Brake operating condition monitoring method and system

Abstract

The application provides a brake operation state monitoring method and system, and relates to the technical field of brake monitoring. Firstly, acoustic emission waveform time sequence and infrared thermal radiation field image time sequence in a brake execution process are acquired. Then, a sound-heat energy transfer correlation path diagram is constructed. Next, sound-heat energy transfer time sequence offset features are identified. Then, the sound-heat energy transfer time sequence offset features are used to generate a hysteresis coupling path diagram. Finally, the hysteresis coupling path diagram is used to determine the contact interface wear state grade of the brake friction pair. The application can accurately depict the energy transfer path and characteristics by collecting physical information in the braking process and revealing the internal relationship of the signals, and can reliably determine the wear state of the brake, thereby improving the operation safety and reliability of the brake.

Description

Technical Field

[0001] This invention relates to the field of brake monitoring technology, and more specifically, to a method and system for monitoring the operating status of a brake. Background Technology

[0002] During the operation of the brake, its performance stability and reliability are directly related to the safe operation of the entire device. When the brake is applied, complex physical changes occur at the contact interface of the friction pair, which trigger the generation of various physical signals, such as acoustic emission signals and infrared thermal radiation signals.

[0003] Currently, methods for monitoring the operating status of brakes mainly suffer from the following problems. Firstly, existing monitoring methods often analyze only a single physical signal, such as acoustic emission signals or infrared thermal radiation signals. However, there are complex correlations between various physical signals during braking, making it difficult to comprehensively and accurately reflect the actual operating status of the brake through single-signal analysis. Secondly, while some methods consider multiple signals, they fail to delve into the intrinsic relationships between these signals during analysis. This results in an inability to effectively capture the energy transfer characteristics of the brake friction pair contact interface and the changing patterns of wear conditions, leading to low accuracy and reliability of the monitoring results, which is insufficient to meet the needs of precise brake operating status monitoring in practical engineering applications. Summary of the Invention

[0004] In view of the aforementioned problems, and in conjunction with the first aspect of the present invention, an embodiment of the present invention provides a method for monitoring the operating status of a brake, the method comprising:

[0005] Acquire the timing sequence of acoustic emission waveforms generated by the contact interface of the friction pair and the timing sequence of infrared thermal radiation field images generated by the friction surface of the brake disc during the braking process; Based on the waveform morphology characteristics of the acoustic emission waveform in the time sequence and the thermal radiation field distribution morphology characteristics in the time sequence of the infrared thermal radiation field image, an acoustic-thermal energy transfer correlation path diagram is constructed. Identify the temporal offset relationship between the inflection point of acoustic emission waveform morphology change and the inflection point of thermal radiation field distribution morphology change in the acoustic-thermal energy transfer correlation path diagram to obtain the acoustic-thermal energy transfer temporal offset characteristics. The acoustic and thermal energy transfer time sequence offset features are used to perform path shape correction processing on the acoustic and thermal energy transfer association path diagram to generate a hysteresis coupling path diagram that reflects the energy transfer hysteresis characteristics of the friction pair contact interface. The wear status level of the contact interface of the brake friction pair is determined based on the morphological difference between the hysteresis segment of the acoustic emission waveform and the leading segment of the thermal radiation field distribution in the hysteresis coupling path diagram.

[0006] Furthermore, embodiments of the present invention also provide a brake operating status monitoring system, comprising: A processor; a machine-readable storage medium for storing machine-executable instructions of the processor; wherein the processor is configured to perform the above-described brake operating state monitoring method by executing the machine-executable instructions.

[0007] In another aspect, embodiments of the present invention also provide a computer program product, the computer program product including machine-executable instructions, the machine-executable instructions being stored in a computer-readable storage medium, the processor of the brake operating status monitoring system reading the machine-executable instructions from the computer-readable storage medium, the processor executing the machine-executable instructions, causing the brake operating status monitoring system to perform the above-described brake operating status monitoring method.

[0008] Based on the above, by acquiring the time sequence of acoustic emission waveforms generated at the contact interface of the friction pair during brake operation and the time sequence of infrared thermal radiation field images generated by the friction surface of the brake disc, an acoustic-thermal energy transfer correlation path diagram is constructed based on the waveform morphology characteristics of the acoustic emission waveform and the distribution morphology characteristics of the thermal radiation field. This diagram depicts the energy transfer path during braking. The temporal offset relationship between the inflection point of the acoustic emission waveform morphology change and the inflection point of the thermal radiation field distribution morphology change in the acoustic-thermal energy transfer correlation path diagram is identified, yielding the acoustic-thermal energy transfer temporal offset characteristics. This further clarifies the time characteristics during energy transfer. Using this temporal offset characteristic, the correlation path diagram is processed to correct the path morphology, generating a hysteresis coupling path diagram that reflects the hysteresis characteristics of energy transfer at the contact interface of the friction pair. This more realistically reflects the energy transfer situation during braking. Finally, based on the morphological difference between the hysteresis segment of the acoustic emission waveform morphology and the leading segment of the thermal radiation field distribution morphology in the hysteresis coupling path diagram, the wear state level of the contact interface of the brake friction pair is determined. This achieves accurate judgment of the brake wear state and effectively improves the safety and reliability of brake operation. Attached Figure Description

[0009] Figure 1 This is a schematic diagram of the execution flow of the brake operation status monitoring method provided in the embodiment of the present invention.

[0010] Figure 2 This is a schematic diagram of exemplary hardware and software components of the brake operation status monitoring system provided in an embodiment of the present invention. Detailed Implementation

[0011] Figure 1 This is a flowchart illustrating a brake operation status monitoring method according to an embodiment of the present invention, which will be described in detail below.

[0012] Step S110: Obtain the timing sequence of the acoustic emission waveform generated by the contact interface of the friction pair and the timing sequence of the infrared thermal radiation field image generated by the friction surface of the brake disc during the braking process.

[0013] Step S111: The acoustic emission sensing unit deployed on the force transmission path between the brake caliper and the friction lining starts acquiring acoustic emission waveforms. The waveform start point and waveform end point of each acoustic emission event are identified from the original voltage waveform output by the acoustic emission sensing unit. The waveform segment between the waveform start point and waveform end point of each acoustic emission event is taken as an acoustic emission waveform unit. All acoustic emission waveform units are arranged in the time sequence of the waveform start point to form an initial sequence of acoustic emission waveform units.

[0014] For example, the acoustic emission sensing unit is installed at the rigid connection between the brake caliper and the friction lining contact backplate. When braking pressure is applied, the fracture or plastic deformation of the micro-protrusions at the friction pair contact interface releases transient elastic waves, which are transmitted to the acoustic emission sensing unit through the structure and converted into a raw voltage waveform. The raw voltage waveform is continuously monitored, and a dual threshold method using short-time energy and zero-crossing rate is used to identify the start and end boundaries of each acoustic emission event. Specifically, the cumulative energy value of the raw voltage waveform within a sliding time window is calculated. When the cumulative energy value exceeds a preset high energy threshold, it is determined as the waveform start point; from the waveform start point onwards, when the cumulative energy value is lower than a preset low energy threshold and remains below a preset stable duration, it is determined as the waveform end point. Each continuous waveform data segment from the waveform start point to the waveform end point is extracted and marked as an acoustic emission waveform unit. All acoustic emission waveform units are sequentially stored in a sequence data structure according to the chronological order of their waveform start points, forming an initial sequence of acoustic emission waveform units.

[0015] Step S112: At the start of the infrared thermal radiation imaging unit deployed in the normal direction of the brake disc friction working surface, the temperature field distribution array of the brake disc friction working surface in each frame of the original thermal image sequence output by the infrared thermal radiation imaging unit is extracted, and the temperature values ​​of each element point in the temperature field distribution array are arranged in the time order of the image frames to form the initial sequence of the infrared thermal radiation field image unit.

[0016] For example, the infrared thermal radiation imaging unit is deployed in the normal direction of the brake disc friction working surface, so that the optical field of view of the infrared thermal radiation imaging unit completely covers the entire brake disc friction area. The infrared thermal radiation imaging unit outputs a sequence of raw thermal images at a fixed frame rate. For each frame in the sequence, the region of interest containing only the brake disc friction working surface is first extracted using an image segmentation algorithm. Within the region of interest, each pixel corresponds to a temperature reading at a spatial location, and these temperature readings are arranged according to the row and column coordinates of the pixels, forming a two-dimensional temperature field distribution array. The two-dimensional temperature field distribution array of each frame is defined as an infrared thermal radiation field image unit. All infrared thermal radiation field image units are arranged in order of their acquisition time to form the initial sequence of infrared thermal radiation field image units.

[0017] Step S113: Extract the moment when the waveform amplitude reaches its maximum value from each acoustic emission waveform unit in the initial sequence of the acoustic emission waveform units, use this moment as the waveform feature anchor point of the acoustic emission waveform unit, and arrange the waveform feature anchor points of all acoustic emission waveform units in chronological order to form an acoustic emission waveform feature anchor point sequence.

[0018] For each acoustic emission waveform unit in the initial sequence, the discrete voltage amplitude data points of the acoustic emission waveform unit are traversed to find the sampling point with the largest absolute amplitude, and the absolute timestamp corresponding to the sampling point is recorded. This timestamp is the waveform feature anchor point of that acoustic emission waveform unit, representing the moment when the energy release of the acoustic emission event is most intense. The waveform feature anchor points of all acoustic emission waveform units are extracted and arranged sequentially according to the order of their corresponding acoustic emission waveform units in the initial sequence, forming an acoustic emission waveform feature anchor point sequence.

[0019] Step S114: Extract the position of the element point in the temperature field distribution array where the temperature value reaches the maximum value from each frame image of the initial sequence of the infrared thermal radiation field image unit, and use the position of the element point as the thermal radiation field feature anchor point of the frame image. Arrange the thermal radiation field feature anchor points of all frames images in chronological order to form a thermal radiation field feature anchor point sequence.

[0020] For each frame of an infrared thermal radiation field image unit in the initial sequence, a global search is performed on the temperature values ​​of all array elements in the temperature field distribution array of the infrared thermal radiation field image unit to find the array element with the highest temperature value. The coordinate position of the above array element in the image coordinate system is recorded. The above coordinate position is the thermal radiation field feature anchor point of the infrared thermal radiation field image unit of that frame, representing the location where the heat source is most concentrated on the friction surface at that moment. The thermal radiation field feature anchor points of all frames of infrared thermal radiation field image units are extracted and arranged sequentially according to the order of their corresponding infrared thermal radiation field image units in the initial sequence of infrared thermal radiation field image units to form a thermal radiation field feature anchor point sequence.

[0021] Step S115: Pair and associate the acoustic emission waveform feature anchor point sequence with the anchor points in the thermal radiation field feature anchor point sequence that have the same acquisition time sequence number. Form a spatiotemporal association unit pair between the paired and associated acoustic emission waveform unit and the infrared thermal radiation field image unit. Arrange all spatiotemporal association unit pairs in chronological order to form a set of synchronous association unit pairs between the acoustic emission waveform time sequence and the infrared thermal radiation field image time sequence.

[0022] The acoustic emission waveform feature anchor sequence and the thermal radiation field feature anchor sequence are asynchronous on the time axis, but they can be aligned using their respective acquisition timing numbers. Specifically, anchor points with the same sequence number in both sequences are considered to correspond to the same stage of the same braking process in physical time. The acoustic emission waveform unit that generates the acoustic emission waveform feature anchor point is associated with the infrared thermal radiation field image unit that generates the thermal radiation field feature anchor point, forming a spatiotemporal associated unit pair. All spatiotemporal associated unit pairs formed in this way are arranged according to the anchor point sequence number of the spatiotemporal associated unit pairs, thus forming a set of synchronous associated unit pairs.

[0023] Step S116: Extract the waveform duration segment and waveform rise time segment of the acoustic emission waveform unit from each spatiotemporal association unit pair in the set of synchronous association unit pairs. Use the length relationship between the waveform duration segment and the waveform rise time segment as the acoustic emission waveform time segment feature of the spatiotemporal association unit pair. Arrange the acoustic emission waveform time segment features of all spatiotemporal association unit pairs in chronological order to form an acoustic emission waveform time segment feature sequence.

[0024] For each spatiotemporal correlated unit pair in the set of synchronized correlated unit pairs, two time metrics are calculated from the corresponding acoustic emission waveform unit. The waveform duration segment is defined as the length of time from the waveform start time stamp to the waveform end time stamp of the acoustic emission waveform unit. The waveform rise time segment is defined as the length of time from the waveform start time stamp to the moment when the waveform amplitude reaches its maximum value. The ratio of the waveform rise time segment to the waveform duration segment constitutes a time ratio feature characterizing the waveform morphology, which is the acoustic emission waveform time segment feature of the spatiotemporal correlated unit pair. The acoustic emission waveform time segment features of all spatiotemporal correlated unit pairs are arranged in the order of their corresponding spatiotemporal correlated unit pairs to form an acoustic emission waveform time segment feature sequence.

[0025] Step S117: Extract the relative positional relationship between the highest temperature array element and the geometric center point of the temperature field distribution array in each spatiotemporal association unit pair of the synchronous association unit pair set. Use this relative positional relationship as the thermal radiation field offset feature of the spatiotemporal association unit pair. Arrange the thermal radiation field offset features of all spatiotemporal association unit pairs in chronological order to form a thermal radiation field offset feature sequence.

[0026] For each spatiotemporal correlated unit pair in the set of synchronous correlated unit pairs, the coordinates of the element point with the highest temperature value are located from the temperature field distribution array of the infrared thermal radiation field image unit corresponding to the spatiotemporal correlated unit pair. The Euclidean distance and direction vector between this coordinate position and the geometric center point of the temperature field distribution array are then calculated. These Euclidean distance and direction vector together constitute a two-dimensional feature describing the degree and direction of heat source offset; this two-dimensional feature is the thermal radiation field offset feature of the spatiotemporal correlated unit pair. The thermal radiation field offset features of all spatiotemporal correlated unit pairs are arranged in the order of their corresponding spatiotemporal correlated unit pairs, forming a thermal radiation field offset feature sequence.

[0027] Step S118: Standardize the acoustic emission waveform time period feature sequence and the thermal radiation field shift feature sequence respectively. Construct a spatiotemporal correlation feature pair based on each feature value in the standardized acoustic emission waveform time period feature sequence and the feature value of the corresponding time position in the standardized thermal radiation field shift feature sequence. Arrange all spatiotemporal correlation feature pairs in chronological order to form a spatiotemporal correlation feature pair sequence.

