Intelligent health assessment analysis method and system based on health management

Abstract

The invention relates to the technical field of intelligent health assessment, in particular to an intelligent health assessment analysis method and system based on health management, and the method comprises the following steps: obtaining a direction conflict path in a monitoring period, tracking an offset change, positioning a synchronous bifurcation position, extracting a trend switching section and a rhythm repetition path, extending a fluctuation range, and dividing a bifurcation structure. And identifying trend starting and ending and tracking a direction trend to obtain a health state grading judgment label group. In the invention, through identifying direction conflicts and rhythm fluctuation changes in an index path, constructing a continuous track structure and extracting a trend differentiation paragraph, and combining a path continuation sequence and fluctuation rhythm distribution, rhythm tracking of state transaction and trend turning is completed. A label sequence with direction perception and rhythm identification features is generated under multi-time sequence change, the response ability of health state division to trend evolution is enhanced, and the capture ability of evaluation to potential state change and the adaptability of classification and layering are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent health assessment, in particular to an intelligent health assessment analysis method and system based on health management. BACKGROUND

[0002] The technical field of intelligent health assessment involves methods for quantitatively analyzing and determining the health status of individuals, including multi-source health data collection, standardized processing, feature extraction, health index modeling, and evaluation rule formulation. This field widely integrates artificial intelligence, physiological parameter detection, and risk prediction models to provide data-driven health status assessment for users, including health behavior recognition, physiological trend monitoring, chronic disease risk prediction, health classification, and evaluation index reasoning. Traditional intelligent health assessment analysis methods based on health management involve constructing an index system based on user basic information and periodic detection data, classifying and determining health status by limiting index thresholds, and using structured questionnaire scoring, rule matching, and statistical threshold determination. These methods rely on manual input and pre-set rules to map and classify health levels based on physical examination items and subjective health questionnaire content from national basic health management standards.

[0003] The structured questionnaire and periodic detection index threshold classification method lacks processing capacity when facing continuous change paths, and cannot track the trend divergence or rhythm repetition characteristics generated by the trend of the index in time evolution. When the index change does not constitute a threshold violation but implies a fluctuation rhythm and turning trend, the original health level remains unchanged, masking potential state changes. Especially in the initial stage of frequent fluctuations or trend changes in individual health status, the static rule processing mode lacks support for trend analysis and fluctuation rhythm, limiting the continuity of state evolution chain recognition and leading to health determination lag and risk signal masking. SUMMARY

[0004] To solve the technical problems existing in the prior art, the embodiments of the present application provide an intelligent health assessment analysis method based on health management, comprising the following steps:

[0005] To achieve the above-mentioned purpose, the present application adopts the following technical scheme: an intelligent health assessment analysis method based on health management, comprising the following steps:

[0006] S1: Obtain the index group of the fluctuation marked section in the health monitoring period, extract the path with inconsistent direction in time, track the offset change of the continuous section, locate the position with direction divergence and synchronous change, and obtain the health index conflict section set;

[0007] S2: identifying direction changeover points along a time sequence based on the path in the health index conflict section set, dividing the path by location, extracting a path with consistent direction, and obtaining a state change candidate trend section;

[0008] S3: reading up and down fluctuation directions of each section based on the continuous path in the state change candidate trend section, analyzing fluctuations in time sequence, identifying rhythm repeating paths, extending fluctuation ranges, and obtaining a health evaluation main channel structure section;

[0009] S4: extracting trend turning points based on the health evaluation main channel structure section, identifying trends with inconsistent directions, dividing path structures according to continuous regions separated according to trend direction repetition, and obtaining a health state double-track bifurcation section group;

[0010] S5: identifying dominant trend starting, changing and ending positions based on the path trend in the health state double-track bifurcation section group, tracking direction trends along a time sequence, analyzing trend continuation modes, and obtaining a health state grading judgment label group.

[0011] As a further scheme of the present application, the health index conflict section set includes a path direction inconsistency section, a change trend overlap section, and a conflict overlap interval, the state change candidate trend section includes a direction consistent path section, a trend turning position, and a path splitting segment, the health evaluation main channel structure section includes a rhythm repeating section, a fluctuation sequence mode, and a fluctuation extension segment, the health state double-track bifurcation section group includes a trend divergence track section, a track direction difference point, and a trend isolation section, and the health state grading judgment label group includes a dominant trend starting point, a trend continuation feature, and a trend termination marker.

[0012] As a further scheme of the present application, the direction changeover point refers to a critical position point where an increase and decrease change relationship of index values at adjacent time positions is identified in trend section extraction, and the direction changeover point is extracted from an upturn to a downturn and a downturn to an upturn.

[0013] The extended fluctuation range refers to extending an index path section with the same fluctuation rhythm sequence along a time axis, analyzing a synchronous section with continuous rhythm matching, and connecting the direction consistency of up and down nodes in a rhythm feature screening process.

[0014] As a further scheme of the present application, the trend turning point refers to a space-time node where a direction trend changes is extracted in a main channel structure section, and the trend turning point refers to a reverse node where a trend upturn turns to a downturn and a downturn turns to an upturn.

[0015] The trend continuation mode refers to analyzing the continuity and change trend of a trend extension path according to the trend direction relationship and turning and convergence performance of adjacent sections in a health state label.

[0016] As a further scheme of the present application, the specific steps of S1 are:

[0017] S101: Obtain the index group of fluctuation marked sections in the health monitoring period, read the trend direction of the index path in the same time slice, locate the position where the direction of the paths is inconsistent, divide the path set with direction difference according to the time slice, and obtain the direction difference path set;

[0018] S102: Based on the path data in the direction difference path set, read the offset behavior in the continuous section in time sequence, the offset direction of the corresponding path at the adjacent time position, identify the path interval with synchronous direction change in the same time range, and obtain the synchronous offset interval set;

[0019] S103: Based on the path trajectory in the synchronous offset interval set, track the path range with direction difference and adjacent change time at the continuous time position, filter the time position with inconsistent direction and synchronous offset, and obtain the health index conflict section set.

[0020] As a further scheme of the present application, the specific steps of S2 are:

[0021] S201: Based on the index path in the health index conflict section set, extract the direction trend of the path at the continuous position in time sequence, judge the direction change position according to the increase and decrease state of the adjacent points, induce the direction conversion node according to the trend characteristics of the difference direction switching, and obtain the direction change position sequence;

[0022] S202: Based on the direction change position sequence, locate the path section before and after the direction switching, identify the continuous index points in the path segment respectively, compare the starting and ending index values between the adjacent segments in turn, calculate the index offset before and after the direction conversion, and obtain the conversion trend offset index group;

[0023] S203: Based on the conversion trend offset index group, filter the path segment with consistent index trend, locate the trajectory range with unchanged direction according to the time direction, track the continuous trend in the path segment, and obtain the state change candidate trend segment.

