A sports medicine care force adaptation modeling method

Abstract

The application discloses a kind of sports medical nursing stress adaptation modeling method, it is related to sports medical nursing technical field, including the following steps: nursing action constraint information in the postoperative sports medical nursing training sequence of sports medical is carried out multidimensional mapping processing, generates nursing constraint sequence, and corresponding stress data sequence is bound according to time corresponding relationship, forms semantic enhancement stress data sequence;Utilize nursing constraint sequence to construct semantic difference evolution sequence, whether nursing action constraint is switched is judged by difference calculation to adjacent time nursing constraint sequence, and in the case where nursing action constraint switches, difference analysis is carried out to the nursing constraint sequence before and after switching, determine constraint difference degree and form difference representation sequence.The application solves the stress data aliasing problem under nursing action constraint switching, realizes that stress data is accurately segmented according to constraint difference and dynamic adaptation modeling.

Description

Technical Field

[0001] This invention relates to the field of sports medicine and nursing technology, specifically to a method for force adaptation modeling in sports medicine and nursing. Background Technology

[0002] Postoperative stress adaptation modeling in sports medicine typically refers to the process of collecting multidimensional data on a patient's movements, postures, range of motion, speed, and contact force during rehabilitation exercises. This data is collected during the rehabilitation phase following surgery in the sports medicine department, taking into account the needs of various aspects such as sports training, clinical medical assessment, and nursing intervention. The data is then processed and analyzed using a computer system to construct a biomechanical model reflecting the stress state of the human skeletal and muscular systems. This model is combined with individualized medical information (such as surgical method, recovery stage, and functional level) and nursing guidelines (such as load limitations, rehabilitation rhythm, and safety requirements) to analyze and evaluate the stress state during exercise. The process usually includes multiple stages such as sports data collection, medical status assessment, nursing constraint modeling, stress calculation and analysis, and result output and control. Through the synergistic effect of sports, medical care, and nursing, the modeling and adaptation of stress during rehabilitation training is achieved.

[0003] The existing technology has the following shortcomings:

[0004] In the process of force adaptation modeling for postoperative sports medicine nursing, when different nursing action constraints exist in the same training sequence and switch during execution, the force data collected at different stages have fundamentally different constraint conditions due to the differences in the force limitation range and motion control requirements corresponding to different nursing constraints. However, these data are still continuously input into the same modeling calculation process during data organization. Existing technologies cannot segment and isolate or label the force modeling input data according to the degree of constraint difference when switching nursing action constraints. This leads to the aliasing of force data under different nursing constraint conditions during the modeling process, making it impossible for the model to identify the specific nursing constraint state corresponding to the data. As a result, the force modeling results lack specificity, which in turn affects the accuracy of nursing adaptation analysis.

[0005] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0006] The purpose of this invention is to provide a method for force adaptation modeling in sports medicine nursing to solve the problems mentioned in the background art.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a method for modeling stress adaptation in sports medicine nursing, specifically comprising the following steps: S1. Perform multidimensional mapping processing on the nursing action constraint information in the postoperative sports medicine nursing training sequence to generate a nursing constraint sequence, and bind it with the corresponding force data sequence according to the time correspondence to form a semantically enhanced force data sequence. S2. Construct a semantic difference evolution sequence using the nursing constraint sequence. Calculate the difference between nursing constraint sequences at adjacent time points to determine whether nursing action constraints have switched. If nursing action constraints have switched, analyze the difference between the nursing constraint sequences before and after the switch to determine the degree of constraint difference and form a difference representation sequence. S3. Based on the difference representation sequence, the force modeling input data is segmented and isolated or processed according to the degree of constraint difference. The semantically enhanced force data sequence is reconstructed into multiple data intervals corresponding to different nursing action constraints, and a state label vector corresponding to the degree of constraint difference is embedded in each data interval. S4. Perform joint calculations on each data interval and status identifier vector, extract the force adaptation parameters and nursing constraint consistency parameters, generate evaluation coefficients through combined calculations, and classify the data intervals according to preset intervals. S5. In response to the classification results, adjust the organization and calculation participation relationships of the input data corresponding to the data intervals to generate updated force modeling input data.