[0028] All eigenvalues ​​in the acoustic emission waveform time-segment feature sequence are Z-score standardized to achieve zero mean and unit variance. Similarly, each component (range and direction components) in the thermal radiation field migration feature sequence is Z-score standardized to obtain a standardized thermal radiation field migration feature sequence. The nth eigenvalue in the standardized acoustic emission waveform time-segment feature sequence is combined with the nth eigenvalue in the standardized thermal radiation field migration feature sequence to form a spatiotemporally correlated feature pair containing both the acoustic emission waveform time-segment feature component and the thermal radiation field migration feature component. All spatiotemporally correlated feature pairs formed in this way are arranged in order of their index n to form a spatiotemporally correlated feature pair sequence.

[0029] Step S119: Identify the degree of correlation between the acoustic emission waveform time period feature component and the thermal radiation field offset feature component of each spatiotemporal correlation feature pair in the spatiotemporal correlation feature pair sequence, obtain the correlation tightness level of each spatiotemporal correlation feature pair, arrange the correlation tightness levels of all spatiotemporal correlation feature pairs in chronological order to form a correlation tightness level sequence, and use the correlation tightness level sequence as a sequence representing the degree of synchronization correlation between the acoustic emission waveform time sequence and the infrared thermal radiation field image time sequence.

[0030] For each spatiotemporal correlated feature pair in the spatiotemporal correlated feature pair sequence, the comprehensive correlation degree between its acoustic emission waveform time-period feature component and the distance and direction components of the thermal radiation field offset feature component is calculated. Specifically, firstly, the absolute value of the first difference between the acoustic emission waveform time-period feature component and the distance component of the thermal radiation field offset feature component is calculated, and simultaneously, the absolute value of the second difference between the acoustic emission waveform time-period feature component and the direction component of the thermal radiation field offset feature component is calculated. The first and second absolute values ​​of the difference are compared with multiple preset level threshold intervals. When both the first and second absolute values ​​of the difference fall within the first level threshold interval, the correlation strength level of the spatiotemporal correlated feature pair is determined to be level one; when either the first or second absolute value of the difference exceeds the first level threshold interval but remains within the second level threshold interval, it is determined to be level two; and so on, forming multiple correlation strength levels. The correlation strength levels of all spatiotemporal correlated feature pairs are arranged in the order of their corresponding spatiotemporal correlated feature pairs to form a correlation strength level sequence. The above-mentioned correlation level sequence is a characterization sequence of the synchronization correlation between the acoustic emission waveform time series sequence and the infrared thermal radiation field image time series sequence.

[0031] Step S120: Construct an acoustic-thermal energy transfer correlation path diagram based on the waveform morphology characteristics of the acoustic emission waveform in the acoustic emission waveform time sequence and the thermal radiation field distribution morphology characteristics in the infrared thermal radiation field image time sequence.

[0032] In this step, based on the synchronized acoustic emission waveform time sequence and infrared thermal radiation field image time sequence, their respective morphological features are extracted. Through the construction and superposition of feature evolution paths, a graph structure that can intuitively reflect the transmission and evolution relationship of the two energy forms during braking is formed.

[0033] Step S121: Extract the rising waveform curve from the starting point to the peak point and the falling waveform curve from the peak point to the end point from each acoustic emission waveform time sequence unit. Use the rising waveform curve and the falling waveform curve together as the waveform morphology feature of the acoustic emission waveform unit. Arrange the waveform morphology features of all acoustic emission waveform units in chronological order to form a waveform morphology feature time sequence.

[0034] For each acoustic emission waveform unit in the acoustic emission waveform time sequence, connecting all discrete amplitude points from the waveform start point to the waveform feature anchor point forms a curve, which is the rising curve. Connecting all discrete amplitude points from the waveform feature anchor point to the waveform end point forms another curve, which is the falling curve. These rising and falling curves together constitute the waveform morphology features describing the overall contour of the acoustic emission waveform unit. The waveform morphology features of all acoustic emission waveform units are arranged in order of their corresponding acoustic emission waveform units in the acoustic emission waveform time sequence, forming a waveform morphology feature time series.

[0035] Step S122: Extract the radial distribution curve of temperature values ​​decreasing from the center region to the edge region of the temperature field distribution array from each frame of the infrared thermal radiation field image time sequence. Use the radial distribution curve as the thermal radiation field distribution feature of the frame image. Arrange the thermal radiation field distribution features of all frames in chronological order to form a thermal radiation field distribution feature time sequence.

[0036] For each frame of an infrared thermal radiation field image unit in the time sequence of infrared thermal radiation field images, the temperature field distribution array of the infrared thermal radiation field image unit is analyzed. Starting from the geometric center point of the temperature field distribution array, along multiple preset radial directions, temperature values ​​at each sampling point from the center point to the edge region are extracted. The sequence of temperature values ​​extracted along one radial direction constitutes a temperature decay curve. The temperature decay curves extracted in all radial directions are combined to form a radial distribution pattern curve describing the overall trend of temperature decaying radially from the center to the surrounding areas. The above radial distribution pattern curve is the thermal radiation field distribution pattern feature of that frame of infrared thermal radiation field image unit. The thermal radiation field distribution pattern features of all infrared thermal radiation field image units are arranged according to the order of their corresponding infrared thermal radiation field image units in the infrared thermal radiation field image time sequence to form a time series of thermal radiation field distribution pattern features.

[0037] Step S123: Correspond the waveform morphology features at each time position in the waveform morphology feature time series with the thermal radiation field distribution morphology features at the same time position in the thermal radiation field distribution morphology feature time series to form a corresponding correlation pair of waveform morphology features and thermal radiation field distribution morphology features. Arrange all corresponding correlation pairs in chronological order to form a corresponding correlation pair sequence.

[0038] The nth waveform morphology feature in the waveform morphology feature time series is paired with the nth thermal radiation field distribution morphology feature in the thermal radiation field distribution feature time series to form a corresponding correlation pair. This pairing operation is based on the synchronization of time indices, ensuring that the paired features originate from the same time segment of the same braking process. All corresponding correlation pairs formed in this way are arranged in order of their index n, forming a sequence of corresponding correlation pairs.

[0039] Step S124: Extract the steepness attribute of the rising curve and the smoothness attribute of the falling curve of the waveform morphology feature from each corresponding pair of the corresponding pair sequence. Use the steepness attribute and the smoothness attribute as the acoustic emission waveform morphology attribute set of the corresponding pair. Arrange the acoustic emission waveform morphology attribute sets of all corresponding pairs in chronological order to form an acoustic emission waveform morphology attribute sequence.

[0040] For each corresponding correlation pair in the sequence, two attributes are calculated from the waveform morphology features of the corresponding correlation pair. The steepness attribute of the rising curve is characterized by calculating the average slope of the rising curve from the starting point to the peak point; the larger the average slope value, the steeper the rise. The gentleness attribute of the falling curve is characterized by calculating the absolute value of the average slope of the falling curve from the peak point to the end point; the smaller the absolute value of the average slope value, the gentler the fall. The steepness and gentleness attributes together constitute a two-dimensional attribute vector, which is the set of acoustic emission waveform morphology attributes for the corresponding correlation pair. The sets of acoustic emission waveform morphology attributes for all corresponding correlation pairs are arranged in the order of their corresponding correlation pairs to form an acoustic emission waveform morphology attribute sequence.

[0041] Step S125: Extract the radiation radius length attribute and radiation gradient change attribute of the radial distribution morphology curve of the thermal radiation field distribution morphology feature from each corresponding association pair in the corresponding association pair sequence. Use the radiation radius length attribute and radiation gradient change attribute as the thermal radiation field distribution morphology attribute set of the corresponding association pair. Arrange the thermal radiation field distribution morphology attribute sets of all corresponding association pairs in chronological order to form a thermal radiation field distribution morphology attribute sequence.

[0042] For each corresponding pair in the sequence, two attributes are calculated from the thermal radiation field distribution morphology characteristics of the corresponding pair. The radiation radius length attribute is defined as the radial distance from the geometric center of the temperature field distribution array along any radial direction when the temperature value decays to a predetermined proportion (e.g., 10% of the initial temperature value). The average of these radial distances across all radial directions is taken as the attribute value. The radiation gradient change attribute is defined as the average temperature difference between adjacent radial sampling points on the radial distribution morphology curve. This average value reflects the drastic degree of temperature change along the radial direction. The radiation radius length attribute and the radiation gradient change attribute together constitute a two-dimensional attribute vector, which is the set of thermal radiation field distribution morphology attributes for that corresponding pair. The sets of thermal radiation field distribution morphology attributes for all corresponding pairs are arranged in the order of their corresponding pairs to form a thermal radiation field distribution morphology attribute sequence.

[0043] Step S126: Standardize each attribute in the acoustic emission waveform morphology attribute sequence and each attribute in the thermal radiation field distribution morphology attribute sequence respectively. Construct attribute association pairs based on the acoustic emission waveform morphology attribute set in the standardized acoustic emission waveform morphology attribute sequence and the corresponding thermal radiation field distribution morphology attribute set in the standardized thermal radiation field distribution morphology attribute sequence. Arrange all attribute association pairs in chronological order to form an attribute association pair sequence. Identify the matching degree between the acoustic emission waveform morphology attribute set and the thermal radiation field distribution morphology attribute set of each attribute association pair in the attribute association pair sequence to obtain the morphology matching degree level of each attribute association pair.

[0044] All steepness attribute values ​​and all smoothness attribute values ​​in the acoustic emission waveform morphology attribute sequence are standardized to obtain a standardized acoustic emission waveform morphology attribute sequence. Similarly, all radiation radius length attribute values ​​and all radiation gradient change attribute values ​​in the thermal radiation field distribution morphology attribute sequence are standardized to obtain a standardized thermal radiation field distribution morphology attribute sequence. The nth attribute set in the standardized acoustic emission waveform morphology attribute sequence is combined with the nth attribute set in the standardized thermal radiation field distribution morphology attribute sequence to form an attribute association pair. All attribute association pairs are arranged in order of their index n to form an attribute association pair sequence. For each attribute association pair in the attribute association pair sequence, the comprehensive matching degree between the two attribute values ​​in its acoustic emission waveform morphology attribute set and the two attribute values ​​in its thermal radiation field distribution morphology attribute set is calculated. The comprehensive matching degree is quantified by calculating the Euclidean distance between the two attribute sets; the smaller the distance, the higher the matching degree. The calculated Euclidean distance is compared with several preset matching degree level threshold intervals to determine the morphological matching degree level of the attribute association pair.

[0045] Step S127: Take the morphological change trend between each waveform morphological feature in the waveform morphological feature time series and the waveform morphological feature at its adjacent time position as the waveform morphological evolution path, and connect all the waveform morphological evolution paths in chronological order to form an acoustic emission waveform morphological evolution path diagram.

[0046] For a waveform morphology feature time series, starting from the second waveform morphology feature, the morphological change between the current waveform morphology feature and its predecessor is calculated. This morphological change is quantified by calculating the difference between corresponding points of the rising and falling morphological curves in the two waveform morphology features. This morphological change is considered as a directed edge from one feature node to the next feature node; this directed edge represents the waveform morphology evolution path. All waveform morphology features in the waveform morphology feature time series are treated as nodes, and the waveform morphology evolution paths between all adjacent nodes are treated as directed edges, connected in chronological order to form a directed acyclic graph (DAG). This DAG is the acoustic emission waveform morphology evolution path diagram.

[0047] Step S128: Take the distribution change trend between each thermal radiation field distribution morphology feature in the time series of thermal radiation field distribution morphology features and its adjacent time position as the thermal radiation field distribution evolution path, and connect all thermal radiation field distribution evolution paths in chronological order to form a thermal radiation field distribution evolution path diagram.

[0048] For a time series of thermal radiation field distribution morphology features, starting from the second thermal radiation field distribution morphology feature, the distribution change between the current thermal radiation field distribution morphology feature and its predecessor is calculated. This distribution change is quantified by calculating the temperature difference in the corresponding radial direction and corresponding radial distance of the radial distribution curves in the two thermal radiation field distribution morphology features. This distribution change is considered as a directed edge from one feature node to the next feature node; this directed edge represents the thermal radiation field distribution evolution path. All thermal radiation field distribution morphology features in the time series are treated as nodes, and the thermal radiation field distribution evolution paths between all adjacent nodes are treated as directed edges, connected in chronological order to form a directed acyclic graph (DAG). This DAG is the thermal radiation field distribution evolution path graph.

[0049] Step S129: Overlay the acoustic emission waveform morphology evolution path diagram with the thermal radiation field distribution evolution path diagram according to the time axis, and connect the waveform morphology evolution path nodes and thermal radiation field distribution evolution path nodes at the same time position in the overlaid path diagram with the correlation lines to form an acoustic and thermal energy transfer correlation path diagram.

[0050] The acoustic emission waveform morphology evolution path diagram and the thermal radiation field distribution evolution path diagram share the same time axis. These two diagrams are overlaid in the same coordinate system, aligning nodes with the same time index on the time axis. In the overlaid diagram, for each identical time position, a node in the acoustic emission waveform morphology evolution path diagram is connected to a node in the thermal radiation field distribution evolution path diagram by a connecting line. This connecting line represents the correspondence between the acoustic emission waveform morphology and the thermal radiation field distribution morphology at the same moment. Through these overlay and connection operations, a comprehensive graph structure containing both morphological evolution paths and their interrelationships is finally generated; this graph structure is the acoustic-thermal energy transfer correlation path diagram.

[0051] Step S130: Identify the temporal offset relationship between the inflection point of the acoustic emission waveform morphology change and the inflection point of the thermal radiation field distribution morphology change in the acoustic-thermal energy transfer correlation path diagram, and obtain the temporal offset characteristics of acoustic-thermal energy transfer.

[0052] In this step, an in-depth analysis is conducted on the constructed acoustic-thermal energy transfer correlation path diagram. By identifying key turning points in the two evolutionary paths and quantifying the temporal sequence between these turning points, the temporal dynamic characteristics of energy transfer from acoustic emission to thermal radiation are revealed.