[0024] As a further scheme of the present application, the specific steps of S3 are:

[0025] S301: Based on the continuous section in the state change candidate trend segment, read the up and down fluctuation direction of the index path in each section, follow the direction trend change position in time sequence, sequentially connect the fluctuation order, and obtain the fluctuation direction sequence group;

[0026] S302: Based on the fluctuation direction sequence group, the fluctuation paragraph is divided according to the continuity of the in-path direction trend, the rhythm sequence of direction retention and transition is identified, the structure ratio of path segment rhythm continuity and direction continuation strength is calculated, and the rhythm feature matching sequence set is obtained.

[0027] S303: Based on the rhythm feature matching sequence set, the trajectory segment of the same rhythm sequence is extended along the time line, the upper and lower nodes are connected in the form of fluctuation direction synchronous continuation, and the health evaluation main channel structure paragraph is obtained.

[0028] As a further scheme of the present application, the specific steps of S4 are:

[0029] S401: Based on the index path in the health evaluation main channel structure paragraph, the trend fluctuation form is tracked along the path time extension direction, the positions of continuous change direction in the time sequence are compared, and the trend turning point position group is obtained.

[0030] S402: Based on the trend turning point position group, the trend direction between adjacent turning points is tracked along the trajectory paragraph, the path segment direction with inconsistent direction before and after is identified, and the direction divergence trajectory interval group is obtained.

[0031] S403: Based on the path trend difference in the direction divergence trajectory interval group, the trend extension of the trend continuation line is analyzed at the direction change alternation, the interval segment with direction difference and not maintaining the same direction is extracted along the path trend, and the health state double rail bifurcation section group is obtained.

[0032] As a further scheme of the present application, the specific steps of S5 are:

[0033] S501: Based on the time sequence of the path segment in the health state double rail bifurcation section group, the node of the path starting point and the first direction reversal is identified, and the direction convergence segment in the tail trend is connected, to obtain the trend stage boundary sequence group.

[0034] S502: Based on the continuous trend of path direction between adjacent nodes in the trend stage boundary sequence group, the trend extension segment is tracked along the time direction, the trend trend structure with no direction turning is identified, and the trend extension interval set is obtained.

[0035] S503: Based on the trend direction distribution state of the path segment in the trend extension interval set, the region with consistent direction is screened, the trend trend information is sequentially associated, and the health state grading judgment label group is obtained.

[0036] The intelligent health evaluation analysis system based on health management comprises:

[0037] The conflict segment identification module obtains the indicator group of the fluctuating marked segment within the health monitoring cycle, identifies the indicator path with different directional trends within the same time slice, tracks the continuous segment offset changes along the path, locates segments with inconsistent directions and overlapping changes, and obtains the set of conflict segments of health indicators.

[0038] The trend segment extraction module, based on the indicator path of the health indicator conflict segment set, follows the trend change along the time series, locates the position where the upward and downward directions switch, splits the path according to the switching point, and filters out the path with a continuous consistent direction to obtain the candidate trend segment of state change.

[0039] The rhythm path filtering module reads the upward and downward fluctuation direction, analyzes the fluctuation sequence, and determines whether rhythm repetition occurs based on the continuous segments in the candidate trend segments of the state change. It extends the fluctuation interval in the repeated rhythm path to obtain the main channel structure segment of health assessment.

[0040] Based on the time extension relationship of the indicator path in the main channel structure segment of the health assessment, the track path decomposition module extracts the trend turning point, identifies the trajectory segment with inconsistent directional changes, splits the path according to the divergence point, decomposes the directional repetition separation segment, and obtains the health status double track bifurcation section group.

[0041] The label path tracking module identifies the start, transition, and end positions of the trend based on the trend sequence of the path segments in the dual-track bifurcation section group of the health status, tracks the continuation of the trend direction, and obtains the health status classification label group.

[0042] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0043] In this invention, by identifying directional conflicts and rhythm fluctuations in the indicator path, a continuous trajectory structure is constructed and trend differentiation segments are extracted. By combining the path continuation sequence and the distribution of fluctuation rhythm, the rhythm tracking of state anomalies and trend reversals is completed. Under multiple temporal changes, a label sequence with directional perception and rhythm recognition features is generated, which enhances the responsiveness of health state classification to trend evolution and improves the ability of assessment to capture potential state changes and the adaptability of classification and stratification. Attached Figure Description

[0044] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0045] Figure 1 This is a schematic diagram of the steps of the present invention;

[0046] Figure 2 This is a detailed schematic diagram of S1 of the present invention;

[0047] Figure 3 This is a detailed schematic diagram of S2 of the present invention;

[0048] Figure 4 This is a detailed schematic diagram of S3 of the present invention;

[0049] Figure 5 This is a detailed schematic diagram of S4 of the present invention;

[0050] Figure 6 This is a detailed schematic diagram of S5 of the present invention;

[0051] Figure 7 This is a system module diagram of the present invention. Detailed Implementation

[0052] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0053] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0054] In the embodiments of this invention, the terms "image" and "picture" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, their intended meanings are consistent. Similarly, the terms "of," "corresponding (relevant)," and "corresponding" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, their intended meanings are consistent.

[0055] In this embodiment of the invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0056] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0057] Please see Figure 1 This invention provides an intelligent health assessment and analysis method based on health management, comprising the following steps:

[0058] S1: Obtain the indicator group of the fluctuation marker segment within the health monitoring period, extract the indicator path with different directional trends within the same time slice, track the offset changes in continuous segments along the path, locate the segments with inconsistent directions and overlapping changes between paths, and obtain the set of conflict segments of health indicators.

[0059] S2: Based on the indicator path of the health indicator conflict zone, follow the path trend along the time series to identify the process of directional change, locate the position where the direction changes from rising to falling and from falling to rising, split the path segment according to the direction switching point, identify the path where the direction remains consistent, and obtain the candidate trend segment of state change.

[0060] S3: Based on the continuous segments in the candidate trend segments of state change, read the up and down fluctuation direction of the indicator path in each segment, analyze the fluctuation sequence in each segment along the time direction, determine whether there is a phenomenon of repeated fluctuation rhythm, extend the fluctuation range in the path segment where the fluctuation rhythm is continuously repeated, and obtain the main channel structure segment of health assessment.