[0008] Preferably, S1 specifically includes the following steps: Multidimensional mapping processing is performed on the nursing action constraint information in the postoperative sports medicine nursing training sequence. The nursing action constraint information is split according to the constraint type dimension, constraint direction dimension, and constraint intensity dimension, and numerical mapping is performed on each dimension to generate multidimensional mapping results. The multidimensional mapping results are then arranged in chronological order to generate a nursing constraint sequence. The nursing constraint sequence is bound to the corresponding force data sequence according to the time correspondence. By constructing a unified time index, the multidimensional mapping results in the nursing constraint sequence are matched point by point with the force data at the corresponding time position in the force data sequence to form a time-aligned data combination. After establishing the time correspondence, the multidimensional mapping results in the nursing constraint sequence are embedded into the corresponding force data. By combining the multidimensional mapping results and the force data, a semantically enhanced force data sequence containing nursing action constraint information and force information is formed.

[0009] Preferably, S2 specifically includes the following steps: S201. Pair the multidimensional mapping results arranged in time order in the nursing constraint sequence with adjacent time points, perform dimension-wise difference calculation on each pair of adjacent time points to generate the corresponding difference vector, and arrange each difference vector in time order to form a semantic difference evolution sequence. S202. Perform dimension-by-dimensional analysis on the difference vectors in the semantic difference evolution sequence. Calculate the change in the difference vectors in each dimension and compare the change with the preset benchmark change in the corresponding dimension. When the change in any dimension is greater than the preset benchmark change in the corresponding dimension, determine the nursing action constraint switch at the corresponding time position and record the switch position. S203. When nursing action constraints change, select the multidimensional mapping results in the nursing constraint sequence before and after the change position for dimension-by-dimensional comparison calculation, combine and express the differences in each dimension to determine the degree of constraint difference, and arrange the degree of constraint difference in time order to generate a difference representation sequence.

[0010] Preferably, S203 is as follows: When nursing action constraints change, the multidimensional mapping results in the nursing constraint sequence within the corresponding time interval are extracted according to the switching position. The sets of multidimensional mapping results for several consecutive moments before the switching position and the sets of multidimensional mapping results for several consecutive moments after the switching position are obtained respectively, and the correspondence between the two is established in chronological order. The multidimensional mapping results before and after the position switch are compared and calculated dimension by dimension. The difference between the values ​​of the same dimension before and after the switch is calculated to obtain the difference of each dimension. The difference is arranged in the order of the dimensions to form a set of dimension differences. The set of dimension differences is weighted and combined to generate the degree of constraint difference of the corresponding time position. The degree of constraint difference calculated at each time point is arranged in chronological order, and the degree of constraint difference is continuously organized to form a difference characterization sequence consistent with the time axis of the nursing constraint sequence.

[0011] Preferably, S3 specifically includes the following steps: S301. Based on the difference characterization sequence, extract the degree of constraint difference corresponding to each time position in chronological order, compare the degree of constraint difference between adjacent time positions numerically, and when the difference between the degree of constraint difference between adjacent time positions is greater than the preset difference threshold, determine the corresponding time position as the segment boundary, and perform segment isolation processing on the force modeling input data based on the segment boundary to form multiple time-continuous data segments. S302. Reconstruct the semantic enhancement force data sequence according to the segment boundaries, merge the semantic enhancement force data in the same segment interval according to the time order to form multiple data intervals corresponding to different nursing action constraints, and maintain the time order relationship between each data interval. S303. Embed state identifier vectors corresponding to the degree of constraint difference in each data interval, map the degree of constraint difference at each time position into a multi-dimensional numerical vector, and write it into the semantically enhanced force data in the corresponding data interval in chronological order, so that each data interval contains a state identifier vector corresponding to the degree of constraint difference.

[0012] Preferably, S302 is as follows: The semantic enhancement force data sequence is reconstructed according to the segment boundaries. The semantic enhancement force data sequence is segmented according to the time index corresponding to the segment boundaries. Continuous semantic enhancement force data between each segment boundary is extracted to form multiple initial data segments. The semantic enhancement force data within the same segment interval are merged in chronological order. The semantic enhancement force data in each initial data segment are reordered and continuously spliced ​​according to the time index to form multiple time-continuous data intervals defined by segment boundaries, and each data interval corresponds to a single nursing action constraint. The data intervals corresponding to different nursing action constraints are arranged according to the time order of the segment boundaries. The start time position and end time position of each data interval are sequentially associated to establish the time order relationship between the data intervals.

