase or decrease throughout, then this node is not recorded. Finally, the numbers of all nodes with directional change behavior are sorted to generate a humidity change change number sequence. If the result is T4 and T5, it means that these nodes have unstable humidity change direction during the monitoring process, which has analytical value.

[0038] S503: Based on the humidity change jump number sequence, combined with the node index range in the overlapping numbered segments of the anomaly and heat source, the continuous numbered segments that appear in both are selected, the numbered segments are overlaid and analyzed and mapped to the graphic structure, and the identification style parameters are set to generate the monitoring and leakage early warning identification map. Based on the node index range in the humidity change jump sequence and the overlapping numbering segment of anomalies and heat sources, the continuity of the node numbers contained in both is screened. The judgment method is as follows: after arranging the numbers in ascending order, the difference between adjacent numbers is compared. If the number interval is 1, it is considered a continuous numbering segment. For example, if the jump sequence is T4, T5, and the overlapping numbering segment is T4, T5, T6, the intersection of the two is T4, T5, forming a continuous numbering segment. This numbering segment is used as the target segment for overlay analysis. Next, this numbering segment is mapped to the monitoring graphic structure. First, T is extracted from the structure diagram. The spatial coordinates of nodes 4 and T5 are used. For example, T4 is located at 420mm in the axial direction and T5 is located at 460mm in the axial direction. This continuous segment is represented as the area from 420mm to 460mm in the axial direction. This spatial area is marked with a red solid line box on the graphic. Then, the corresponding graphic label style parameters are set, such as filling the node symbol with orange, thickening the boundary box line, and setting the transparency of the border to 50%. The area is also marked as "Monitoring and Leakage Early Warning Area" in the legend. Through the above labeling method, a comprehensive visualization of three types of information—abnormality, heat source, and humidity change—is completed, and a monitoring and leakage early warning identification map is finally constructed.

[0039] Please see Figure 7 The generator temperature signal intelligent monitoring and leakage early warning system includes: The temperature sensing response identification module is used to achieve S1: extract the temperature disturbance injection time at the end of the generator stator cooling channel, call the temperature sensing node data arranged along the axial direction, analyze the temperature curve slope change points, record the corresponding time, calculate the time difference with the disturbance time and match the node number, and generate a temperature sensing response sequence diagram. The heat conduction starting point positioning module is used to achieve S2: based on the temperature response sequence diagram, extract continuous time difference data, calculate the change in time difference between adjacent nodes, compare it with the node axial structure number, and generate an axial heat conduction starting point region map. The heat flux path analysis module is used to implement S3: In the axial heat conduction starting point region map, it calls the multi-cycle heat flux data of the corresponding sensor node, extracts the heat flux difference between adjacent cycles, determines the positive and negative directions and sorts them to generate a heat flux conduction path map; The cooling anomaly identification module is used to implement S4: Based on the heat flux conduction path diagram, extract the periodic pressure difference and temperature rise data measured by the pressure sensor and temperature sensor, construct a periodic pairing sequence and calculate the difference, filter the data segments with the same direction of pressure difference and temperature rise change, and generate a cooling state anomaly area map. The leakage early warning analysis module is used to implement S5: cross-matching the numbers of the abnormal cooling state area map with the axial heat conduction starting point area map, calling the humidity monitoring data collected by the humidity sensor connected to the numbered node, and generating a monitoring and leakage early warning identification map.

[0040] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for intelligent monitoring and leakage early warning of generator temperature signals, characterized in that, Includes the following steps: S1: Extract the temperature disturbance injection time at the end of the generator stator cooling channel, call the temperature sensing node data arranged along the axial direction, analyze the temperature curve slope change points, record the corresponding time, calculate the time difference with the disturbance time and match the node number to generate a temperature sensing response sequence diagram. S2: Based on the temperature response sequence diagram, extract continuous time difference data, calculate the change in time difference between adjacent nodes, compare it with the node axial structure number, and generate an axial heat conduction starting point region diagram. S3: In the axial heat conduction starting point region map, call the heat flux sensor data of the nodes in the region, extract the periodic heat flux data, extract the heat flux difference between adjacent periods, determine the positive and negative directions and sort them to generate a heat flux conduction path map. S4: Based on the heat flux conduction path diagram, extract the periodic pressure difference and temperature rise data measured by the pressure sensor and temperature sensor, construct a periodic pairing sequence and calculate the difference, filter the data segments with the same direction of pressure difference and temperature rise change, and generate a cooling state abnormal area diagram. S5: Cross-match the numbers of the abnormal cooling state area map and the axial heat conduction starting point area map, call the humidity monitoring data collected by the humidity sensor connected to the numbered node, and generate a monitoring and leakage early warning identification map.