[0061] S4: Based on the time extension relationship of the indicator path in the main channel structure of health assessment, extract the adjacent trend turning points in the track, identify the trajectory segments with inconsistent trend change directions, segment the path according to the trajectory divergence points, split the segments in the path with repeated trend direction separation, and obtain the health status double track bifurcation section group.

[0062] S5: Based on the trend sequence of path segments in the dual-track bifurcation section group of health status, identify the starting position, direction change position and ending position of the dominant trend, track the trend change process along the time sequence, and obtain the health status classification label group by corresponding to the continuation of the trend direction in the path segment.

[0063] The set of conflicting health indicators includes segments with inconsistent path directions, segments with overlapping trends, and conflicting intervals. Candidate trend segments for state changes include segments with consistently consistent directions, trend turning points, and path splitting segments. The main channel structure segments for health assessment include segments with repetitive rhythms, fluctuating sequence patterns, and undulating extension segments. The dual-track bifurcation segment group for health status includes segments with diverging trend trajectories, points with differences in trajectory directions, and trend isolation segments. The health status grading and judgment label group includes the starting point of the dominant trend, trend continuation characteristics, and trend termination markers.

[0064] Please see Figure 2 The specific steps of S1 are as follows:

[0065] S101: Obtain the indicator group of the fluctuation marker segment within the health monitoring period, read the trend direction of the indicator path within the same time slice, locate the position where the direction of the path is inconsistent, delineate the path set with direction difference according to the time slice, and obtain the path set with direction difference.

[0066] Firstly, based on the index sampling sequence recorded by each observation point within a preset monitoring period, paths can be divided according to the index type. For example, separate index path sets can be constructed for different monitoring dimensions such as machine vibration intensity, temperature fluctuation, and load stress. Time periods with fluctuation attributes can be marked for each. During this process, it is necessary to determine whether fluctuations exist based on the magnitude of the change in node values ​​within each index path. If the difference in values ​​between adjacent nodes in a continuous node segment exceeds the fluctuation judgment benchmark value corresponding to the path, then that segment is determined to be a fluctuation-marked segment. For example, if the fluctuation judgment benchmark for the temperature path is 1.8℃, and there is a fluctuation between consecutive sampling points within a certain path segment with a difference between ±2.1℃, then that path segment is marked as a fluctuation segment. Based on this, further readings are performed. The direction of change of each indicator path at the same sampling point at the same time can be determined by the positive and negative changes between adjacent sampling point values. For example, when the value change of sampling point x and x+1 on the pressure path is positive, the direction is marked as "up"; otherwise, it is marked as "down". The direction marks of all indicator paths at the same time are aggregated to form the direction distribution set under that time slice. The location of inconsistent direction is identified in this way, that is, the intersection point of different paths showing positive and negative changes at a certain sampling time. Further, the location of the above-mentioned direction difference is continuously tracked according to the time slice. The sets of multiple paths with conflicting directions are merged to obtain the set of all path segments with inconsistent directions in the same time slice within the continuous segment. Finally, the set of paths with different directions is obtained.

[0067] S102: Based on the path data in the set of directional difference paths, read the offset behavior in continuous segments in chronological order, identify the offset direction of the corresponding path at adjacent time positions, identify the path intervals with synchronized directional changes within the same time range, and obtain the set of synchronized offset intervals.

[0068] First, the data sequence of each path at consecutive time points is read, and the direction of numerical change of adjacent nodes is identified. For each path, a difference judgment operation is performed between any two consecutive sampling points. If the value of the later point is higher than that of the earlier point, the corresponding offset direction is "positive", and vice versa. For example, if the values ​​of a certain path at adjacent time points t1 and t2 are 4.2 and 3.7, the offset direction is negative. In this way, the offset direction sequence group of the entire path in the entire segment is generated. Then, the path sequence is paired with the offset sequences of other paths at corresponding time points, and a synchronous scanning operation is performed on each time slice on the timeline. In each time slice, it is determined whether there are multiple paths with the same direction mark at that time point. If three out of the five paths at time t3 are "positive", the offset direction is determined. The offset at that time point is recorded as synchronous. Based on this, these time points with consistent directions are connected to form a synchronous direction segment. To improve the stability of the judgment, a path direction coincidence judgment threshold is used in the path comparison process, set to 60%. That is, if there are 5 paths participating in the judgment in a certain interval, and 3 of them have consistent directions in the whole interval, the interval can be judged as a synchronous offset segment. For example, in the interval from t1 to t10, if the directions of the 1st, 3rd and 5th paths are all "negative", and the others are inconsistent, then the segment meets the synchronous offset judgment requirements. Then, all path segments that meet the direction coincidence threshold are grouped to complete the delineation operation of each synchronous interval segment. These delineation results are organized into a set form according to the time axis order, and finally the synchronous offset interval set is obtained.

[0069] S103: Based on the path trajectory in the set of synchronous offset intervals, track the path range with different directions and adjacent change times at continuous time positions, filter the time positions of synchronous offset with inconsistent directions, and obtain the set of conflict segments of health indicators.

[0070] First, the trend direction of each path needs to be segmented and read in continuous time intervals. Combined with the trend performance of adjacent paths within the same time interval, a directional correlation between paths is established to determine if directional differences exist. During execution, the direction marker sequence of each path within the synchronous offset segment is read first. In the corresponding example, if the direction of path P1 is "negative-negative-negative-positive" in continuous time slices t1 to t5, and the direction of path P2 is "positive-positive-negative-negative", then a directional difference occurs between t3 and t5. Next, the direction markers of other paths within this time period are selected, and the paths are compared to see if inconsistent directional behavior occurs simultaneously. For example, if path P3 also shows "positive-positive-positive" during t3 to t5, it indicates that three paths have mutually divergent directions within this segment. In this case, this segment needs to be marked as a directional difference interval. For all synchronous offset time segments in all path sets, directional consistency assessment is performed. The path comparison method is used to make logical judgments on the direction of each pair of paths at the same time point. If there is an inequality, the time point is recorded as a conflict point. Then, the distribution of these conflict points on the time line is aggregated, and the location segments with inconsistent path directions at multiple consecutive time points are selected as the judgment criteria. To ensure the continuity of the segments, the length of the directional conflict segment is set to be no less than 3 time points. If t4, t5, and t6 all meet the above directional inconsistency conditions, then t4 to t6 is defined as the directional conflict path interval. Subsequently, multiple time segments that meet the continuous difference requirements are marked and converted into a set of time position sequences. Finally, this set of time position sequences is combined with the path index information to form a set of directional inconsistent paths, which is the set of health indicator conflict segments.