[0053] Step S131: Extract the acoustic emission waveform morphology evolution path diagram from the acoustic heat energy transfer correlation path diagram, identify the position where the rising morphology curve of the waveform changes from steep to gentle in the acoustic emission waveform morphology evolution path diagram, take the position of the morphology change as the first type of turning point of the acoustic emission waveform morphology, and record the time position of the first type of turning point.

[0054] The node sequence in the acoustic emission waveform morphology evolution path diagram is traversed. For each node, the steepness attribute in its acoustic emission waveform morphology attribute set is examined. The rate of change of the steepness attribute between adjacent nodes is calculated. When the steepness attribute continuously decreases from a high value to a low value, and the decrease exceeds a preset first change threshold, the starting node position of the above change is determined as the first type of inflection point in the acoustic emission waveform morphology. The temporal position of this node in the acoustic emission waveform morphology evolution path diagram is recorded.

[0055] Step S132: Extract the acoustic emission waveform morphology evolution path diagram from the acoustic heat energy transfer correlation path diagram, identify the position where the waveform morphology descends from a flat to a steep shape in the acoustic emission waveform morphology evolution path diagram, take the position of the morphology change as the second type of turning point of the acoustic emission waveform morphology, and record the time position of the second type of turning point.

[0056] The node sequence in the acoustic emission waveform morphology evolution path diagram is traversed. For each node, the smoothness attribute in its acoustic emission waveform morphology attribute set is examined. The rate of change of the smoothness attribute between adjacent nodes is calculated. When the smoothness attribute continuously rises from a lower value (i.e., a steeper decrease) to a higher value (i.e., a gentler decrease), and the increase exceeds a preset second change threshold, the starting node position of the above change is determined as the second type of inflection point in the acoustic emission waveform morphology. The temporal position of this node in the acoustic emission waveform morphology evolution path diagram is recorded.

[0057] Step S133: Arrange the time positions of the first type of inflection point of the acoustic emission waveform and the time positions of the second type of inflection point of the acoustic emission waveform in chronological order to form a time position sequence of inflection points of the acoustic emission waveform.

[0058] The time positions of all the first type of inflection points of acoustic emission waveforms identified in step S131 are merged with the time positions of all the second type of inflection points of acoustic emission waveforms identified in step S132. Then, all time positions are sorted in chronological order to form a sequence of time positions of inflection points of acoustic emission waveforms.

[0059] Step S134: Extract the thermal radiation field distribution evolution path map from the acoustic and thermal energy transfer correlation path map, identify the distribution change position where the radiation radius length of the radial distribution pattern curve changes from growth to stagnation in the thermal radiation field distribution evolution path map, take the distribution change position as the first type of turning point of the thermal radiation field distribution pattern, and record the time position of the first type of turning point.

[0060] The node sequence in the thermal radiation field distribution evolution path diagram is traversed. For each node, the radiation radius length attribute in its thermal radiation field distribution morphology attribute set is examined. The rate of change of the radiation radius length attribute between adjacent nodes is calculated. When the radiation radius length attribute continuously decreases from a positive value (indicating growth) to near zero (indicating stagnation), and the decrease exceeds a preset third change threshold, the starting node position of the above change is determined as the first type of turning point in the thermal radiation field distribution morphology. The temporal position of this node in the thermal radiation field distribution evolution path diagram is recorded.

[0061] Step S135: Extract the thermal radiation field distribution evolution path map from the acoustic and thermal energy transfer correlation path map, identify the distribution change position where the radiation gradient of the radial distribution morphology curve changes from increasing to decreasing in the thermal radiation field distribution evolution path map, take the distribution change position as the second type of turning point of thermal radiation field distribution morphology, and record the time position of the second type of turning point.

[0062] The node sequence in the thermal radiation field distribution evolution path diagram is traversed. For each node, the radiation gradient change attribute in its thermal radiation field distribution morphology attribute set is examined. The rate of change of the radiation gradient change attribute between adjacent nodes is calculated. When the radiation gradient change attribute continuously decreases from a large value to a small value, and the decrease exceeds a preset fourth change threshold, the starting node position of the above change is determined as the second type of turning point in the thermal radiation field distribution morphology. The temporal position of this node in the thermal radiation field distribution evolution path diagram is recorded.

[0063] Step S136: Arrange the time positions of the first type of inflection point of the thermal radiation field distribution pattern and the time positions of the second type of inflection point of the thermal radiation field distribution pattern in chronological order to form a time position sequence of inflection points of thermal radiation field distribution pattern.

[0064] The time positions of all the first type of inflection points of thermal radiation field distribution patterns identified in step S134 are merged with the time positions of all the second type of inflection points of thermal radiation field distribution patterns identified in step S135. Then, all time positions are sorted in chronological order to form a sequence of time positions of inflection points of thermal radiation field distribution patterns.

[0065] Step S137: Compare the time position of each inflection point in the acoustic emission waveform morphology inflection point time position sequence with the time position of the inflection point with the same inflection point number in the thermal radiation field distribution morphology inflection point time position sequence to obtain the time sequence relationship between each corresponding inflection point pair, and arrange the time sequence relationships between all corresponding inflection point pairs in order of inflection point number to form a time sequence relationship sequence.

[0066] The inflection points in both the acoustic emission waveform morphology inflection point time sequence and the thermal radiation field distribution morphology inflection point time sequence are numbered sequentially according to their occurrence time. Inflection points with the same index in both sequences are selected, for example, the k-th acoustic emission waveform morphology inflection point and the k-th thermal radiation field distribution morphology inflection point, and their time positions are compared. If the time position of the acoustic emission waveform morphology inflection point is less than that of the thermal radiation field distribution morphology inflection point, the chronological order is recorded as the acoustic emission waveform morphology inflection point leading; otherwise, it is recorded as the thermal radiation field distribution morphology inflection point leading. The chronological order of all corresponding inflection point pairs is arranged according to the inflection point index k, forming a chronological order sequence.

[0067] Step S138: Based on the ratio of the number of inflection point pairs in the time sequence where the inflection point of the acoustic emission waveform shape precedes the inflection point of the thermal radiation field distribution shape to the total number of inflection point pairs in the time sequence, determine the degree of precedence of the acoustic emission waveform shape change over the thermal radiation field distribution shape change. Also, based on the ratio of the number of inflection point pairs where the inflection point of the thermal radiation field distribution shape precedes the inflection point of the acoustic emission waveform shape to the total number of inflection point pairs in the time sequence, determine the degree of precedence of the thermal radiation field distribution shape change over the acoustic emission waveform shape change.

[0068] In the statistical time sequence, the total number of inflection point pairs where the acoustic emission waveform morphology inflection point leads is denoted as the first quantity. Dividing the first quantity by the total number of inflection point pairs in the time sequence yields the ratio indicating the degree to which the acoustic emission waveform morphology change leads the thermal radiation field distribution morphology change. Simultaneously, the total number of inflection point pairs in the statistical time sequence where the thermal radiation field distribution morphology inflection point leads is denoted as the second quantity. Dividing the second quantity by the total number of inflection point pairs in the time sequence yields the ratio indicating the degree to which the thermal radiation field distribution morphology change leads the acoustic emission waveform morphology change.

[0069] Step S139: Compare the degree of the acoustic emission waveform shape change leading the thermal radiation field distribution shape change with the degree of the thermal radiation field distribution shape change leading the acoustic emission waveform shape change to obtain the acoustic-thermal energy transfer time sequence offset feature.

[0070] The time-series shift characteristic of acoustic-thermal energy transfer is a feature quantity that comprehensively reflects the dominant relationship between the time-series changes of two morphological changes. Specifically, it calculates the difference between the leading characteristics of acoustic emission waveform morphological changes and the leading characteristics of thermal radiation field distribution morphological changes. If the difference is positive, it indicates that the acoustic emission waveform morphological changes generally lead the thermal radiation field distribution morphological changes; if the difference is negative, the opposite is true. The absolute value of the difference reflects the strength of the leading effect. The quantification result, composed of the positive and negative signs and the absolute value, constitutes the time-series shift characteristic of acoustic-thermal energy transfer.

[0071] Step S140: The acoustic and thermal energy transfer timing offset features are used to perform path shape correction processing on the acoustic and thermal energy transfer association path diagram to generate a hysteresis coupling path diagram that reflects the energy transfer hysteresis characteristics of the friction pair contact interface.

[0072] In this step, based on the identified temporal offset features, the evolution path in the acoustic-thermal energy transfer correlation path diagram is stretched or compressed on the time axis to align the key nodes of the two morphological changes in time, thereby revealing the hysteresis phenomenon in the energy transfer process.

[0073] Step S141: Determine the first correction direction parameter based on the degree of lead of the acoustic emission waveform morphology change over the thermal radiation field distribution morphology change in the acoustic-thermal energy transfer time sequence offset feature. The first correction direction parameter is used to indicate the path segment that needs to be extended in the positive direction of the time axis in the acoustic emission waveform morphology evolution path diagram.

[0074] The leading characteristic of acoustic emission waveform morphology change in the temporal offset features of acoustic-thermal energy transfer reflects the degree to which the acoustic emission waveform morphology change leads the change in thermal radiation field distribution morphology. When this leading characteristic is greater than zero, it indicates that some path segments in the acoustic emission waveform morphology evolution path diagram are ahead of the corresponding thermal radiation field distribution evolution path in time. In this case, the first correction direction parameter is set to indicate the positive direction of the time axis (i.e., the direction of time increase), used to identify the path segments in the acoustic emission waveform morphology evolution path diagram that need to be extended in the positive direction of the time axis to align them with the thermal radiation field distribution evolution path.

[0075] Step S142: Determine the second correction direction parameter based on the degree of lead of the change in the thermal radiation field distribution shape over the change in the acoustic emission waveform shape in the time sequence offset feature of the acoustic and thermal energy transfer. The second correction direction parameter is used to indicate the path segment that needs to be extended in the positive direction of the time axis in the thermal radiation field distribution evolution path diagram.

[0076] The leading characteristic of the change in the thermal radiation field distribution pattern in the time-series offset features of acoustic-thermal energy transfer reflects the degree to which the change in the thermal radiation field distribution pattern leads the change in the acoustic emission waveform pattern. When this leading characteristic is greater than zero, it indicates that some path segments in the thermal radiation field distribution evolution path map are ahead of the corresponding acoustic emission waveform pattern evolution path in time. In this case, the second correction direction parameter is set to indicate the positive direction of the time axis, which is used to identify the path segments in the thermal radiation field distribution evolution path map that need to be extended in the positive direction of the time axis to align them with the acoustic emission waveform pattern evolution path.

[0077] Step S143: Determine the location region of the path segment that needs to be extended in the acoustic emission waveform morphology evolution path diagram according to the first correction direction parameter, and perform path extension operation on the acoustic emission waveform morphology evolution path in the location region in accordance with the direction indicated by the first correction direction parameter to obtain the acoustic emission waveform morphology evolution path diagram after extension and correction.

[0078] Based on the specific values ​​of the leading degree feature of acoustic emission waveform morphology change in the acoustic-thermal energy transfer time-series offset characteristics, the length of the path segment to be extended is determined. In the acoustic emission waveform morphology evolution path diagram, all nodes marked by the time position sequence of acoustic emission waveform morphology inflection points are located. Using these nodes as centers, a time window proportional to the leading degree feature value is extended in the positive direction of the time axis. The path segment within this time window is the location region to be extended. For each node within this location region, its time coordinate is increased by a correction amount. The magnitude of the correction amount is related to the distance of the node from the starting point of the window and the leading degree feature value, thereby stretching the path on the time axis. After all node coordinates are corrected, the nodes are reconnected to form the extended and corrected acoustic emission waveform morphology evolution path diagram.

[0079] Step S144: Determine the location region of the path segment that needs to be extended in the thermal radiation field distribution evolution path map according to the second correction direction parameter, and extend the thermal radiation field distribution evolution path in the location region according to the direction indicated by the second correction direction parameter to obtain the thermal radiation field distribution evolution path map after extension and correction.

[0080] Based on the specific values ​​of the leading characteristic of the change in the thermal radiation field distribution pattern in the time-series offset characteristics of acoustic and thermal energy transfer, the length of the path segment to be extended is determined. In the thermal radiation field distribution evolution path diagram, all nodes marked by the time position sequence of the inflection point of the thermal radiation field distribution pattern are located. Using these nodes as centers, a time window proportional to the leading characteristic value is extended in the positive direction of the time axis. The path segment within this time window is the location region to be extended. For each node within this location region, a correction amount is added to its time coordinate. The magnitude of the correction amount is related to the distance of the node from the starting point of the window and the leading characteristic value, thereby stretching the path on the time axis. After all node coordinates are corrected, the nodes are reconnected to form the extended and corrected thermal radiation field distribution evolution path diagram.

[0081] Step S145: Overlay the extended and corrected acoustic emission waveform morphology evolution path diagram and the extended and corrected thermal radiation field distribution evolution path diagram on the time axis, so that the inflection points with the same inflection point number in the two evolution path diagrams after the extension and correction are aligned in time position, to obtain a preliminary aligned acoustic and thermal energy transfer correlation path diagram, and identify the degree of morphological matching between the acoustic emission waveform morphology evolution path and the thermal radiation field distribution evolution path at the same time position, and take the path nodes with a morphological matching degree lower than the preset matching threshold as the node positions that need further correction, forming a set of node positions to be corrected.

[0082] The extended and corrected acoustic emission waveform morphology evolution path diagram and the extended and corrected thermal radiation field distribution evolution path diagram are re-overlaid on the same time axis. Due to the extension and correction operation, the previously time-misaligned inflection points are now basically aligned in time position. For each identical time position in the overlaid diagram, the acoustic emission waveform morphology node and the thermal radiation field distribution morphology node at that position are extracted, and the morphological matching degree between the acoustic emission waveform morphology attribute set and the thermal radiation field distribution morphology attribute set of the two nodes is calculated (e.g., calculating the Euclidean distance of the attribute sets). The calculated morphological matching degree is compared with a preset matching threshold. All node positions with a morphological matching degree lower than the above preset matching threshold are marked as node positions that need further correction, and all these node positions constitute the set of node positions to be corrected.

[0083] Step S146: Based on the difference direction between the acoustic emission waveform morphology attribute and the thermal radiation field distribution morphology attribute at each node position in the set of node positions to be corrected, determine the local correction direction parameter at each node position to be corrected. According to the local correction direction parameter at each node position to be corrected, perform local path morphology fine-tuning on the acoustic emission waveform morphology evolution path or thermal radiation field distribution evolution path at the corresponding node position in the initially aligned acoustic-thermal energy transfer association path diagram, so that the matching degree between the acoustic emission waveform morphology and the thermal radiation field distribution morphology at the node position is improved after fine-tuning.