[0013] Preferably, S4 specifically includes the following steps: S401. Perform joint calculation on each data interval and status identifier vector, extract semantically enhanced force data and corresponding status identifier vector in each data interval in chronological order, perform intra-interval numerical aggregation processing on semantically enhanced force data to obtain force change, and perform intra-interval vector aggregation processing on status identifier vector to obtain constraint change, and use force change and constraint change as force adaptation parameter and nursing constraint consistency parameter, respectively. S402. The stress adaptation parameters and nursing constraint consistency parameters are weighted and superimposed to generate evaluation coefficients corresponding to each data interval, and the one-to-one correspondence between the evaluation coefficients and the data intervals is maintained. S403. The evaluation coefficient is compared with the preset interval one by one. The evaluation coefficient is determined by the upper and lower boundary values ​​of multiple preset intervals. The numerical interval to which the evaluation coefficient belongs is determined. The corresponding data interval is classified according to the numerical interval to which the evaluation coefficient belongs. Each data interval is divided into different categories and each data interval is assigned a classification label corresponding to the numerical interval to which it belongs.

[0014] Preferably, S401 specifically refers to: The semantically enhanced force data and corresponding state identifier vectors of each data interval are jointly calculated. The semantically enhanced force data and corresponding state identifier vectors of each data interval are extracted in chronological order. The semantically enhanced force data are traversed according to the time index and the force values ​​at each time position are extracted. The force values ​​at each time position are accumulated and averaged within the interval to obtain the force change. Based on the obtained force change, the state identifier vectors in each data interval are traversed in chronological order. The state identifier vectors at each time position are accumulated and normalized one dimension at a time position to form a vector aggregation result within the interval. This vector aggregation result is used as the constraint change. The changes in force and the changes in constraints are used as force adaptation parameters and nursing constraint consistency parameters, respectively. A correspondence between the changes in force and the changes in constraints is established so that each data interval corresponds to a set of force adaptation parameters and nursing constraint consistency parameters.

[0015] Preferably, S5 is as follows: In response to the classification results, the organization of the input data corresponding to the data intervals is adjusted. The data intervals are grouped according to the classification labels of each data interval, and the semantic enhancement data in each data interval is rearranged according to the order corresponding to the classification labels. The data intervals with the same classification label are continuously spliced ​​according to the time index. At the same time, the data connection relationship between data intervals with different classification labels is re-established to form a classification-driven input data organization structure. Based on the adjustment of the input data organization, the calculation participation relationship of the corresponding data intervals is adjusted. The corresponding participation coefficients are assigned to each data interval according to the classification identifier, and the participation coefficients are written into the semantically enhanced force data in the corresponding data intervals. By uniformly integrating the participation coefficients of each data interval, updated force modeling input data is generated.

[0016] The technical effects and advantages provided by the present invention in the above technical solution are as follows: 1. This invention introduces a multidimensional mapping expression of nursing action constraint information into the force adaptation modeling process of postoperative sports medicine nursing, and constructs a semantic difference evolution sequence by combining difference calculation, thereby achieving accurate identification and quantitative description of nursing action constraint switching behavior. On this basis, by analyzing the degree of constraint difference and driving the force modeling input data to perform segmented isolation or state identification processing, the force data under different nursing constraint conditions that originally existed mixed in the same modeling process are effectively distinguished, avoiding the problem of force data aliasing in the modeling process. At the same time, by embedding state identification vectors corresponding to the degree of constraint difference in the data interval, the force data has clear constraint semantic attributes, so that subsequent modeling calculations can accurately identify the data source under different nursing action constraint states, improving the pertinence and interpretability of the force modeling results.