[0071] Please see Figure 3 The specific steps of S2 are as follows:

[0072] S201: Based on the indicator path of the health indicator conflict segment concentration, the directional trend of the path at continuous positions is extracted along the time series. The position of directional change is judged according to the increase or decrease of adjacent points. The directional change nodes are summarized according to the trend characteristics of differentiated directional switching to obtain the sequence of directional change positions.

[0073] First, a single path is selected as the operation sample. The direction of indicator value changes at consecutive positions is read sequentially along its time series. For example, in path P1, the indicator value sequences at time points t1, t2, and t3 are v1 = 1.2, v2 = 1.4, and v3 = 1.1, respectively. By observing the direction difference between each pair of time points, we can see that t1 to t2 is an increase, and t2 to t3 is a decrease. Therefore, t2 is determined to be the point of direction change. In this process, it is not necessary to directly calculate the difference, but it is necessary to determine the increase or decrease status of adjacent points. Then, all consecutive time points in path P1 are traversed, recording the direction status of the indicator value between each pair of time points. These direction statuses are then sequentially collected into a direction sequence. Next, adjacent direction items in the direction sequence are judged. When there is a "rise → fall" or "fall → rise" direction switch... At that time, the position is marked as a direction change node. For example, if the direction sequence is "↑↑↑↓↓", then the middle position is marked as the switching point. The same process is performed on all paths in sequence to obtain a list of direction switching positions of different paths in their respective time periods. Next, these switching nodes need to be aggregated to summarize the situation where multiple paths switch directions at similar time points into the same type of direction conversion behavior. Then, according to the difference characteristics of the direction switching positions between paths, different types of direction conversion points are grouped. For example, if a certain type of switching point is mainly concentrated at the inflection point of the trend from rising to falling, then this type can be classified as a downward conversion feature. Then, they are numbered and marked. Finally, all direction conversion points are numbered and output to form a clear sequence of direction change markers, thus obtaining the direction change position sequence.

[0074] S202: Based on the position sequence of direction change, locate the path segment before and after the direction switch, identify continuous indicator points in the path segment, compare the starting and ending indicator values ​​between adjacent segments in turn, calculate the indicator offset before and after the direction change, and obtain the conversion trend offset indicator group.

[0075] The specific formula for calculating the index offset before and after the direction change is as follows:

[0076]

[0077] Where, ΔD i A represents the index offset before and after the direction change of the i-th path segment. i λ represents the weighted average amplitude of the continuous index point sequence of the i-th path segment. i The index coefficient representing the proportion of points with the same direction in the i-th path segment, n i P represents the total number of index points within the i-th path segment. ij This represents the j-th actually measured health indicator value in the i-th path segment. This represents the trend fit estimate of the j-th indicator point in the i-th path segment, ∈ ij The trend residual rate represents the value at the j-th point in the i-th segment;

[0078] Formula operation logic: The first term in the formula reflects the stability of the overall path in the macro trend by multiplying the amplitude value with the trend direction and normalizing it. The second term is the weighted sum of the absolute differences between the indicator points and the trend line, which is then averaged and squared to measure the interference of local fluctuations on the overall trend judgment. The two terms are combined by the sum of squares and the square root to form the macro and micro deviations of the path.

[0079] Using heart rate variability (HRV) as the monitoring object, five consecutive time points were collected as follows:

[0080] 50ms, 60ms, 55ms, 65ms, 70ms;

[0081] The index point sequence P i1 =50, P i2 =60, P i3 =55, P i4 =65, P i5 =70;

[0082]

[0083] A i The average of the five points is 60ms, and the sum of the absolute values ​​of the corresponding differences is 10+0+5+5+10=30ms;

[0084] Taking the average, we get A i =30 / 5=6ms;

[0085] λ i =0.8;

[0086] Point residual rate:

[0087] Point 1 is ∈ i1 =|50-52| / 52≈0.0385;

[0088] Point 2 is ∈ i2 =|60-62| / 62≈0.0323;

[0089] The remaining values ​​are 0.0351, 0.0317, and 0.0294, respectively.

[0090] The weighted absolute differences were obtained as 0.077, 0.0646, 0.0702, 0.0634, and 0.0588.

[0091] The summation is 0.334, and the average value is 0.334 / 5 ≈ 0.0668 ms;

[0092] Substitute into the calculation:

[0093]

[0094] The square of the second term is (0.0668). 2 ≈0.00446, adding the two results in 0.9261;

[0095] Taking the square root, we get:

[0096]

[0097] Interpretation of results and numerical significance: In the path of change of the segment indicator, the average offset in the trend fitting level before and after the direction change is about 0.9623ms. Combined with the corresponding benchmark threshold of 1ms, the indicator fluctuation is in a critical state and belongs to a slight trend deviation.

[0098] Explanation of the innovative aspects of the formula:

[0099] The advantage of the formula is that by simultaneously introducing three structured parameters—path amplitude, trend consistency, and trend residual rate—the indicator deviation assessment can cover trend stability, short-term disturbances, and overall consistency, thereby improving the sensitivity of identifying subtle trend changes and the ability to determine stratification in health status assessment.

[0100] S203: Based on the trend offset indicator group, filter the path segments with continuous and consistent indicator trends, locate the trajectory range with unchanged direction according to the time direction, track the continuous trend within the path segment, and obtain the candidate trend segments of state change.

[0101] First, each indicator path in the indicator group is selected as a separate processing object. The indicator values ​​of consecutive points along the path in the time series are read sequentially, and the continuity of the path trend is determined based on the direction of value changes. For example, in indicator path A, if the indicator values ​​are 1.1, 1.3, 1.4, 1.5, and 1.7 in five consecutive time positions t1 to t5, then the path segment can be determined to have a continuous upward trend within this time range and is marked as a consistent trend segment. Further filtering is performed to identify all path segments with similar continuous trend directions, excluding those with a change in direction between any two adjacent positions. Simultaneously, to eliminate fluctuation interference, a direction continuity confirmation criterion needs to be set, stipulating that a valid continuous trend segment is considered when the number of points with a consistent direction is not less than five. Subsequently, these filtering... The continuous trend segments are scanned sequentially along the time direction to determine their direction maintenance status. The direction of adjacent points on each path segment is judged, and discontinuous paths are eliminated, retaining only areas with consistent trends. Then, the longest sub-segment in each segment where the trend direction has not changed is selected as the main trend segment, and its start and end positions are obtained to form a continuous trajectory range. Next, the trend continuation characteristics of each main trend segment are tracked, and its direction maintenance status at subsequent time positions is extracted to observe whether the trend continues to expand. In path B, if it is in an upward trend from t1 to 6 and continues to rise in t7 and t8, its extended trend segment range is extended to t8. Finally, the trend characteristics of all processed path segments are combined, and path segments with consistent continuity are summarized to obtain candidate trend segments with state changes.