[0084] For each node in the set of nodes to be corrected, the difference between the acoustic emission waveform morphology and the thermal radiation field distribution morphology at that node is analyzed. Specifically, the deviation directions of the steepness attribute and the radiation radius length attribute, as well as the deviation directions of the smoothness attribute and the radiation gradient change attribute, are compared. Based on these deviation directions, local correction direction parameters are determined. For example, if the steepness attribute of the acoustic emission waveform morphology is greater than the normalized value of the radiation radius length attribute of the thermal radiation field distribution morphology, the local correction direction parameter indicates that the acoustic emission waveform morphology evolution path at that node needs to be fine-tuned in the direction of decreasing morphology attribute value. Based on the above local correction direction parameter, a small displacement is made to the position of that node in the evolution path diagram, the magnitude of which is proportional to the degree of deviation. By iterating the above fine-tuning process, the morphology matching degree at that node is improved to above a preset matching threshold. The above fine-tuning operation is performed on all nodes to be corrected to obtain the acoustic and thermal energy transfer correlation path diagram after local path morphology fine-tuning.

[0085] Step S147: The distribution of the degree of morphological matching between the acoustic emission waveform morphology evolution path and the thermal radiation field distribution evolution path in the acoustic-thermal energy transfer correlation path diagram after local path morphology fine-tuning is taken as the hysteresis degree distribution feature, and the hysteresis degree distribution feature is embedded into the acoustic-thermal energy transfer correlation path diagram after local path morphology fine-tuning, and the hysteresis coupling path diagram reflecting the energy transfer hysteresis characteristics of the friction pair contact interface is output.

[0086] In the acoustic-thermal energy transfer path diagram after local path morphology fine-tuning, there exists a final morphology matching degree value between the acoustic emission waveform morphology node and the thermal radiation field distribution morphology node at each time position. These morphology matching degree values ​​at all time positions are arranged chronologically to form a hysteresis degree distribution feature. This hysteresis degree distribution feature is then used as attribute information for each node at each time position and embedded into the acoustic-thermal energy transfer path diagram after local path morphology fine-tuning. Specifically, a "hysteresis degree" attribute field is added to each node in the diagram (including acoustic emission waveform morphology nodes and thermal radiation field distribution morphology nodes), and its value is the morphology matching degree value at that time position. The graph structure output after the above embedding operation is a hysteresis coupling path diagram reflecting the hysteresis characteristics of energy transfer at the friction pair contact interface.

[0087] Step S150: Determine the wear status level of the contact interface of the brake friction pair based on the morphological difference between the hysteresis segment of the acoustic emission waveform and the leading segment of the thermal radiation field distribution in the hysteresis coupling path diagram.

[0088] In this step, the hysteresis coupling path diagram is analyzed in depth. By identifying the path segment (hysteresis segment) with the most significant hysteresis phenomenon in the diagram and the corresponding leading segment, the morphological difference between the two segments is quantified, and this difference is mapped to the wear state level of the friction pair.

[0089] Step S151: Extract the path segment from the hysteresis coupling path diagram where the hysteresis degree distribution characteristic value in the acoustic emission waveform morphology evolution path exceeds the preset hysteresis threshold, take the path segment as the acoustic emission waveform morphology hysteresis segment, and record the start and end positions of the hysteresis segment in the acoustic emission waveform morphology evolution path.

[0090] In the hysteresis coupling path graph, all nodes on the acoustic emission waveform morphology evolution path are traversed, and the hysteresis degree attribute value of each node is read. The path segment consisting of a sequence of nodes whose hysteresis degree attribute values ​​continuously exceed a preset hysteresis threshold (which represents the upper limit of acceptable energy transfer synchronicity) is marked as an acoustic emission waveform morphology hysteresis segment. The starting and ending node indices of this path segment in the acoustic emission waveform morphology evolution path are recorded as the position information of the hysteresis segment.

[0091] Step S152: Extract path segments from the hysteresis coupling path diagram where the hysteresis degree distribution characteristic value is lower than the preset hysteresis threshold in the thermal radiation field distribution evolution path. Take the path segment as the leading segment of the thermal radiation field distribution form and record the starting and ending positions of the leading segment in the thermal radiation field distribution evolution path.

[0092] In the hysteresis coupling path graph, all nodes on the thermal radiation field distribution evolution path are traversed, and the hysteresis degree attribute value of each node is read. The path segment formed by the sequence of nodes whose hysteresis degree attribute values ​​are continuously lower than the above-mentioned preset hysteresis threshold is marked as the leading segment of the thermal radiation field distribution pattern. The starting node index and ending node index of this path segment in the thermal radiation field distribution evolution path are recorded as the position information of this leading segment.

[0093] Step S153: Compare the starting position of the acoustic emission waveform hysteresis segment with the starting position of the thermal radiation field distribution leading segment to obtain the first relative offset direction of the starting position of the acoustic emission waveform hysteresis segment relative to the starting position of the thermal radiation field distribution leading segment.

[0094] Compare the time index of the starting node of the hysteresis segment of the acoustic emission waveform morphology in the acoustic emission waveform morphology evolution path with the time index of the starting node of the leading segment of the thermal radiation field distribution morphology in the thermal radiation field distribution evolution path. If the starting time index of the hysteresis segment is less than the starting time index of the leading segment, the first relative offset direction is recorded as the hysteresis segment starting earlier than the leading segment starting; otherwise, it is recorded as the hysteresis segment starting later than the leading segment starting.

[0095] Step S154: Compare the termination position of the acoustic emission waveform hysteresis segment with the termination position of the thermal radiation field distribution leading segment to obtain the second relative offset direction of the termination position of the acoustic emission waveform hysteresis segment relative to the termination position of the thermal radiation field distribution leading segment.

[0096] Compare the time index of the termination node of the hysteresis segment of the acoustic emission waveform morphology in the acoustic emission waveform morphology evolution path with the time index of the termination node of the leading segment of the thermal radiation field distribution morphology in the thermal radiation field distribution evolution path. If the termination time index of the hysteresis segment is less than the termination time index of the leading segment, the second relative offset direction is recorded as the termination of the hysteresis segment leading the termination of the leading segment; otherwise, it is recorded as the termination of the hysteresis segment lagging behind the termination of the leading segment.

[0097] Step S155: Determine the relative positional relationship between the acoustic emission waveform morphology hysteresis segment and the thermal radiation field distribution morphology leading segment in the hysteresis coupling path diagram based on the first relative offset direction and the second relative offset direction, and quantify the relative positional relationship as a first morphological difference characteristic value.

[0098] The first and second relative offset directions together describe the relative positional relationship between the hysteresis segment and the leading segment on the time axis. If both are "the start of the hysteresis segment precedes the start of the leading segment" and "the end of the hysteresis segment lags behind the end of the leading segment," it indicates that the hysteresis segment covers and exceeds the leading segment in time, and this relative positional relationship is quantified as a larger first morphological difference feature value (e.g., set as value A). If both are "the start of the hysteresis segment lags behind the start of the leading segment" and "the end of the hysteresis segment precedes the end of the leading segment," it indicates that the hysteresis segment is completely contained by the leading segment, and this relative positional relationship is quantified as a smaller first morphological difference feature value (e.g., set as value B). Other combinations correspond to intermediate values.

[0099] Step S156: Extract the average steepness attribute of the rising curve and the average smoothness attribute of the falling curve of the acoustic emission waveform morphology within the hysteresis segment from the hysteresis coupling path diagram, and use the average steepness attribute and the average smoothness attribute as the morphological attribute features of the hysteresis segment of the acoustic emission waveform morphology.

[0100] For all nodes within the hysteresis segment of the acoustic emission waveform, calculate the average value of the steepness attribute and the average value of the smoothness attribute in the acoustic emission waveform morphological attribute set corresponding to these nodes. These two average values ​​together form a two-dimensional vector, which represents the morphological attribute feature of the hysteresis segment of the acoustic emission waveform.

[0101] Step S157: Extract the average growth rate attribute of the radiation radius length and the average change amplitude attribute of the radiation gradient of the thermal radiation field distribution pattern in the leading segment of the thermal radiation field distribution pattern from the hysteresis coupling path diagram, and use the average growth rate attribute and the average change amplitude attribute as the morphological attribute features of the leading segment of the thermal radiation field distribution pattern.

[0102] For all nodes within the leading segment of the thermal radiation field distribution pattern, calculate the first-order difference average of the radiation radius length attribute in the corresponding thermal radiation field distribution pattern attribute set for each node, as the average growth rate attribute; calculate the average value of the radiation gradient change attribute, as the average change amplitude attribute. The above average growth rate attribute and average change amplitude attribute together constitute a two-dimensional vector, which is the pattern attribute feature of the leading segment of the thermal radiation field distribution pattern.

[0103] Step S158: Standardize the morphological attribute features of the acoustic emission waveform hysteresis segment and the morphological attribute features of the thermal radiation field distribution leading segment, respectively. Calculate the attribute difference between the standardized morphological attribute features of the acoustic emission waveform hysteresis segment and the standardized morphological attribute features of the thermal radiation field distribution leading segment to obtain a second morphological difference feature value. According to a preset weight, weightedly fuse the first morphological difference feature value and the second morphological difference feature value to obtain the comprehensive morphological difference between the acoustic emission waveform hysteresis segment and the thermal radiation field distribution leading segment.

[0104] Z-score standardization is applied to the two components of the morphological attribute features (two-dimensional vector) of the hysteresis segment of the acoustic emission waveform, yielding the standardized hysteresis segment morphological attribute feature vector. Similarly, Z-score standardization is applied to the two components of the morphological attribute features of the leading segment of the thermal radiation field distribution, yielding the standardized leading segment morphological attribute feature vector. The Euclidean distance between the two standardized vectors is calculated; this distance value is the second morphological difference feature value. Subsequently, the first and second morphological difference feature values ​​are weighted and fused. The fusion formula is: the comprehensive morphological difference equals the first weight coefficient multiplied by the first morphological difference feature value plus the second weight coefficient multiplied by the second morphological difference feature value. The sum of the first and second weight coefficients is 1; the specific values ​​can be preset according to the emphasis on spatial location differences and morphological attribute differences in the actual application scenario.

[0105] Step S159: Compare the morphological difference degree with the morphological difference degree range corresponding to multiple preset wear state levels, determine the wear state level range to which the morphological difference degree belongs, and take the wear state level corresponding to the wear state level range as the wear state level of the contact interface of the brake friction pair.

[0106] A mapping table is pre-established, defining multiple consecutive wear condition levels (e.g., light wear, moderate wear, heavy wear), with each wear condition level corresponding to a numerical range of morphological difference. The calculated overall morphological difference is compared with the ranges in the mapping table to determine which wear condition level's numerical range the overall morphological difference falls into. This wear condition level is then determined as the current wear condition level of the brake friction pair's contact interface.

[0107] Step S160: Determine the wear status level of the contact interface of the brake friction pair based on the morphological difference between the hysteresis segment of the acoustic emission waveform and the leading segment of the thermal radiation field distribution in the hysteresis coupling path diagram.

[0108] This step aims to conduct a more refined evaluation of the wear state level determined in the aforementioned step S150. By introducing factors such as the transition region characteristics between the hysteresis and leading stages, the time length ratio, the comparison of the severity of morphological changes, and the matching of the hysteresis degree distribution pattern, a more comprehensive wear state determination system is constructed.

[0109] Step S161: Extract the transition region morphological features between the rising and falling curves of the acoustic emission waveform morphology within the hysteresis segment from the hysteresis coupling path diagram, and use these transition region morphological features as the transition region features of the hysteresis segment of the acoustic emission waveform morphology.

[0110] Within the hysteresis segment of the acoustic emission waveform, for each acoustic emission waveform morphology node, the region between the end of its rising morphology curve (i.e., the waveform feature anchor point) and the beginning of its falling morphology curve is identified. The curvature change pattern within this region is calculated, including the number of inflection points where curvature changes from positive to negative and the magnitude of the curvature change. This curvature change pattern is quantified to form the transition region characteristics of the hysteresis segment of the acoustic emission waveform morphology, describing the transition region morphology.

[0111] Step S162: Extract the morphological features of the connecting region between the radiation radius growth region and the radiation gradient change region in the radial distribution curve of the thermal radiation field distribution pattern in the leading segment of the hysteresis coupling path diagram, and use the morphological features of the connecting region as the morphological features of the connecting region of the leading segment of the thermal radiation field distribution pattern.

[0112] Within the leading segment of the thermal radiation field distribution pattern, for each node in the thermal radiation field distribution pattern, the transition region between the end of its radiation radius growth phase and the beginning of its radiation gradient change phase is identified. The coupling relationship between the radiation radius growth rate and the radiation gradient change rate within this region is calculated, such as the ratio or correlation between their rates of change. This coupling relationship is quantified to form the transition region morphological characteristics of the leading segment of the thermal radiation field distribution pattern, describing the morphology of the transition region.

[0113] Step S163: Calculate the regional morphological difference between the transition region characteristics of the hysteresis segment of the acoustic emission waveform and the connection region characteristics of the leading segment of the thermal radiation field distribution to obtain the third morphological difference characteristic value.

[0114] The cosine similarity between the feature vectors of the transition region of the acoustic emission waveform hysteresis segment and the feature vectors of the connecting region of the leading segment of the thermal radiation field distribution is calculated. The cosine similarity value ranges from -1 to 1; the smaller the value, the greater the directional difference between the two feature vectors, i.e., the greater the morphological difference. The calculated cosine similarity is then transformed, for example, by subtracting the cosine similarity from 1, to obtain the third morphological difference feature value. The larger this third morphological difference feature value, the greater the regional morphological difference.

[0115] Step S164: Extract the ratio of the duration of the hysteresis segment of the acoustic emission waveform on the time axis to the duration of the leading segment of the thermal radiation field distribution on the time axis from the hysteresis coupling path diagram, and use this ratio as the characteristic value of the time length ratio between the hysteresis segment and the leading segment.