2. The intelligent monitoring and leakage early warning method for generator temperature signals according to claim 1, characterized in that, The temperature response sequence diagram includes a sequence of temperature response node numbers, a time interval between each node response, and the time corresponding to the slope change point. The axial heat conduction starting point region diagram includes the heat conduction starting number position, the region with the smallest time difference, and the concentrated area of ​​temperature rise abrupt change. The heat flux conduction path diagram includes heat flux change direction indicators, a chain path of node numbers, and heat flux difference during the cycle. The cooling state abnormal region diagram includes a section where pressure difference and temperature rise change synchronously, a continuous abnormal numbering segment, and a cooling path number matching sequence. The monitoring and leakage early warning identification diagram includes a number overlap matching interval, a humidity change jump numbering segment, and an early warning risk identification indicator.

3. The intelligent monitoring and leakage early warning method for generator temperature signals according to claim 1, characterized in that, The point where the slope of the temperature curve changes refers to the location where the rate of temperature change over time at the temperature sensing node changes abruptly.

4. The intelligent monitoring and leakage early warning method for generator temperature signals according to claim 1, characterized in that, The humidity change sequence of the overlapping numbered areas refers to the set of humidity sensor data changes extracted in chronological order in the area where the cooling anomaly area and the heat conduction starting point are numbered.

5. The intelligent monitoring and leakage early warning method for generator temperature signals according to claim 1, characterized in that, The specific steps of S1 are as follows: S101: Obtain the temperature disturbance injection time at the end of the generator stator cooling channel, the temperature sensing node data arranged along the axial direction, perform differential calculation on each set of temperature time series, monitor the slope change rate at time points, make judgments based on the change characteristics in adjacent time periods, record the timestamps corresponding to the abrupt change points, and generate a set of temperature sensing node slope change time points. S102: Call the set of time points for slope change of the temperature sensing node, retrieve the temperature disturbance injection time, calculate the time difference between the change time point and the injection time, and index the time difference with the node number to obtain the temperature sensing node response delay number mapping set. S103: Based on the temperature sensing node response delay number mapping set, aggregate the response delay values ​​in order of node number to construct the corresponding time series, convert it into an image generation input format, and establish a temperature sensing response sequence diagram.

6. The intelligent monitoring and leakage early warning method for generator temperature signals according to claim 1, characterized in that, The specific steps of S2 are as follows: S201: Based on the temperature response sequence diagram, extract the continuous data column composed of response time differences, perform the difference calculation between adjacent time differences in the order of node numbers, use numerical operations between sequentially adjacent items, construct the difference sequence, and then arrange it in a structured manner according to the node index order to generate the sequence of changes in adjacent time differences. S202: Call the adjacent time difference change sequence and retrieve the node layout axial logic structure. Compare the position index in the difference sequence with the node number in the axial structure, record the position number of the corresponding difference, and obtain the response difference node number. S203: Based on the response difference node number, call the node temperature change data, identify the temperature rise point and mark it as the temperature rise number, perform cross-retrieval on the coordinate range of the two numbers, extract the node set corresponding to the intersection interval, construct the spatial identification range of the set in the axial structure, and establish an axial heat conduction starting point region map.