[0102] Please see Figure 4 The specific steps of S3 are as follows:

[0103] S301: Based on the continuous segments in the candidate trend segments of state change, read the up and down fluctuation direction of the indicator path in each segment, follow the position of the direction trend change along the time sequence, and connect the fluctuation sequence in order to obtain the fluctuation direction sequence group.

[0104] First, multiple indicator paths are extracted from each segment. By reading the numerical changes of the indicators at adjacent positions point by point, the directional state is identified as either upward or downward. The adjacent directional states are then recorded to construct a directional trend chain. For example, if path A has a value of 7.2 at position x1 and 7.5 at x2, then x1 to x2 can be considered an upward segment. Similarly, if path B has a value of 8.1 at x3 and 7.8 at x4, then this segment is a downward segment. The directional trends of multiple paths, such as path A, path B, and path C, are then recorded sequentially along the timeline, thus forming a complete directional chain of indicator paths within the segment. During the path direction recording process, the position of each node is adjusted according to the time sequence. The algorithm sorts the rows to ensure the continuity of subsequent directional changes. Then, it extracts the directional change positions on the sorted structure. For example, if the directional state of path C reverses from continuous rise to fall, this node is recorded as a key turning point. The algorithm then continues to extract the directional sequence of subsequent fluctuations. Next, it constructs a complete directional change path in chronological order. Then, it connects the preceding and following trends by using the directional reversal nodes as connection points. For example, if path D is falling at x6 and then turns to rising after x7, this structure is connected as an fluctuation unit in the directional change chain. Finally, it completes the temporal combination and connection of the indicator directional states on all continuous segments to obtain the fluctuation directional sequence group.

[0105] S302: Based on the fluctuation direction sequence group, the fluctuation segments are divided according to the continuity of the directional trend within the path, the rhythm sequence of directional maintenance and change is identified, the structural ratio of the rhythm continuity of the path segment to the intensity of directional continuation is calculated, and the rhythm feature matching sequence set is obtained.

[0106] The formula for calculating the structural ratio of rhythmic continuity to directional continuity intensity of a path segment is as follows:

[0107]

[0108] Among them, R k M represents the structural ratio of the rhythmic continuity of the path segment to the directional continuity intensity of the fluctuation segment k. k d represents the total number of directional nodes in the fluctuation segment k. k,m ω represents the number of times the direction of the m-th node in the fluctuation segment k is maintained. k,m This represents the rhythm weight of a node in the path direction sequence. μ represents the average frequency of directional changes in the fluctuation segment k after weighting. k ε represents the average frequency of unweighted overall directional changes, and ε represents a constant offset term to prevent the denominator from being zero.

[0109] Formula operation logic: In the formula, the summation operation ∑ω in the numerator part k,m · d k,mThe logic is to integrate the rhythm weight and maintenance strength of nodes to construct a cumulative and continuous trend performance; the denominator part +ε is used to capture the degree of deviation in rhythm frequency and avoids the masking effect of directional fluctuations in the form of absolute deviation, thus making the ratio more valuable for identifying the consistency of direction and rhythm.

[0110] There are 4 nodes in the fluctuation section, numbered from m=1 to m=4 respectively;

[0111] d k,m Obtained through trend image sampling, set as: 4, 5, 3, 4 (times);

[0112] v k,m Recorded within a 1-second period using a direction change counter;

[0113] The results were: 1.2, 1.0, 1.3, 1.1 (times / second);

[0114] The rhythm matching degree is obtained by standard sequence template matching quantization:

[0115] Match rates of 90%, 80%, 95%, and 85%;

[0116] The weights are mapped to ω respectively. k,m =0.9, 0.8, 1.1, 1.0;

[0117] Calculate the mean frequency of directions:

[0118]

[0119] Calculate the rhythm-weighted average frequency:

[0120]

[0121] Calculate the numerator:

[0122]

[0123] Calculate the denominator:

[0124]

[0125] Substituting into the formula, we get:

[0126]

[0127] Interpretation of results and numerical significance: In the current path segment, the rhythm matching weight is high and the number of directional continuations is stable, with small frequency changes and deviations, showing overall consistency and rhythmic coherence.

[0128] Explanation of the innovative aspects of the formula:

[0129] The advantage of the formula lies in the introduction of a weighted rhythm weight ω. k,m with weighted frequency mean Coupled identification of directional consistency and rhythm offset within the path was performed, avoiding the drawback of traditional frequency analogy values ​​ignoring rhythm structure, and enhancing the controllability and accuracy of feature selection.

[0130] S303: Based on the rhythm feature matching sequence set, the trajectory segments of the same rhythm sequence are extended along the time line, and the upper and lower nodes are connected in a synchronous manner according to the direction of fluctuation to obtain the main channel structure segment of health assessment.

[0131] First, the index path trajectory segment corresponding to each rhythm sequence is read. The continuous intervals of rhythmic fluctuations are marked within the index path. The temporal order of continuous rhythmic segments is extracted and arranged. Then, these trajectory segments with the same rhythmic structure are extended forward and backward in time. When processing rhythmic path A, if its rhythmic pattern is "rising-falling-rising-falling," and the same pattern appears in path B, then the segments with the same rhythmic structure corresponding to the two paths are marked as rhythmically consistent segments. Subsequently, extension processing is performed based on the position of these consistent segments on the time axis. That is, using the start and end points of the original segment as reference points, the continuation structure of the same rhythmic pattern is searched forward or backward, and all segments that can be extended under the same fluctuation structure are included. Each segment is categorized into extendable trajectory segments. Then, it is determined whether the fluctuation direction of these trajectory segments is consistent at adjacent nodes. By identifying the direction, it is analyzed whether the nodes in each segment are synchronous at the rhythm switching points. For example, if the upward or downward directions of paths A and C are completely consistent in four consecutive rhythm nodes, they are determined to be synchronous continuable segments. The entire segment from the starting point to the ending point is included in the continuation matching structure and is connected with the preceding extended trajectory segments. This makes multiple rhythm trajectory segments form a continuous channel in the timeline. Then, from all the confirmed extended segments, the set of segments with consistent rhythm structure and continuous direction connection is identified, and finally, the main channel structure segment of health assessment is obtained.

[0132] Please see Figure 5 The specific steps of S4 are as follows:

[0133] S401: Based on the indicator path in the main channel structure of health assessment, track the trend fluctuations along the time extension direction of the path, compare the position of the continuous change direction in the time series, and obtain the trend turning point position group.