[0116] Calculate the time span of the hysteresis segment in the acoustic emission waveform morphology within the hysteresis coupling path diagram, i.e., subtract the time index of the starting node from the time index of the terminating node, to obtain the duration of the hysteresis segment. Similarly, calculate the duration of the leading segment in the thermal radiation field distribution morphology. Divide the duration of the hysteresis segment by the former to obtain the time length ratio characteristic value.

[0117] Step S165: Extract the ratio of the drastic change in the acoustic emission waveform morphology within the hysteresis segment to the drastic change in the thermal radiation field distribution morphology within the leading segment from the hysteresis coupling path diagram, and use this ratio as a comparative feature value of the drastic change in the morphology between the hysteresis segment and the leading segment.

[0118] The average Euclidean distance of the acoustic emission waveform morphology attribute set between all adjacent nodes within the hysteresis segment is calculated as the degree of morphological change drastically in the hysteresis segment. Similarly, the average Euclidean distance of the thermal radiation field distribution morphology attribute set between all adjacent nodes within the leading segment is calculated as the degree of morphological change drastically in the leading segment. Dividing the degree of morphological change drastically in the hysteresis segment by the former yields a comparative characteristic value for the degree of morphological change drastically.

[0119] Step S166: Standardize the third morphological difference feature value, the time length ratio feature value, and the morphological change intensity comparison feature value. Combine the standardized third morphological difference feature value, time length ratio feature value, and morphological change intensity comparison feature value to construct a morphological difference vector between the lag segment and the leading segment. Use this morphological difference vector as a refined representation of the comprehensive morphological difference.

[0120] Z-score standardization was applied to the third morphological difference feature value, the time length ratio feature value, and the morphological change intensity comparison feature value to ensure they are on the same dimension. The three standardized feature values ​​were then concatenated sequentially to form the morphological difference vector between the hysteresis segment and the leading segment.

[0121] Step S167: Compare the morphological difference vector with the morphological difference vector reference range corresponding to multiple preset wear state levels in vector space to determine the vector space region to which the morphological difference vector belongs, and take the wear state level corresponding to the vector space region as the first intermediate wear state level.

[0122] A reference database is pre-established, which stores the typical distribution ranges (e.g., convex hull regions in three-dimensional space) of the morphological difference vectors corresponding to multiple wear state levels (e.g., light wear, moderate wear, heavy wear). The distance from the current morphological difference vector to each typical distribution range (e.g., the shortest distance to the convex hull boundary) is calculated, and the wear state level corresponding to the closest distribution range is selected as the first intermediate wear state level.

[0123] Step S168: Extract the spatial distribution pattern of the hysteresis degree distribution feature value in the acoustic emission waveform morphology evolution path from the hysteresis coupling path diagram, and use the spatial distribution pattern as the hysteresis degree distribution pattern feature of the acoustic emission waveform morphology.

[0124] The hysteresis attribute values ​​of all nodes along the acoustic emission waveform morphology evolution path are considered as a one-dimensional sequence. Time series analysis is performed on this sequence to extract its statistical characteristics, such as mean, variance, skewness, kurtosis, and autocorrelation coefficient. These statistical characteristics collectively constitute the acoustic emission waveform morphology hysteresis distribution pattern characteristics, describing the distribution pattern of hysteresis along the acoustic emission waveform morphology evolution path.

[0125] Step S169: Extract the spatial distribution pattern of the hysteresis degree distribution feature value in the thermal radiation field distribution evolution path from the hysteresis coupling path diagram, and use the spatial distribution pattern as the hysteresis degree distribution pattern feature of the thermal radiation field distribution morphology.

[0126] Similarly, the hysteresis attribute values ​​of all nodes along the evolution path of the thermal radiation field distribution are considered as a one-dimensional sequence. The same time series analysis as in step S168 is performed on this one-dimensional sequence to extract statistical features such as mean, variance, skewness, kurtosis, and autocorrelation coefficient, which constitute a feature vector of the hysteresis distribution pattern of the thermal radiation field distribution.

[0127] Step S1610: Perform pattern matching degree identification on the acoustic emission waveform hysteresis distribution pattern feature and the thermal radiation field distribution hysteresis distribution pattern feature to obtain the pattern matching degree level, and correct the first intermediate wear state level according to the pattern matching degree level to obtain the wear state level of the contact interface of the brake friction pair.

[0128] The dynamic time warping distance between the feature vectors of the acoustic emission waveform hysteresis distribution pattern and the thermal radiation field hysteresis distribution pattern is calculated. The dynamic time warping distance measures the similarity between two time series of different lengths. The calculated dynamic time warping distance is compared with several preset pattern matching degree threshold ranges to determine the pattern matching degree. According to preset correction rules, the pattern matching degree is used as a correction factor to adjust the first intermediate wear state level. For example, if the pattern matching degree is low (indicating a large difference between the two distribution patterns), the wear state level may be increased by one sub-level; if the pattern matching degree is high, it remains unchanged or is decreased. The final output wear state level is the wear state level of the contact interface of the brake friction pair after the above correction.

[0129] Step S171: Extract the evolution trend of the morphological difference between the hysteresis segment of the acoustic emission waveform and the leading segment of the thermal radiation field distribution on the time axis from the hysteresis coupling path diagram, and use this evolution trend as the morphological difference time evolution curve.

[0130] Arrange the comprehensive morphological dissimilarity values ​​(or the norm of the morphological dissimilarity vector) at all time points in the hysteresis coupling path diagram in chronological order to form a sequence. Smooth this sequence (e.g., using a moving average filter) to remove noise, and the resulting smooth curve is the morphological dissimilarity time evolution curve.

[0131] Step S172: Identify the time point in the morphological difference time evolution curve where the morphological difference value changes abruptly, take the time point as the morphological difference change point, and record the direction of change of morphological difference before and after the change point.

[0132] Calculate the first derivative of the morphological difference over time. Identify the time points where the absolute value of the first derivative exceeds a preset mutation threshold as mutation points. Record the sign of the first derivative before and after the mutation point; a positive sign indicates that the morphological difference increases after that point, and a negative sign indicates that it decreases. This sign indicates the direction of change in morphological difference.

[0133] Step S173: Determine the change direction of the wear state of the brake friction pair contact interface based on the change direction of the morphological difference before and after the abrupt change point, and take this change direction as the wear state change direction feature.

[0134] If the direction of change in morphological difference before and after the abrupt change is positive, it indicates that the wear condition is intensifying (e.g., changing from moderate wear to heavy wear); if the direction of change is negative, it indicates that the wear condition is alleviating (e.g., changing from heavy wear to moderate wear). The above-mentioned change direction is recorded as the wear condition change direction characteristic.

[0135] Step S174: Extract the ratio of the distribution density of the hysteresis segment of the acoustic emission waveform on the time axis to the distribution density of the leading segment of the thermal radiation field on the time axis from the hysteresis coupling path diagram, and use this ratio as the distribution density ratio feature of the hysteresis segment and the leading segment.

[0136] The ratio of the total duration of the hysteresis segment of the acoustic emission waveform to the entire monitoring time is calculated as the distribution density of the hysteresis segment. Similarly, the ratio of the total duration of the leading segment of the thermal radiation field distribution to the entire monitoring time is calculated as the distribution density of the leading segment. Dividing the distribution density of the hysteresis segment by the former yields the distribution density ratio characteristic.

[0137] Step S175: Extract the repetition pattern matching degree between the morphological feature repetition pattern of the acoustic emission waveform in the hysteresis segment and the morphological feature repetition pattern of the thermal radiation field distribution in the leading segment from the hysteresis coupling path diagram, and use the repetition pattern matching degree as the morphological feature repetition pattern matching degree feature.

[0138] The time series consisting of the acoustic emission waveform morphology attribute set of each node in the hysteresis segment of the acoustic emission waveform and the time series consisting of the thermal radiation field distribution morphology attribute set of each node in the leading segment of the thermal radiation field distribution are extracted separately. The similarity between the two time series is calculated using a dynamic time warping algorithm, and this similarity value is the morphological feature repetition pattern matching degree feature.

[0139] Step S176: Construct a dynamic evolution feature vector of wear state based on the wear state change direction feature, the distribution density ratio feature, and the morphological feature repetition pattern matching degree feature.

[0140] The wear state change direction features (quantified into numerical values, e.g., +1 for positive direction and -1 for negative direction), distribution density ratio features, and morphological repetition pattern matching degree features are concatenated into a wear state dynamic evolution feature vector.

[0141] Step S177: Compare the dynamic evolution feature vector of the wear state with the reference range of the dynamic evolution feature vector of the wear state corresponding to multiple preset wear state evolution stages in vector space to determine the vector space region to which the dynamic evolution feature vector of the wear state belongs, and take the wear state evolution stage corresponding to the vector space region as the wear state evolution stage of the contact interface of the brake friction pair.

[0142] A reference database is pre-established, storing the typical distribution ranges of the dynamic evolution feature vectors of wear states for each of the multiple wear state evolution stages (e.g., initial break-in stage, stable wear stage, severe wear stage). The distance from the current wear state dynamic evolution feature vector to each typical distribution range is calculated, and the wear state evolution stage corresponding to the closest distribution range is selected.

[0143] Step S178: Determine the expected change trend of the wear state level of the contact interface of the brake friction pair according to the wear state evolution stage, and use the expected change trend as auxiliary judgment information for brake operation state monitoring.

[0144] Based on the identified wear state evolution stages, combined with domain knowledge or historical data statistical patterns, the future wear state level change trend is predicted. For example, if the current stage is stable wear, the expected trend may be a slow increase in the wear state level; if the current stage is severe wear, the expected trend may be a rapid increase in the wear state level. These expected trends are output as auxiliary information for monitoring the brake's operating status.

[0145] Step S179: Extract the curve showing the change in the degree of morphological matching between the acoustic emission waveform morphology evolution path and the thermal radiation field distribution evolution path over time from the acoustic-thermal energy transfer correlation path diagram.

[0146] In the acoustic-thermal energy transfer path diagram, there is a degree of morphological matching between the acoustic emission waveform morphology node and the thermal radiation field distribution morphology node at each time position (i.e., the degree of matching calculated in step S145). The morphological matching degree values ​​at all time positions are arranged in chronological order to form a curve of matching degree change.

[0147] Step S1710: Identify the transition point in the matching degree change curve where the matching degree value changes from a high matching state to a low matching state, take the transition point as the starting point of energy transfer path distortion, and record the time position of the distortion starting point.

[0148] The first derivative of the matching degree change curve is calculated, and a high matching state threshold and a low matching state threshold are set. When the matching degree value continuously decreases from above the high matching state threshold to below the low matching state threshold, the time position corresponding to the first point in the decrease process that falls below the low matching state threshold is marked as the starting point of energy transfer path distortion.

[0149] Step S1711: Extract the morphological matching trend between the acoustic emission waveform morphology evolution path and the thermal radiation field distribution evolution path in the time region after the distortion start point from the acoustic-thermal energy transfer correlation path diagram, and use this recovery trend as the self-recovery capability feature of the energy transfer path.

[0150] After the point where the energy transfer path distortion begins, the matching degree change curve is continuously monitored. Matching degree change data is extracted from the point of distortion until the matching degree value recovers to above the high-matching threshold. The average rate of increase of the matching degree value during this period is calculated; this average rate of increase is the self-recovery capability characteristic of the energy transfer path. The faster the rate of increase, the stronger the self-recovery capability.

[0151] Step S1712: Determine the distortion sensitivity level of the energy transfer path at the contact interface of the brake friction pair based on the time position of the distortion start point of the energy transfer path and the self-recovery capability characteristics of the energy transfer path.

[0152] The earlier the distortion initiation point of the energy transfer path occurs, and the smaller the self-recovery capability characteristic value of the energy transfer path (i.e., the slower the recovery), the higher the distortion sensitivity level. A two-dimensional mapping table can be established to combine the distortion initiation point time location and the self-recovery capability characteristic value to different distortion sensitivity levels (e.g., high sensitivity, medium sensitivity, low sensitivity).

[0153] Step S1713: Extract the variation law of the morphological difference between the hysteresis segment of the acoustic emission waveform and the leading segment of the thermal radiation field distribution from the hysteresis coupling path diagram during multiple braking processes, and use this variation law as the feature of wear accumulation during multiple braking processes.

[0154] For multiple consecutive braking processes, the comprehensive morphological difference (or morphological difference vector) corresponding to each braking process is calculated. The comprehensive morphological difference values ​​of these multiple braking processes are arranged in chronological order to form a sequence. This sequence is then fitted to obtain its trend (e.g., linear growth, exponential growth, logarithmic growth, etc.). This trend represents the cumulative wear characteristics of multiple braking processes.

[0155] Step S1714: Based on the distortion sensitivity level and the cumulative wear characteristics of multiple braking cycles, determine the cumulative change trend of the wear state level of the contact interface of the brake friction pair, and use this cumulative change trend as the basis for predicting the remaining service life of the brake.

[0156] The distortion sensitivity level is used as a weighting factor to correct the wear accumulation rate described by the cumulative wear characteristics of multiple braking cycles. For example, if the distortion sensitivity level is high, the wear accumulation rate is increased by a proportional coefficient; if it is low, it is decreased. The corrected wear accumulation rate, combined with the current wear status level, can be extrapolated to predict the time required to reach the preset scrap threshold (i.e., the highest wear status level). This time is the predicted value of the remaining service life.

[0157] For example, the method may further include: step S181: extracting the time interval between the peak point of the waveform amplitude and the starting point of the waveform of each acoustic emission waveform unit from the acoustic emission waveform time sequence, using the time interval as the waveform rise time feature of the acoustic emission waveform unit, and arranging the waveform rise time features of all acoustic emission waveform units in chronological order to form a waveform rise time feature sequence.

[0158] For each acoustic emission waveform unit in the acoustic emission waveform time sequence, calculate the time difference between its waveform feature anchor point (i.e., the moment when the waveform amplitude reaches its maximum value) and the waveform start point. This difference is the waveform rise time feature of that acoustic emission waveform unit. Arrange the waveform rise time features of all acoustic emission waveform units according to the order of their corresponding acoustic emission waveform units in the acoustic emission waveform time sequence to form a waveform rise time feature sequence.

[0159] Step S182: Extract the temperature difference between the highest temperature element and the lowest temperature element in the temperature field distribution array of each frame image from the time sequence of infrared thermal radiation field images. Use this temperature difference as the thermal radiation field temperature span feature of the frame image. Arrange the thermal radiation field temperature span features of all frames images in time order to form a thermal radiation field temperature span feature sequence.