7. The intelligent monitoring and leakage early warning method for generator temperature signals according to claim 1, characterized in that, The specific steps for S3 are as follows: S301: Based on the region located in the axial heat conduction starting point region map, call the multi-cycle heat flux data of the corresponding sensing nodes in the region, extract the heat flux value of each node in two adjacent cycles, subtract the cycle values, construct the heat flux change difference column of the nodes, and generate the heat flux change value sequence between nodes. S302: Call the sequence of heat flux change values ​​between nodes, perform positive and negative direction judgment on each heat flux difference, mark the conduction direction attribute according to the difference sign, and rearrange the numerical sequence with direction attribute according to the node number order to obtain the node heat flux direction identification sequence. S303: Based on the node heat flux direction identification sequence, using the number as an index, the heat flux directions are linked in a chain to construct a continuous path structure, and the heat flux conduction path diagram is established using the path structure as an input configuration item.

8. The intelligent monitoring and leakage early warning method for generator temperature signals according to claim 1, characterized in that, The specific steps of S4 are as follows: S401: Based on the node sequence in the heat flux conduction path diagram, extract the corresponding cooling path number, and call the periodic pressure difference data and temperature rise data of the node corresponding to the number to construct a pairing sequence of pressure difference value and temperature rise value in two adjacent cycles. Rearrange according to the number index structure to generate a cooling path periodic pairing value sequence. S402: Call the cooling path periodic pairing value sequence, perform numerical difference calculation on the pressure difference value and temperature rise value under adjacent periods, make consistency judgment based on the sign direction of the pressure difference change and temperature rise change, filter the data index segments with the same direction, and obtain the synchronous change number segment of the cooling path. S403: Based on the synchronous change of the numbering segment of the cooling path, extract the continuity feature of the numbering, perform numbering aggregation processing on the continuous numbering segment, map the aggregation result to the spatial distribution position of the node in the path map, and generate a cooling state abnormal area map.

9. The intelligent monitoring and leakage early warning method for generator temperature signals according to claim 1, characterized in that, The specific steps of S5 are as follows: S501: Based on the node numbers in the abnormal cooling state area map and the axial heat conduction starting point area map, perform a bidirectional cross-matching operation, sequentially determine the equality of number values ​​in the number sequences of the two maps, extract the overlapping number intervals and perform index aggregation processing to obtain the number segments that overlap between the abnormality and the heat source. S502: Call the numbered segment where the anomaly and heat source coincide, extract the humidity monitoring sequence of the corresponding node, construct a trend change curve for the humidity data under each node number according to the periodic sequence, and perform sign judgment on the change direction between adjacent points, record the position number of the change direction jump, and generate a humidity change jump number sequence. S503: Based on the humidity change jump number sequence, combined with the node index range in the overlapping numbering segment of the anomaly and heat source, select the continuous numbering segment that appears in both, perform overlay analysis on the numbering segment and map it to the graphic structure, set the identification style parameters, and generate the monitoring and leakage early warning identification map.

10. A generator temperature signal intelligent monitoring and leakage early warning system, characterized in that, The system is used to implement the intelligent monitoring and leakage early warning method for generator temperature signals according to any one of claims 1-9, the system comprising: The temperature sensing response identification module is used to achieve S1: extract the temperature disturbance injection time at the end of the generator stator cooling channel, call the temperature sensing node data arranged along the axial direction, analyze the temperature curve slope change points, record the corresponding time, calculate the time difference with the disturbance time and match the node number, and generate a temperature sensing response sequence diagram. The heat conduction starting point positioning module is used to achieve S2: based on the temperature sensing response sequence diagram, extract continuous time difference data, calculate the change in time difference between adjacent nodes, compare it with the node axial structure number, and generate an axial heat conduction starting point region map. The heat flux path analysis module is used to implement S3: In the axial heat conduction starting point region map, it calls the multi-cycle heat flux data of the corresponding sensing node, extracts the heat flux difference between adjacent cycles, determines the positive and negative directions and sorts them to generate a heat flux conduction path map; The cooling anomaly identification module is used to implement S4: Based on the heat flux conduction path diagram, extract the periodic pressure difference and temperature rise data measured by the pressure sensor and temperature sensor, construct a periodic pairing sequence and calculate the difference, filter the data segments with the same direction of pressure difference and temperature rise change, and generate a cooling state anomaly area map. The leakage early warning analysis module is used to implement S5: cross-match the numbering of the abnormal cooling state area map with the axial heat conduction starting point area map, call the humidity monitoring data collected by the humidity sensor connected to the numbered node, and generate a monitoring and leakage early warning identification map.

Images