[0134] First, extract consecutive nodes according to the time sequence involved in each path, read the indicator value corresponding to each node, and determine the trend of the current segment by the direction of value change between nodes. For example, if the value of path P1 is 5.2 at position a and 5.6 at position b, the direction of this segment can be determined to be upward. If the value at position c is 5.3, then the trend from b to c is downward. Continue to extract the indicator values ​​of subsequent nodes along the path sequence, and connect them sequentially according to the direction of change to form trend segments. Then, aggregate the trend segments formed by each path to form a complete continuous trend sequence. In this trend sequence, identify the start and end nodes corresponding to path segments with the same continuous trend direction, and determine the direction change within any consecutive segment. The node positions are extracted centrally. For example, in path P2, if ab is rising, bc is falling, and cd is falling, then position b can be added to the candidate set as a trend direction change position. Then, the same action is performed in other paths such as P3 and P4. The changes in the trend direction are scanned one by one according to the time sequence, and each turning point is marked. Then, these position changes in direction are compared in all paths. The effective turning point is located according to whether it is a continuous rise to fall or a continuous fall to rise. Nodes that only have slight fluctuations but do not constitute a continuous trend change are eliminated. Finally, the direction reversal nodes in the indicator paths under all main channel structure segments are centrally located to obtain the trend turning point position group.

[0135] S402: Based on the trend turning point location group, track the trend direction between adjacent turning points along the trajectory segment, identify the path segment direction with inconsistent changes before and after the direction, and obtain the direction divergence trajectory interval group.

[0136] First, the trajectory segments between two adjacent turning points within each path are extended, and the trend of the indicator direction between each segment is extracted. Directional change segments are constructed between nodes, and the indicator values ​​within each segment are extracted sequentially. The trend direction presented by each segment is determined. For example, if the value of a trajectory segment continuously increases from node A to node B, it can be identified as an upward trend segment; if the value continuously decreases from node B to node C, it is a downward trend segment. This process is repeated to establish a complete trend sequence, while recording the corresponding time interval and path number. Next, the trend direction changes of adjacent segments are compared path by path in the path group to identify segments with inconsistent directions. For example, when consecutive segments on path P1 show... If path P1 rises and then falls, while path P2 rises and falls within the same time frame, then the segments containing these two paths constitute segments with inconsistent directions. The same identification operation is then performed on all paths, and the path segments that meet the condition of inconsistency before and after the direction change are integrated into the extraction interval. During the extraction process, segments with direction change amplitudes less than the set fluctuation intensity benchmark value are excluded. The benchmark value can be set to 0.2, referencing the median of historical fluctuation amplitudes on that path. In actual operation, if the difference between the upper and lower values ​​of a turning point in path P3 is only 0.05 and the direction change is not obvious, it is not included in the identification results. Finally, all trajectory segments with differing direction changes are integrated to obtain a group of directional divergent trajectory intervals.

[0137] S403: Based on the path direction differences in the directional divergence trajectory interval group, analyze the trend extension of the trend continuation line at the alternation of direction changes, extract the interval segments with different directions and not maintaining the same direction of advancement along the path direction, and obtain the healthy state double-track bifurcation segment group.

[0138] First, locate the sequence of changes in path direction within each interval. Read the extension trend of each path segment between two turning points, and label it as either rising or falling based on continuity. Then, observe the trend lines at the points where different paths alternate in direction, extracting the fluctuation trends between adjacent paths. During this process, determine if there is any deviation in direction extension based on the consistency of direction in continuous path segments. If a segment shows a reverse trend relative to the path when connecting, it is recorded as a deviation point. For example, path A continuously rises between segments T1 and T3, while path B continuously falls within the same segment, forming an alternation in direction at node T2. Next, analyze the paths before and after this alternation point. The trend extension structure is displayed, and features of the subsequent directional trend within the continuation path are extracted to determine whether the continuation direction continues to differ from the direction of the starting alternation point within a certain range. If multiple consecutive nodes have the same direction and maintain an opposite relationship with another path, the directional behavior is defined as a non-co-directional advancement segment. Then, a minimum extension requirement threshold is set according to the extension length of the difference between paths. This threshold can be adjusted by the average value of the path continuation segments in historical data. For example, in a detection, the minimum number of points in the directional extension segment is set to 3, that is, only three consecutive extension nodes with the same direction can constitute a difference segment. Finally, the path segments with directional deviation and meeting the minimum extension requirement are classified and extracted to the healthy state double-track bifurcation segment group.

[0139] Please see Figure 6 The specific steps of S5 are as follows:

[0140] S501: Based on the time sequence of path segments in the healthy state double-track bifurcation segment group, identify the starting point of the path and the node where the direction first reverses, connect the direction convergence segment in the tail trend, and obtain the trend stage boundary sequence group.

[0141] First, the starting position of each path segment is read. Then, each node from the starting point of a path segment is checked for the corresponding indicator's fluctuation direction to determine if there is a continuous upward or downward trend. When the fluctuation direction differs from the starting direction, and this change is the first reversal in the current segment, that node is recorded as the first reversal node. For example, if the fluctuation direction at the starting point of path X is positive, and the direction first changes to negative at the 5th time point, then the 5th node is recorded as the reversal point. This identification process is repeated for each path. Next, at the end of the path segment, it is determined whether the subsequent direction maintains a continuous trend. The sequence of consecutive direction nodes at the end is read, and the last segment of the sequence is judged for a continuous trend. If the path forms three or more consecutive nodes with the same direction at the end, it is considered... For example, if path Y descends without reversal between the fourth and second to last nodes, then the tail direction is a descending convergence segment. Connect this segment from the reversal point to the segment with the same tail direction to construct the stage segment of the path. For all path segments, complete the time positioning actions of the starting point, the first reversal node, and the tail direction convergence segment. Arrange these three types of key points in all path segments in chronological order and uniformly number their time positions. For example, if the three key points of path A are located at nodes 3, 7, and 10, then the three points are numbered A-1, A-2, and A-3 respectively. After completing the above operations, merge the time point sets composed of the three points in each path to obtain the trend stage boundary sequence group.

[0142] S502: Based on the continuous trend of the path direction between adjacent nodes in the trend stage boundary sequence group, the trend extension segment is traced along the time direction to identify the trend direction structure where the direction has not changed, and the set of trend extension intervals is obtained.