[0160] For each frame of an infrared thermal radiation field image unit in the time sequence of infrared thermal radiation field images, the temperature value of the element point with the largest temperature value (i.e., the thermal radiation field feature anchor point) is found from the temperature field distribution array of the infrared thermal radiation field image unit, and the temperature value of the element point with the smallest temperature value is also found. The difference between these two temperature values ​​is calculated, and this difference is the thermal radiation field temperature span feature of that frame of image. The thermal radiation field temperature span features of all frames of images are arranged according to the order of their corresponding infrared thermal radiation field image units in the time sequence of infrared thermal radiation field images to form a thermal radiation field temperature span feature sequence.

[0161] Step S183: Associate the waveform rise time feature sequence with the waveform rise time feature and the thermal radiation field temperature span feature sequence that have the same time number to form a correlation pair sequence of rise time and temperature span.

[0162] Take the nth feature value from the waveform rise time feature sequence and combine it with the nth feature value from the thermal radiation field temperature span feature sequence to form a correlation pair. All correlation pairs formed in this way are arranged in order of their index n to form a correlation pair sequence of rise time and temperature span.

[0163] Step S184: Identify the degree of trend consistency between the trend of waveform rise time characteristics and the trend of thermal radiation field temperature span characteristics from the correlation pair sequence, and take the degree of trend consistency as the trend consistency characteristic of acoustic and thermal energy transfer.

[0164] Calculate the first-order difference sequences (i.e., the differences between adjacent feature values) for both the waveform rise time characteristic sequence and the thermal radiation field temperature span characteristic sequence. Then calculate the Pearson correlation coefficient between these two first-order difference sequences. This correlation coefficient reflects the synchronicity of the changing trends of the two features; the closer its value is to 1, the more consistent the changing trends. This correlation coefficient represents the consistency characteristic of the acoustic and thermal energy transfer trend.

[0165] Step S185: Determine the energy transfer stability level of the brake friction pair contact interface based on the consistency characteristics of the acoustic and thermal energy transfer trend, and use the energy transfer stability level as supplementary information to the brake operating status monitoring results.

[0166] Multiple energy transfer stability levels are pre-defined, each corresponding to a numerical range of consistent acoustic and thermal energy transfer trends. For example, if the correlation coefficient is greater than a first threshold, it is classified as a high stability level; if it is between the first and second thresholds, it is classified as a medium stability level; and if it is less than the second threshold, it is classified as a low stability level. These stability levels serve as supplementary information to the brake's operating status monitoring results, used to comprehensively evaluate brake performance.

[0167] For example, the method may further include: step S191: extracting the morphological difference time evolution curves of the acoustic emission waveform morphological hysteresis segment and the thermal radiation field distribution morphological leading segment from the hysteresis coupling path diagram, and inputting the morphological difference time evolution curves into a sequence prediction model based on a hybrid structure of convolutional neural network and long short-term memory network.

[0168] The morphological difference time evolution curve described in step S171 is extracted from the hysteresis coupling path graph and used as input data to a pre-built and trained sequence prediction model. The main architecture of this sequence prediction model consists of stacked convolutional neural network layers and long short-term memory network layers.

[0169] Step S192: The sequence prediction model extracts local time pattern features from the input morphological difference time evolution curve through its convolutional neural network layer, and obtains multiple morphological difference local feature vectors at different time scales. The multiple morphological difference local feature vectors are concatenated along the time dimension to generate a multi-scale morphological difference feature sequence.

[0170] In the sequence prediction model, the input morphological difference time evolution curve is first fed into a convolutional neural network layer containing multiple parallel convolutional kernels. Each convolutional kernel has a different size (e.g., sizes 3, 5, and 7) to capture local waveform patterns at different time scales. Each convolutional kernel slides across the input sequence, performing convolution operations to generate a corresponding feature map. These feature maps are the local feature vectors of morphological difference at the corresponding time scale. Subsequently, these multiple feature maps are concatenated along the time dimension to form a higher-dimensional multi-scale morphological difference feature sequence.

[0171] Step S193: The sequence prediction model performs time-series dependency modeling on the multi-scale morphological difference feature sequence through its long short-term memory network layer, captures the long-term and short-term evolution relationship of the local feature vector of morphological difference in the time series, and generates a time-series state vector that reflects the inherent evolution law of morphological difference.

[0172] The multi-scale morphological dissimilarity feature sequence generated in step S192 is input into a Long Short-Term Memory (LSTM) network layer. The forget gate, input gate, and output gate within the LTM layer work together to process each time step in the sequence and update its internal state. After traversing the entire input sequence, the hidden state of the last time step in the LTM layer, or the vector obtained by pooling the hidden states of all time steps, is used as the temporal state vector. This vector encodes the long-term and short-term dependencies of the entire input sequence, reflecting the intrinsic evolutionary law of morphological dissimilarity.

[0173] Step S194: Extract the time series of the leading degree feature of the acoustic emission waveform morphology change leading the thermal radiation field distribution morphology change from the acoustic and thermal energy transfer time series offset features. Input the time series of the leading degree feature into the attention mechanism layer in the sequence prediction model. The attention mechanism layer calculates the attention weight distribution of the importance of the leading degree feature to the prediction of the current morphological difference at different time steps.

[0174] From the temporal offset features of acoustic and thermal energy transfer obtained in step S138, the leading degree feature value at each time position is extracted to form a time series. This time series is then input into the attention mechanism layer of the sequence prediction model. The attention mechanism layer uses the leading degree feature time series as the query and the temporal state vector sequence output by the Long Short-Term Memory network layer as the key and value. By calculating the similarity between the query and each key, the attention weight corresponding to each time step is obtained. The above attention weight distribution reflects the degree of influence of the leading degree feature on the prediction of the morphological difference at the current moment at different historical time points.

[0175] Step S195: Perform feature fusion between the temporal state vector and the leading degree feature after attention weighting to generate a comprehensive state representation vector containing leading temporal information.

[0176] The attention weights calculated in step S194 are weighted and summed with the time series of the leading degree features to obtain a weighted leading degree feature vector. This weighted leading degree feature vector is then concatenated with the time series state vector generated in step S193 to form a longer comprehensive state representation vector, which simultaneously contains the evolutionary law of morphological difference and leading time series information.

[0177] Step S196: Input the comprehensive state representation vector into the fully connected output layer of the sequence prediction model, and the fully connected output layer maps the comprehensive state representation vector into a morphological difference prediction sequence for a specified future prediction period. The morphological difference prediction sequence contains predicted morphological difference values ​​at multiple future time points.

[0178] The integrated state representation vector is fed into the fully connected output layer of the sequence prediction model. This fully connected output layer typically contains multiple neurons, and its output dimension is equal to the number of future time steps to be predicted. The fully connected output layer performs a linear transformation on the input integrated state representation vector (multiplying it by a weight matrix and adding a bias), generating a vector of the same length as the number of future time steps. Each element in this vector represents the predicted morphological dissimilarity value at the corresponding future time point. All these values ​​are arranged in chronological order to form the morphological dissimilarity prediction sequence.

[0179] Step S197: Identify the time points in the morphological difference prediction sequence where the predicted morphological difference value exceeds the preset wear warning threshold, and take the earliest time point that exceeds the warning threshold as the warning time point when the wear state level of the contact interface may deteriorate.

[0180] The morphological difference prediction sequence is iterated, and each predicted value is compared with a preset wear warning threshold. The time point corresponding to the first predicted value exceeding the threshold is recorded, and this time point is the warning time point. This warning time point indicates that if the current wear trend continues, the wear state level of the contact interface of the brake friction pair will deteriorate around this time point.

[0181] Step S198: Extract the rate of change sequence of morphological attribute features of the acoustic emission waveform morphology hysteresis segment within the historical time window from the hysteresis coupling path diagram, perform correlation analysis between the rate of change sequence and the morphological difference prediction sequence, and calculate the correlation coefficient between the rate of change of morphological attribute features and the morphological difference prediction value.

[0182] The morphological attributes (such as steepness and smoothness) of the acoustic emission waveform hysteresis segment within the most recent historical time window (e.g., the past 10 braking processes) are extracted, and their rate of change is calculated to form a rate of change sequence. The Pearson correlation coefficient between the above rate of change sequence and the morphological difference prediction sequence is calculated, and this coefficient is the correlation coefficient.

[0183] Step S199: Based on the magnitude of the correlation coefficient, adjust the confidence level of the predicted morphological difference value corresponding to the warning time point, generate a wear state evolution prediction result with confidence assessment, and use the prediction result as the decision basis for preventive maintenance of the brake.

[0184] The correlation coefficient reflects the correlation between historical data used for prediction and future prediction results. A high absolute value of the correlation coefficient (e.g., greater than a preset strong correlation threshold) indicates high reliability of the prediction and high accuracy of the warning time point. Conversely, a low absolute value of the correlation coefficient indicates low reliability of the prediction and potential deviation in the warning time point. Based on the correlation coefficient, a confidence level is assigned to the predicted morphological difference value corresponding to the warning time point. The final output is a comprehensive prediction result including the warning time point, the predicted morphological difference value, and its confidence level. This result can be used to guide preventative maintenance decisions for brakes.

[0185] For example, the method may further include: step S1100: obtaining multiple hysteresis coupling path graphs generated in multiple complete braking cycles to form a hysteresis coupling path graph training sample set.

[0186] For multiple complete braking cycles, steps S110 to S147 are executed respectively to generate a hysteresis coupling path diagram corresponding to each braking cycle. The above multiple hysteresis coupling path diagrams and their corresponding known contact interface wear state level labels (which can be obtained through other independent measurement methods, such as post-disassembly analysis) together constitute a training sample set.

[0187] Step S1101: Construct a wear state representation learning model based on self-attention mechanism and variational autoencoder. The wear state representation learning model includes an encoder, a latent space sampling layer and a decoder. The encoder is used to map the input hysteresis coupling path graph into a low-dimensional latent space feature distribution. The decoder is used to reconstruct the hysteresis coupling path graph from the latent space features.

[0188] Construct a deep learning model that employs a variational autoencoder architecture and incorporates a self-attention mechanism. The model consists of three parts: an encoder, a latent space sampling layer, and a decoder. The encoder aims to compress the input high-dimensional graph structure data (hysteresis-coupled path graph) into a low-dimensional latent space feature distribution (typically assumed to be a multidimensional Gaussian distribution). The decoder aims to reconstruct a hysteresis-coupled path graph as similar as possible to the input from the vectors sampled from this latent space feature distribution.

[0189] Step S1102: Input each hysteresis coupling path map in the training sample set of the hysteresis coupling path map into the graph neural network layer of the encoder. The graph neural network layer performs feature aggregation on the acoustic emission waveform morphology evolution path nodes and thermal radiation field distribution evolution path nodes in the hysteresis coupling path map to generate graph embedding features for each path node.

[0190] The first layer of the encoder is a graph neural network layer. For an input hysteresis-coupled path graph, the graph neural network layer utilizes the structural information of the graph (edge ​​connections between nodes) for message passing and feature aggregation. Specifically, for each node in the graph (whether it's a node representing an acoustic emission waveform or a node representing a thermal radiation field distribution), the graph neural network layer aggregates the features of its neighboring nodes (including nodes of another type connected by connecting lines and adjacent nodes on the same path), and generates new features for that node through a learnable linear transformation and activation function. After iterations through multiple layers of graph neural networks, the final feature representation of each node integrates its own attributes, neighborhood information, and global graph structural information; this feature is the graph embedding feature of that node.

[0191] Step S1103: Input the graph embedding features into the self-attention layer of the encoder. The self-attention layer calculates the global dependencies between different nodes in the path graph, generates a node feature sequence that enhances global context information, and inputs it into the pooling layer of the encoder to obtain a fixed-dimensional global feature vector of the entire hysteresis coupled path graph.

[0192] The graph embedding features of all nodes generated by the graph neural network layer are arranged into a node feature sequence according to the order of the nodes in the path graph. This sequence is then input into the self-attention layer in the encoder. The self-attention layer captures the global dependencies between nodes by calculating the correlation between any two node features in the sequence and generates a new feature representation for each node that incorporates global information. The node feature sequence enhanced by self-attention is then fed into a pooling layer (e.g., global average pooling or global max pooling). The pooling layer aggregates the features of all nodes to generate a fixed-length global feature vector for the entire hysteresis-coupled path graph.

[0193] Step S1104: Input the global feature vector into the latent space sampling layer of the wear state characterization learning model. The latent space sampling layer maps the global feature vector into a latent space mean vector and a latent space variance vector that follow a multidimensional Gaussian distribution. The latent space feature vector characterizing the hysteresis coupling path graph is sampled from the multidimensional Gaussian distribution.

[0194] The global feature vector is fed into the latent space sampling layer. This layer contains two independent linear transformation modules: one module maps the global feature vector to the latent space mean vector; the other module maps the global feature vector to the latent space variance vector. These mean and variance vectors define a multidimensional Gaussian distribution. The latent space sampling layer randomly samples a latent space feature vector representing the input hysteresis coupling path graph from this multidimensional Gaussian distribution. This sampling operation ensures the continuity of the latent space learned by the model, facilitating subsequent generation and interpolation operations.

[0195] Step S1105: Input the latent space feature vector into the feature reconstruction module of the decoder. The feature reconstruction module first reshapes the latent space feature vector into graph node initialization features with the same topology as the input hysteresis coupling path graph. Then, it transmits information between nodes through graph convolutional layers to gradually reconstruct the features of acoustic emission waveform morphology evolution path nodes and thermal radiation field distribution evolution path nodes, and finally reconstructs the complete hysteresis coupling path graph.

[0196] The decoder first inputs the latent space feature vector into a feature reconstruction module. This module maps the latent space feature vector into a vector with dimensions matching the total initial feature dimension of all nodes in the original graph through a fully connected layer. Then, this vector is reshaped according to the node order and node feature dimensions of the original graph to obtain the initial features of each node. Subsequently, these initial features are fed as input into multiple graph convolutional layers. Each graph convolutional layer transmits information between nodes according to the topology of the original graph, progressively updating the node feature representations. After iteration through multiple graph convolutional layers, the reconstructed features of each node are finally generated. By assembling the reconstructed features of all nodes according to the node order and connection relationships of the original graph, a complete hysteresis-coupled path graph can be reconstructed.