[0143] First, the acquired trend boundary time point sequence is read from each path. Path segments between adjacent nodes are delineated according to these time points. Then, the direction of each path segment is identified, and the indicator values ​​corresponding to the first and last nodes of the segment are extracted. Their magnitude relationship is determined to ascertain whether the overall fluctuation trend of the current segment is upward or downward. Next, the direction of each intermediate node within the path segment is extended to check if it is consistent with the overall trend of the segment. If the direction of an intermediate node is inconsistent with the overall direction, the segment is marked as a trend interruption segment. If the direction of an intermediate node is consistent with the overall direction, the segment is marked as a trend continuity segment. For example, in path P, the indicator value at the beginning and end of the path segment between time points T1 and T4 increases from 35 to 40. The indicator values ​​at intermediate time points T2 and T3 are 36 and 38 respectively, both in an upward direction. Therefore, T1- The T4 path segment is recorded as a continuous upward segment. Conversely, if the value of an intermediate node decreases, it is not recorded as a trend extension segment. Subsequently, the continuous trend segments in each path are continuously tracked along the time direction. The starting and ending points are labeled and grouped into a trend segment number. For example, the continuous trend segment in path A starts at time 5 and ends at time 11, and is numbered A-Trend-1. This number and the corresponding time range information are recorded, and its fluctuation direction is retained. Then, the same action is repeated for the path segment starting at the next node. The structure of the continuous trend in each path segment is identified segment by segment until all adjacent trend boundary nodes of each path are processed. If there are continuous trend segments with the same time range in multiple paths, their path numbers are recorded and grouped for statistics. Finally, the numbers and corresponding time ranges of all trend continuous path segments are collected to obtain the trend extension interval set.

[0144] S503: Based on the trend direction distribution of path segments in the trend extension interval set, filter areas with consistent directions, sequentially associate trend information, and obtain health status classification label groups.

[0145] First, the trend direction record for each path is extracted from the set of trend extension intervals. The direction indicator value of each path segment within a continuous time range is read, with positive values ​​representing a continuous upward trend, negative values ​​representing a continuous downward trend, and zero values ​​indicating a stable state. The direction indicator values ​​of adjacent path segments are compared sequentially. If the direction indicator values ​​of consecutive segments are equal and non-zero, the path segment is marked as a region with consistent direction. For example, if the direction values ​​of path X in extension intervals A, B, and C are +1, +1, +1 respectively, then A to C are merged into a region with consistent direction. If the direction value of segment B is -1, it is considered a direction interruption, and the region with consistent direction is re-marked. The process is repeated to find the next continuing direction segment, obtaining continuous segments with consistent direction on each path. Then, the corresponding time intervals are obtained from these regions with consistent direction. The absolute change range of indicators within these intervals is read, and the corresponding classification threshold groups are used to assess the directional level. In an upward trend, an amplitude of less than 0.3 is defined as a low amplitude segment, between 0.3 and 0.7 as a medium amplitude segment, and greater than 0.7 as a high amplitude segment. In a downward trend, the same intervals are used to divide the level in reverse. Each segment with the same direction is labeled with the corresponding level. The indicator name of the path segment is read, and the label level combination is collected according to the indicator category. Then, the number of paths with the same direction in the same time period is extracted from multiple paths. If there are 3 or more paths with an upward trend in the time period, the overall trend level in the time period is increased by one level. Conversely, if the trend is downward, the overall level is decreased by one level. By superimposing the path direction level in this way, the comprehensive health trend label is labeled, and finally, the health status classification label group is obtained.

[0146] Please see Figure 7 An intelligent health assessment and analysis system based on health management includes:

[0147] The conflict segment identification module obtains the indicator group of the fluctuating marked segment within the health monitoring cycle, identifies the indicator path with different directional trends within the same time slice, tracks the continuous segment offset changes along the path, locates segments with inconsistent directions and overlapping changes, and obtains the set of conflict segments of health indicators.

[0148] The trend segment extraction module is based on the indicator path of the health indicator conflict zone. It follows the trend change along the time series, locates the position where the upward and downward directions switch, splits the path according to the switching point, and filters out the path with a continuous consistent direction to obtain the candidate trend segment of state change.

[0149] The rhythm path filtering module reads the direction of up and down fluctuations, analyzes the order of fluctuations, and determines whether rhythm repetition occurs based on continuous segments in the candidate trend segments of state changes. It then extends the fluctuation range in the repeated rhythm path to obtain the main channel structure segment of health assessment.

[0150] The track path decomposition module extracts trend turning points and identifies trajectory segments with inconsistent directional changes based on the time extension relationship of indicator paths in the main channel structure segment of health assessment. It then splits the path according to the divergence point, decomposes the directional repetition separation segment, and obtains the health status double track bifurcation section group.

[0151] The label path tracking module identifies the start, transition, and end positions of trends based on the trend sequence of path segments in the dual-track bifurcation segment group of health status, tracks the continuation of trend direction, and obtains the health status classification label group.

[0152] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. An intelligent health assessment and analysis method based on health management, characterized in that, Includes the following steps: S1: Obtain the indicator group of the fluctuating marked segment within the health monitoring period, extract the inconsistent paths within the time period, track the offset changes of continuous segments, locate the positions where the directions diverge and the changes are synchronized, and obtain the set of conflicting health indicator segments. S2: Based on the path of the health indicator conflict segment set, identify the direction rise and fall switching points along the time series, divide the path according to the position, extract the path with the same direction, and obtain the candidate trend segment of state change. S3: Based on the continuous path in the candidate trend segment of the state change, read the upward and downward fluctuation direction of each segment, analyze the fluctuations in time order, identify the rhythm repeating path, extend the fluctuation range, and obtain the main channel structure segment of health assessment. S4: Based on the main channel structure segment of the health assessment, extract the trend turning point, identify the trend with inconsistent direction, divide the path structure according to the continuous area of ​​repeated separation of trend direction, and obtain the dual-track bifurcation segment group of health status. S5: Based on the path trend in the dual-track bifurcation segment group of the health status, identify the starting, changing and ending positions of the dominant trend, track the directional trend along the time sequence, analyze the trend continuation mode, and obtain the health status classification label group.

2. The intelligent health assessment and analysis method based on health management according to claim 1, characterized in that, The set of conflicting health indicators includes segments with inconsistent path directions, segments with overlapping trends, and conflicting intervals. The candidate trend segments for state changes include segments with consistently consistent directions, trend turning points, and path splitting segments. The main channel structure segments for health assessment include segments with repetitive rhythms, fluctuation sequence patterns, and undulating extension segments. The dual-track bifurcation segment group for health status includes trend divergence trajectory segments, trajectory direction difference points, and trend isolation segments. The health status grading and judgment label group includes the starting point of the dominant trend, trend continuation characteristics, and trend termination markers.