[0197] Step S1106: During model training, the parameters of the wear state representation learning model are optimized by minimizing the reconstruction error between the reconstructed hysteresis coupling path graph and the input hysteresis coupling path graph, and minimizing the KL divergence between the latent space feature distribution and the standard Gaussian distribution. After training, the corresponding known contact interface wear state level labels are extracted from the hysteresis coupling path graphs in the training sample set. The latent space feature vector corresponding to each graph is associated with its wear state level label to form a training set. A multilayer perceptron classifier is trained using the training set, and the multilayer perceptron classifier is used as the wear state level discrimination model.

[0198] The training objective of the wear state representation learning model consists of two parts. The first part is the reconstruction error, which aims to make the hysteresis coupling path map reconstructed by the decoder as close as possible to the original hysteresis coupling path input by the encoder. Figure 1 The difference between the two loss functions is typically measured using mean squared error or cross-entropy loss. The second part is KL divergence, which aims to make the latent space feature distribution (defined by mean and variance) learned by the latent space sampling layer as close as possible to a standard Gaussian distribution (mean = 0, variance = 1). The weighted sum of the two loss functions is used as the total loss for backpropagation to optimize the model parameters. After the model is trained, its encoder can map any input hysteresis coupling path graph into a well-distributed latent space feature vector. Subsequently, using all hysteresis coupling path graphs in the training sample set, the latent space feature vector corresponding to each graph is extracted by the trained encoder. The above latent space feature vectors are associated with their corresponding known wear state level labels to form a new training set. A multilayer perceptron classifier is trained using the above training set. This classifier can learn the mapping relationship from latent space feature vectors to wear state levels. The trained multilayer perceptron classifier is the wear state level discrimination model.

[0199] Step S1107: For a new braking process, the real-time hysteresis coupling path graph generated is input into the encoder of the trained wear state representation learning model to obtain its real-time latent space feature vector. The real-time latent space feature vector is then input into the wear state level discrimination model, which directly outputs the wear state level of the contact interface of the brake friction pair.

[0200] When wear condition monitoring is required for a new braking process, a real-time hysteresis coupling path diagram for that process is first generated using the aforementioned method. This real-time hysteresis coupling path diagram is then input into the encoder of a pre-trained wear condition representation learning model, which outputs a latent space feature vector. This latent space feature vector is then input into a similarly trained wear condition level discrimination model (i.e., a multilayer perceptron classifier). The wear condition level discrimination model performs classification calculations on this latent space feature vector and outputs a category label, which represents the wear condition level of the contact interface of the current brake friction pair.

[0201] In one exemplary embodiment, a brake operation status monitoring system is provided. This system can be a terminal, server, etc., and its internal structure diagram can be as follows: Figure 2 As shown, the brake operation status monitoring system includes a processor, memory, input / output interface, communication interface, display unit, and input device. The processor, memory, and input / output interface are connected via a system bus, and the communication interface, display unit, and input device are also connected to the system bus via the input / output interface. The processor provides computing and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The input / output interface is used for exchanging information between the processor and external devices. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, mobile cellular networks, near-field communication, or other technologies. When the computer program is executed by the processor, it implements a brake operation status monitoring method. The display unit is used to generate a visually visible image and can be a display screen, projection device, or virtual reality imaging device. The display screen can be an LCD screen or an e-ink screen. The input device can be a touch layer covering the display screen, or a button, trackball, or touchpad set on the housing of the brake operation status monitoring system, or an external keyboard, touchpad, or mouse, etc.

[0202] It should be noted that, in order to simplify the description of the present invention and thus help to understand one or more embodiments of the invention, multiple features may sometimes be grouped into one embodiment, drawing or description thereof in the foregoing description of the embodiments of the present invention.

Claims

1. A method for monitoring the operating status of a brake, characterized in that, The method includes: Acquire the timing sequence of acoustic emission waveforms generated by the contact interface of the friction pair and the timing sequence of infrared thermal radiation field images generated by the friction surface of the brake disc during the braking process; Based on the waveform morphology characteristics of the acoustic emission waveform in the time sequence and the thermal radiation field distribution morphology characteristics in the time sequence of the infrared thermal radiation field image, an acoustic-thermal energy transfer correlation path diagram is constructed. Identify the temporal offset relationship between the inflection point of acoustic emission waveform morphology change and the inflection point of thermal radiation field distribution morphology change in the acoustic-thermal energy transfer correlation path diagram to obtain the acoustic-thermal energy transfer temporal offset characteristics. The acoustic and thermal energy transfer timing offset characteristics are used to perform path shape correction processing on the acoustic and thermal energy transfer association path diagram to generate a hysteresis coupling path diagram that reflects the energy transfer hysteresis characteristics of the friction pair contact interface. The wear status level of the contact interface of the brake friction pair is determined based on the morphological difference between the hysteresis segment of the acoustic emission waveform and the leading segment of the thermal radiation field distribution in the hysteresis coupling path diagram.

2. The brake operating status monitoring method according to claim 1, characterized in that, The acquisition of the acoustic emission waveform timing sequence generated by the friction pair contact interface and the infrared thermal radiation field image timing sequence generated by the brake disc friction surface during braking execution includes: At the start of acoustic emission waveform acquisition, the acoustic emission sensing unit deployed on the force transmission path between the brake caliper and the friction lining starts acquiring acoustic emission waveforms. The waveform start point and waveform end point of each acoustic emission event are identified from the raw voltage waveform output by the acoustic emission sensing unit. The waveform segment between the waveform start point and waveform end point of each acoustic emission event is taken as an acoustic emission waveform unit. All acoustic emission waveform units are arranged in the time sequence of the waveform start point to form an initial sequence of acoustic emission waveform units. At the start of infrared thermal radiation imaging unit deployed in the normal direction of the brake disc friction working surface, the temperature field distribution array of the brake disc friction working surface in each frame of the original thermal image sequence output by the infrared thermal radiation imaging unit is extracted, and the temperature values ​​of each element point in the temperature field distribution array are arranged in the time order of the image frames to form the initial sequence of infrared thermal radiation field image unit. Extract the moment when the waveform amplitude reaches its maximum value from each acoustic emission waveform unit in the initial sequence of acoustic emission waveform units, and use this moment as the waveform feature anchor point of the acoustic emission waveform unit. Arrange the waveform feature anchor points of all acoustic emission waveform units in chronological order to form an acoustic emission waveform feature anchor point sequence. Extract the position of the element point where the temperature value reaches the maximum value in the temperature field distribution array from each frame image of the initial sequence of the infrared thermal radiation field image unit, and use the position of the element point as the thermal radiation field feature anchor point of the frame image. Arrange the thermal radiation field feature anchor points of all frames images in chronological order to form a thermal radiation field feature anchor point sequence. The acoustic emission waveform feature anchor point sequence and the thermal radiation field feature anchor point sequence are paired and associated with anchor points that have the same acquisition time sequence number. The paired and associated acoustic emission waveform unit and the infrared thermal radiation field image unit are combined to form a spatiotemporal association unit pair. All spatiotemporal association unit pairs are arranged in chronological order to form a set of synchronous association unit pairs between the acoustic emission waveform time sequence and the infrared thermal radiation field image time sequence. The waveform duration segment and waveform rise time segment of the acoustic emission waveform unit are extracted from each spatiotemporal association unit pair in the set of synchronous association unit pairs. The length relationship between the waveform duration segment and the waveform rise time segment is used as the acoustic emission waveform time segment feature of the spatiotemporal association unit pair. The acoustic emission waveform time segment features of all spatiotemporal association unit pairs are arranged in chronological order to form an acoustic emission waveform time segment feature sequence. From each spatiotemporal association unit pair in the set of synchronous association units, the relative positional relationship between the highest temperature array element point and the geometric center point of the temperature field distribution array of the infrared thermal radiation field image unit is extracted. This relative positional relationship is used as the thermal radiation field offset feature of the spatiotemporal association unit pair. The thermal radiation field offset features of all spatiotemporal association unit pairs are arranged in chronological order to form a thermal radiation field offset feature sequence. The acoustic emission waveform time period feature sequence and the thermal radiation field shift feature sequence are standardized respectively. Spatiotemporal correlation feature pairs are constructed based on each feature value in the standardized acoustic emission waveform time period feature sequence and the feature value of the corresponding time position in the standardized thermal radiation field shift feature sequence. All spatiotemporal correlation feature pairs are arranged in chronological order to form a spatiotemporal correlation feature pair sequence. The correlation degree between the acoustic emission waveform time period feature component and the thermal radiation field offset feature component of each spatiotemporal correlation feature pair in the spatiotemporal correlation feature pair sequence is identified to obtain the correlation tightness level of each spatiotemporal correlation feature pair. The correlation tightness level of all spatiotemporal correlation feature pairs is arranged in chronological order to form a correlation tightness level sequence. The correlation tightness level sequence is used as a sequence to represent the synchronization correlation degree between the acoustic emission waveform time sequence and the infrared thermal radiation field image time sequence.

3. The brake operating status monitoring method according to claim 1, characterized in that, The step of constructing an acoustic-thermal energy transfer correlation path map based on the waveform morphology characteristics of the acoustic emission waveform in the time sequence and the thermal radiation field distribution morphology characteristics in the time sequence of the infrared thermal radiation field image includes: Extract the rising waveform curve from the starting point to the peak point and the falling waveform curve from the peak point to the end point from each acoustic emission waveform time sequence unit. Use the rising waveform curve and the falling waveform curve together as the waveform morphology feature of the acoustic emission waveform unit. Arrange the waveform morphology features of all acoustic emission waveform units in time order to form a waveform morphology feature time sequence. Extract the radial distribution curve of temperature values ​​decreasing from the center region to the edge region of the temperature field distribution array from each frame of the infrared thermal radiation field image time sequence. Use the radial distribution curve as the thermal radiation field distribution feature of the frame image. Arrange the thermal radiation field distribution features of all frames in time sequence to form a thermal radiation field distribution feature time sequence. The waveform morphology features at each time position in the waveform morphology feature time series are correlated with the thermal radiation field distribution morphology features at the same time position in the thermal radiation field distribution morphology feature time series, forming a corresponding correlation pair between waveform morphology features and thermal radiation field distribution morphology features. All corresponding correlation pairs are arranged in chronological order to form a sequence of corresponding correlation pairs. Extract the steepness attribute of the rising curve and the smoothness attribute of the falling curve of the waveform morphology feature from each corresponding pair of the corresponding pair sequence. Use the steepness attribute and the smoothness attribute as the acoustic emission waveform morphology attribute set of the corresponding pair. Arrange the acoustic emission waveform morphology attribute sets of all corresponding pairs in chronological order to form an acoustic emission waveform morphology attribute sequence. Extract the radiation radius length attribute and radiation gradient change attribute of the radial distribution morphology curve of the thermal radiation field distribution morphology feature from each corresponding pair of the corresponding pair of the sequence. Use the radiation radius length attribute and radiation gradient change attribute as the thermal radiation field distribution morphology attribute set of the corresponding pair of the sequence ... Each attribute in the acoustic emission waveform morphology attribute sequence and each attribute in the thermal radiation field distribution morphology attribute sequence are standardized respectively. Attribute association pairs are constructed based on the acoustic emission waveform morphology attribute set in the standardized acoustic emission waveform morphology attribute sequence and the corresponding thermal radiation field distribution morphology attribute set in the standardized thermal radiation field distribution morphology attribute sequence. All attribute association pairs are arranged in chronological order to form an attribute association pair sequence. The degree of matching between the acoustic emission waveform morphology attribute set and the thermal radiation field distribution morphology attribute set of each attribute association pair in the attribute association pair sequence is identified to obtain the morphology matching degree level of each attribute association pair. The morphological change trend between each waveform morphological feature in the time series of the waveform morphological features and the waveform morphological features at adjacent time positions is taken as the waveform morphological evolution path, and all waveform morphological evolution paths are connected in chronological order to form an acoustic emission waveform morphological evolution path diagram. The distribution change trend between each thermal radiation field distribution morphology feature in the time series and its adjacent time position is taken as the thermal radiation field distribution evolution path, and all thermal radiation field distribution evolution paths are connected in chronological order to form a thermal radiation field distribution evolution path diagram. The acoustic emission waveform morphology evolution path diagram and the thermal radiation field distribution evolution path diagram are superimposed along the time axis, and the waveform morphology evolution path nodes at the same time position in the superimposed path diagram are connected with the thermal radiation field distribution evolution path nodes through the correlation line to form an acoustic and thermal energy transfer correlation path diagram.

4. The brake operating status monitoring method according to claim 1, characterized in that, The identification of the temporal offset relationship between the inflection points of acoustic emission waveform morphology changes and the inflection points of thermal radiation field distribution morphology changes in the acoustic-thermal energy transfer correlation path diagram, to obtain the temporal offset characteristics of acoustic-thermal energy transfer, includes: The acoustic emission waveform morphology evolution path diagram is extracted from the acoustic and thermal energy transfer correlation path diagram. The position where the rising morphology curve of the waveform changes from steep to gentle is identified in the acoustic emission waveform morphology evolution path diagram. This morphology change position is taken as the first type of turning point of the acoustic emission waveform morphology, and the time position of the first type of turning point is recorded. The acoustic emission waveform morphology evolution path diagram is extracted from the acoustic heat energy transfer correlation path diagram. The position where the waveform morphology descends from a flat to a steep change is identified in the acoustic emission waveform morphology evolution path diagram. This position is taken as the second type of turning point of the acoustic emission waveform morphology, and the time position of the second type of turning point is recorded. Arrange the time positions of the first type of inflection point of the acoustic emission waveform morphology and the time positions of the second type of inflection point of the acoustic emission waveform morphology in chronological order to form a time position sequence of inflection points of acoustic emission waveform morphology. Extract the thermal radiation field distribution evolution path map from the acoustic and thermal energy transfer correlation path map, identify the distribution change position where the radiation radius length of the radial distribution morphology curve changes from growth to stagnation in the thermal radiation field distribution evolution path map, take the distribution change position as the first type of turning point of thermal radiation field distribution morphology, and record the time position of the first type of turning point. Extract the thermal radiation field distribution evolution path map from the acoustic and thermal energy transfer correlation path map, identify the distribution change position where the radiation gradient of the radial distribution morphology curve changes from increasing to decreasing in the thermal radiation field distribution evolution path map, take the distribution change position as the second type of turning point of thermal radiation field distribution morphology, and record the time position of the second type of turning point. Arrange the time positions of the first type of inflection point of the thermal radiation field distribution pattern and the time positions of the second type of inflection point of the thermal radiation field distribution pattern in chronological order to form a time position sequence of inflection points of thermal radiation field distribution pattern. The time position of each inflection point in the acoustic emission waveform morphology inflection point time position sequence is compared with the time position of the inflection point with the same inflection point number in the thermal radiation field distribution morphology inflection point time position sequence to obtain the time sequence relationship between each corresponding inflection point pair. The time sequence relationship between all corresponding inflection point pairs is arranged in the order of the inflection point number to form a time sequence relationship. Based on the ratio of the number of inflection point pairs in the time sequence where the inflection point of the acoustic emission waveform morphology precedes the inflection point of the thermal radiation field distribution morphology to the total number of inflection point pairs in the time sequence, the degree of precedence of the acoustic emission waveform morphology change over the thermal radiation field distribution morphology change is determined. Similarly, the degree of precedence of the thermal radiation field distribution morphology change over the acoustic emission waveform morphology change is determined by the ratio of the number of inflection point pairs where the inflection point of the thermal radiation field distribution morphology precedes the inflection point of the acoustic emission waveform morphology to the total number of inflection point pairs in the time sequence. By comparing the degree of the lead between the acoustic emission waveform morphology change and the thermal radiation field distribution morphology change with the degree of the lead between the thermal radiation field distribution morphology change and the acoustic emission waveform morphology change, the acoustic-thermal energy transfer time sequence offset feature is obtained.