3. The intelligent health assessment and analysis method based on health management according to claim 1, characterized in that, The aforementioned direction shift point refers to identifying the relationship between the increase and decrease of indicator values ​​at adjacent time positions during trend segment extraction, and extracting the critical position point from rising to falling and from falling to rising. The extended fluctuation range refers to the process of extending the index path segment with the same fluctuation rhythm sequence along the time axis during the rhythm feature screening process, analyzing the synchronous segment of continuous rhythm matching, and connecting the directional consistency of the upper and lower nodes.

4. The intelligent health assessment and analysis method based on health management according to claim 1, characterized in that, The trend turning point refers to the spatial and temporal nodes in the main channel structure segment where the direction and trend change, and the reversal nodes from an upward trend to a downward trend and from a downward trend to an upward trend. The trend continuation method refers to analyzing the continuity and changing trend of the trend extension path in the health status label based on the trend direction relationship and turning point and convergence performance of adjacent segments.

5. The intelligent health assessment and analysis method based on health management according to claim 1, characterized in that, The specific steps of S1 are as follows: S101: Obtain the indicator group of the fluctuation marker segment within the health monitoring period, read the trend direction of the indicator path within the same time slice, locate the position where the direction of the path is inconsistent, delineate the path set with direction difference according to the time slice, and obtain the path set with direction difference. S102: Based on the path data in the set of directional difference paths, read the offset behavior in continuous segments in chronological order, identify the offset direction of the corresponding path at adjacent time positions, identify the path intervals with synchronized directional changes within the same time range, and obtain the set of synchronized offset intervals. S103: Based on the path trajectory in the set of synchronous offset intervals, track the path ranges with different directions and adjacent change times at continuous time positions, filter the time positions of synchronous offset with inconsistent directions, and obtain the set of health indicator conflict segments.

6. The intelligent health assessment and analysis method based on health management according to claim 1, characterized in that, The specific steps of S2 are as follows: S201: Based on the indicator path of the health indicator conflict segment set, extract the directional trend of the path at continuous positions along the time series, determine the position of directional change according to the increase or decrease of adjacent points, summarize the directional change nodes according to the trend characteristics of differentiated directional switching, and obtain the directional change position sequence. S202: Based on the direction change position sequence, locate the path segments before and after the direction switch, identify continuous indicator points in the path segments, compare the start and end indicator values ​​between adjacent segments in turn, calculate the indicator offset before and after the direction switch, and obtain the conversion trend offset indicator group. S203: Based on the aforementioned conversion trend offset index group, filter path segments with continuous and consistent index trends, locate the trajectory range with unchanged direction according to the time direction, track the continuous trend within the path segment, and obtain candidate trend segments for state change.

7. The intelligent health assessment and analysis method based on health management according to claim 1, characterized in that, The specific steps for S3 are as follows: S301: Based on the continuous segments in the candidate trend segments of the state change, read the up and down fluctuation direction of the indicator path in each segment, follow the position of the direction trend change along the time sequence, and connect the fluctuation sequence in order to obtain the fluctuation direction sequence group. S302: Based on the wave direction sequence group, divide the wave segments according to the continuity of the directional trend within the path, identify the rhythm sequence of directional maintenance and change, calculate the structural ratio of the rhythm continuity of the path segment to the directional continuation intensity, and obtain the rhythm feature matching sequence set. S303: Based on the rhythm feature matching sequence set, extend the trajectory segments of the same rhythm sequence along the timeline, and connect the upper and lower nodes in a synchronous manner according to the direction of fluctuation to obtain the main channel structure segment of health assessment.

8. The intelligent health assessment and analysis method based on health management according to claim 1, characterized in that, The specific steps of S4 are as follows: S401: Based on the indicator path in the main channel structure segment of the health assessment, track the trend fluctuation pattern along the path time extension direction, compare the position of the continuous change direction in the time series, and obtain the trend turning point position group. S402: Based on the trend turning point location group, track the trend direction between adjacent turning points along the trajectory segment, identify the path segment direction with inconsistent changes before and after the direction, and obtain the direction divergence trajectory interval group. S403: Based on the path direction differences in the directional divergence trajectory interval group, analyze the trend extension of the trend continuation lines at the alternation of direction changes, extract the interval segments with different directions that do not maintain the same direction of advancement along the path direction, and obtain the healthy state double-track bifurcation segment group.

9. The intelligent health assessment and analysis method based on health management according to claim 1, characterized in that, The specific steps of S5 are as follows: S501: Based on the time sequence of the path segments in the healthy state bifurcated segment group, identify the starting point of the path and the node where the direction first reverses, connect the direction convergence segment in the tail trend, and obtain the trend stage boundary sequence group. S502: Based on the continuous trend of the path direction between adjacent nodes in the trend stage boundary sequence group, track the trend extension segment along the time direction, identify the trend direction structure where the direction has not changed, and obtain the trend extension interval set. S503: Based on the trend direction distribution of the path segments in the trend extension interval set, filter out areas where the direction remains consistent, and sequentially associate trend information to obtain a health status classification label group.

10. An intelligent health assessment and analysis system based on health management, characterized in that, The system is used to implement the intelligent health assessment and analysis method based on health management as described in any one of claims 1-9, and the system comprises: The conflict segment identification module obtains the indicator group of the fluctuating marked segment within the health monitoring cycle, identifies the indicator path with different directional trends within the same time slice, tracks the continuous segment offset changes along the path, locates segments with inconsistent directions and overlapping changes, and obtains the set of conflict segments of health indicators. The trend segment extraction module, based on the indicator path of the health indicator conflict segment set, follows the trend change along the time series, locates the position where the upward and downward directions switch, splits the path according to the switching point, and filters out the path with a continuous consistent direction to obtain the candidate trend segment of state change. The rhythm path filtering module reads the upward and downward fluctuation direction, analyzes the fluctuation sequence, and determines whether rhythm repetition occurs based on the continuous segments in the candidate trend segments of the state change. It extends the fluctuation interval in the repeated rhythm path to obtain the main channel structure segment of health assessment. Based on the time extension relationship of the indicator path in the main channel structure segment of the health assessment, the track path decomposition module extracts the trend turning point, identifies the trajectory segment with inconsistent directional changes, splits the path according to the divergence point, decomposes the directional repetition separation segment, and obtains the health status double track bifurcation section group. The label path tracking module identifies the start, transition, and end positions of the trend based on the trend sequence of the path segments in the dual-track bifurcation section group of the health status, tracks the continuation of the trend direction, and obtains the health status classification label group.

Images