5. The brake operating status monitoring method according to claim 1, characterized in that, The step of using the time-series offset characteristics of acoustic and thermal energy transfer to perform path shape correction processing on the acoustic and thermal energy transfer correlation path diagram to generate a hysteresis coupling path diagram reflecting the energy transfer hysteresis characteristics of the friction pair contact interface includes: The first correction direction parameter is determined based on the degree of precedence of the acoustic emission waveform morphology change over the thermal radiation field distribution morphology change in the acoustic-thermal energy transfer time-series offset characteristics. The first correction direction parameter is used to indicate the path segment that needs to be extended in the positive direction of the time axis in the acoustic emission waveform morphology evolution path diagram. The second correction direction parameter is determined based on the degree of precedence of the change in the distribution shape of the thermal radiation field over the change in the shape of the acoustic emission waveform in the time-series offset characteristics of the acoustic and thermal energy transfer. The second correction direction parameter is used to indicate the path segment that needs to be extended in the positive direction of the time axis in the thermal radiation field distribution evolution path diagram. Based on the first correction direction parameter, determine the location region of the path segment that needs to be extended in the acoustic emission waveform morphology evolution path diagram, and extend the acoustic emission waveform morphology evolution path in the location region according to the direction indicated by the first correction direction parameter to obtain the acoustic emission waveform morphology evolution path diagram after extension and correction. The location region of the path segment that needs to be extended in the thermal radiation field distribution evolution path map is determined according to the second correction direction parameter. The thermal radiation field distribution evolution path in the location region is extended in the direction indicated by the second correction direction parameter to obtain the thermal radiation field distribution evolution path map after extension and correction. The acoustic emission waveform morphology evolution path diagram after extension and correction is overlaid with the thermal radiation field distribution evolution path diagram after extension and correction on the time axis. This aligns the inflection points with the same inflection point number in the two evolution path diagrams in time position, resulting in a preliminary aligned acoustic and thermal energy transfer correlation path diagram. The degree of morphological matching between the acoustic emission waveform morphology evolution path and the thermal radiation field distribution evolution path at the same time position is then identified. Path nodes with a morphological matching degree lower than a preset matching threshold are identified as node positions that need further correction, forming a set of node positions to be corrected. Based on the difference direction between the acoustic emission waveform morphology attribute and the thermal radiation field distribution morphology attribute at each node position in the set of node positions to be corrected, the local correction direction parameter at each node position to be corrected is determined. According to the local correction direction parameter at each node position to be corrected, the local path morphology evolution path of the acoustic emission waveform morphology or the thermal radiation field distribution evolution path at the corresponding node position in the initially aligned acoustic and thermal energy transfer association path diagram is fine-tuned, so that the matching degree between the acoustic emission waveform morphology and the thermal radiation field distribution morphology at the node position is improved after fine-tuning. The distribution of the degree of shape matching between the acoustic emission waveform evolution path and the thermal radiation field distribution evolution path in the acoustic-thermal energy transfer correlation path diagram after local path shape fine-tuning is used as the hysteresis degree distribution feature. The hysteresis degree distribution feature is then embedded into the acoustic-thermal energy transfer correlation path diagram after local path shape fine-tuning, and a hysteresis coupling path diagram reflecting the energy transfer hysteresis characteristics of the friction pair contact interface is output.

6. The brake operating status monitoring method according to claim 1, characterized in that, The determination of the wear state level of the contact interface of the brake friction pair based on the morphological difference between the hysteresis segment of the acoustic emission waveform and the leading segment of the thermal radiation field distribution in the hysteresis coupling path diagram includes: Extract the path segments in the acoustic emission waveform morphology evolution path from the hysteresis coupling path diagram where the hysteresis degree distribution characteristic value exceeds the preset hysteresis threshold. Take the path segment as the acoustic emission waveform morphology hysteresis segment and record the start and end positions of the hysteresis segment in the acoustic emission waveform morphology evolution path. Extract path segments from the hysteresis coupling path diagram where the hysteresis degree distribution characteristic value is lower than the preset hysteresis threshold in the thermal radiation field distribution evolution path. Take these path segments as the leading segments of the thermal radiation field distribution morphology and record the starting and ending positions of these leading segments in the thermal radiation field distribution evolution path. The starting position of the acoustic emission waveform hysteresis segment is compared with the starting position of the thermal radiation field distribution leading segment to obtain the first relative offset direction of the starting position of the acoustic emission waveform hysteresis segment relative to the starting position of the thermal radiation field distribution leading segment. The position of the hysteresis segment of the acoustic emission waveform is compared with the position of the leading segment of the thermal radiation field distribution to obtain the second relative offset direction of the hysteresis segment of the acoustic emission waveform relative to the position of the leading segment of the thermal radiation field distribution. The relative positional relationship between the acoustic emission waveform morphology hysteresis segment and the thermal radiation field distribution morphology leading segment in the hysteresis coupling path diagram is determined based on the first relative offset direction and the second relative offset direction, and the relative positional relationship is quantified as the first morphological difference characteristic value. The average steepness of the rising curve and the average smoothness of the falling curve of the acoustic emission waveform within the hysteresis coupling path diagram are extracted from the acoustic emission waveform morphology hysteresis segment. The average steepness and average smoothness are used as the morphological attribute features of the acoustic emission waveform morphology hysteresis segment. The average growth rate of the radiation radius and the average change magnitude of the radiation gradient are extracted from the hysteresis coupling path diagram and used as the morphological attribute features of the leading segment of the thermal radiation field distribution. The morphological attribute features of the acoustic emission waveform hysteresis segment and the morphological attribute features of the thermal radiation field distribution leading segment are standardized respectively. The attribute difference degree of the standardized morphological attribute features of the acoustic emission waveform hysteresis segment and the standardized morphological attribute features of the thermal radiation field distribution leading segment is calculated to obtain a second morphological difference degree feature value. According to the preset weight, the first morphological difference degree feature value and the second morphological difference degree feature value are weighted and fused to obtain the comprehensive morphological difference degree between the acoustic emission waveform hysteresis segment and the thermal radiation field distribution leading segment. The morphological difference is compared with the morphological difference range corresponding to multiple preset wear state levels to determine the wear state level range to which the morphological difference belongs, and the wear state level corresponding to the wear state level range is used as the wear state level of the contact interface of the brake friction pair.

7. The brake operating status monitoring method according to claim 1, characterized in that, The method of determining the wear state level of the contact interface of the brake friction pair based on the morphological difference between the hysteresis segment of the acoustic emission waveform and the leading segment of the thermal radiation field distribution in the hysteresis coupling path diagram further includes: The transition region morphological features between the rising and falling curves of the acoustic emission waveform morphology within the hysteresis coupling path diagram are extracted, and these transition region morphological features are used as the transition region features of the hysteresis segment of the acoustic emission waveform morphology. The morphological features of the connecting region between the radiation radius growth region and the radiation gradient change region in the radial distribution curve of the thermal radiation field distribution pattern in the leading segment of the hysteresis coupling path diagram are extracted, and the morphological features of the connecting region are taken as the morphological features of the connecting region of the leading segment of the thermal radiation field distribution pattern. The morphological difference between the transition region features of the hysteresis segment of the acoustic emission waveform and the connection region features of the leading segment of the thermal radiation field distribution is calculated to obtain the third morphological difference feature value. The ratio of the duration of the hysteresis segment of the acoustic emission waveform on the time axis to the duration of the leading segment of the thermal radiation field distribution on the time axis is extracted from the hysteresis coupling path diagram, and this ratio is used as the characteristic value of the time length ratio between the hysteresis segment and the leading segment. The ratio of the drastic change in the acoustic emission waveform morphology within the hysteresis segment to the drastic change in the thermal radiation field distribution morphology within the leading segment is extracted from the hysteresis coupling path diagram. This ratio is used as a comparative feature value for the drastic change in the morphology of the hysteresis segment and the leading segment. The third morphological difference feature value, the time length ratio feature value, and the morphological change intensity comparison feature value are standardized. The standardized third morphological difference feature value, time length ratio feature value, and morphological change intensity comparison feature value are combined to construct the morphological difference vector of the lag segment and the leading segment. This morphological difference vector is used as a refined representation of the comprehensive morphological difference. The morphological difference vector is compared with the morphological difference vector reference range corresponding to multiple preset wear state levels in vector space position to determine the vector space region to which the morphological difference vector belongs, and the wear state level corresponding to the vector space region is taken as the first intermediate wear state level. The spatial distribution pattern of the hysteresis degree distribution feature value in the acoustic emission waveform morphology evolution path is extracted from the hysteresis coupling path diagram, and this spatial distribution pattern is used as the feature of the acoustic emission waveform morphology hysteresis degree distribution pattern. The spatial distribution pattern of the hysteresis degree distribution feature value in the thermal radiation field distribution evolution path is extracted from the hysteresis coupling path diagram, and this spatial distribution pattern is used as the feature of the hysteresis degree distribution pattern of the thermal radiation field distribution morphology. The hysteresis distribution pattern of the acoustic emission waveform and the hysteresis distribution pattern of the thermal radiation field are used to identify the pattern matching degree to obtain the pattern matching degree level. The first intermediate wear state level is then corrected based on the pattern matching degree level to obtain the wear state level of the contact interface of the brake friction pair.

8. The brake operating status monitoring method according to claim 1, characterized in that, The method further includes: The evolution trend of the morphological difference between the hysteresis segment of acoustic emission waveform and the leading segment of thermal radiation field distribution on the time axis is extracted from the hysteresis coupling path diagram, and this evolution trend is used as the time evolution curve of morphological difference. Identify the time point in the morphological difference time evolution curve where the morphological difference value changes abruptly, take this time point as the morphological difference abrupt change point, and record the direction of change of morphological difference before and after the abrupt change point. The direction of change of wear state at the contact interface of the brake friction pair is determined based on the direction of change of morphological difference before and after the abrupt change point of morphological difference, and this direction of change is used as the characteristic of wear state change direction. The ratio of the distribution density of the hysteresis segment of the acoustic emission waveform on the time axis to the distribution density of the leading segment of the thermal radiation field on the time axis is extracted from the hysteresis coupling path diagram, and this ratio is used as the distribution density ratio feature of the hysteresis segment and the leading segment. The morphological feature repetition pattern matching degree between the acoustic emission waveform morphology in the hysteresis segment and the thermal radiation field distribution morphology in the leading segment is extracted from the hysteresis coupling path diagram, and this repetition pattern matching degree is used as the morphological feature repetition pattern matching degree feature. Based on the wear state change direction characteristics, the distribution density ratio characteristics, and the morphological feature repetition pattern matching degree characteristics, a wear state dynamic evolution feature vector is constructed; The wear state dynamic evolution feature vector is compared with the reference range of wear state dynamic evolution feature vectors corresponding to multiple preset wear state evolution stages in vector space position to determine the vector space region to which the wear state dynamic evolution feature vector belongs, and the wear state evolution stage corresponding to the vector space region is taken as the wear state evolution stage of the contact interface of the brake friction pair. The expected trend of the wear state level of the contact interface of the brake friction pair is determined based on the wear state evolution stage, and this expected trend is used as auxiliary judgment information for brake operation status monitoring.

9. The brake operating status monitoring method according to claim 1, characterized in that, The method further includes: Extract the curve showing the change in the degree of morphological matching between the acoustic emission waveform morphology evolution path and the thermal radiation field distribution evolution path over time from the acoustic-thermal energy transfer correlation path diagram; Identify the transition point in the matching degree change curve where the matching degree value changes from a high matching state to a low matching state, take this transition point as the starting point of energy transfer path distortion, and record the time position of the distortion starting point. The morphological matching trend between the acoustic emission waveform morphology evolution path and the thermal radiation field distribution evolution path in the time region after the distortion initiation point is extracted from the acoustic-thermal energy transfer correlation path diagram, and this recovery trend is used as the self-recovery capability feature of the energy transfer path. Based on the time position of the energy transfer path distortion initiation point and the self-recovery capability characteristics of the energy transfer path, the distortion sensitivity level of the energy transfer path at the contact interface of the brake friction pair is determined. The variation law of the morphological difference between the hysteresis segment of acoustic emission waveform and the leading segment of thermal radiation field distribution morphology in multiple braking processes is extracted from the hysteresis coupling path diagram, and this variation law is used as the feature of wear accumulation in multiple braking processes. Based on the distortion sensitivity level and the cumulative wear characteristics of multiple braking cycles, the cumulative change trend of the wear state level of the contact interface of the brake friction pair is determined, and this cumulative change trend is used as the basis for predicting the remaining service life of the brake.

10. A brake operating status monitoring system, characterized in that, include: processor; A machine-readable storage medium for storing machine-executable instructions of the processor; The processor is configured to execute the brake operating status monitoring method according to any one of claims 1 to 9 by executing the machine-executable instructions.

Images