[0017] 2. This invention extracts force adaptation parameters and nursing constraint consistency parameters by jointly calculating various data intervals and state identifier vectors, and generates evaluation coefficients through combined calculations. This enables a quantitative assessment of the matching degree between the force state and nursing constraints in different data intervals. Simultaneously, the data intervals are classified based on the evaluation coefficients, and the organization and calculation participation relationships of the input data are adjusted in response to the classification results. This results in differences in the participation of different categories of data in subsequent modeling calculations, thereby constructing a dynamically adjustable force modeling input data structure. This allows the modeling process to adaptively adjust according to changes in nursing action constraints, further improving modeling accuracy and adaptability, and enhancing the stability and applicability of the overall modeling process in complex nursing constraint switching scenarios. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0019] Figure 1 This is a schematic diagram of the process of the present invention. Detailed Implementation

[0020] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, they are provided so that the description of this disclosure will be more complete and fully convey the concept of the exemplary embodiments to those skilled in the art.

[0021] This invention provides, for example Figure 1 The method for force adaptation modeling in sports medicine nursing, as shown, specifically includes the following steps: S1. Perform multidimensional mapping processing on the nursing action constraint information in the postoperative sports medicine nursing training sequence to generate a nursing constraint sequence, and bind it with the corresponding force data sequence according to the time correspondence to form a semantically enhanced force data sequence. In this embodiment, S1 specifically includes the following steps: Multidimensional mapping processing is performed on the nursing action constraint information in the postoperative sports medicine nursing training sequence. The nursing action constraint information is split according to the constraint type dimension, constraint direction dimension, and constraint intensity dimension, and numerical mapping is performed on each dimension to generate multidimensional mapping results. The multidimensional mapping results are then arranged in chronological order to generate a nursing constraint sequence. In the implementation process, the timeline of postoperative sports medicine nursing training sequences can be analyzed first to extract the nursing action constraint information corresponding to each moment. For example, for joint mobility restrictions, weight-bearing restrictions, or movement control requirements, each type of constraint can be decomposed into structured data items. Subsequently, mapping can be performed around the constraint type dimension, constraint direction dimension, and constraint intensity dimension. The constraint type can be converted into discrete numerical identifiers through category coding, the constraint direction can be converted into direction vector coding through spatial direction identifiers, and the constraint intensity can be converted into continuous or graded values ​​through interval division or level division, thus forming a multi-dimensional numerical combination. The values ​​of each dimension at the same moment are combined into a multi-dimensional mapping result and arranged sequentially according to the time order of the training sequence to generate a nursing constraint sequence. For example, in a leg raise training process, the limitation of knee flexion range can be encoded as a type identifier, the limitation direction can be corresponding to the flexion-extension direction coding, and the limitation intensity can be represented as the limitation level. The combination of these three forms a multi-dimensional mapping result at that moment, which forms a continuous sequence over time. In this way, the originally scattered nursing constraints can be transformed into a continuous data expression that can participate in modeling calculations.

[0022] Postoperative sports medicine nursing training sequence refers to the set of movement behavior and nursing constraint information recorded chronologically during the rehabilitation process; nursing movement constraint information refers to the restrictions imposed on the patient's movement process, including range of motion, directional control, and load limitation; multidimensional mapping processing refers to the process of converting the original constraint information from text or rule form into multidimensional numerical expression; constraint type dimension is used to distinguish different categories of nursing constraints, constraint direction dimension is used to describe the spatial direction attribute of the constraint, and constraint intensity dimension is used to represent the magnitude of the constraint; numerical mapping refers to converting the information of each dimension into a computable numerical expression; multidimensional mapping result refers to the vectorized expression composed of multiple dimension values ​​at the same point in time; the nursing constraint sequence is a continuous data sequence formed by arranging the multidimensional mapping results of all time points in chronological order, used as input for subsequent force modeling.

[0023] The nursing constraint sequence is bound to the corresponding force data sequence according to the time correspondence. By constructing a unified time index, the multidimensional mapping results in the nursing constraint sequence are matched point by point with the force data at the corresponding time position in the force data sequence to form a time-aligned data combination. In the implementation process, the time stamp information corresponding to the nursing constraint sequence and the force data sequence can be obtained separately. Both types of data are then mapped onto the same time reference axis, and a unified time index is constructed to achieve time scale consistency. Specifically, timestamp resampling or time axis discretization can be used to map data with different sampling frequencies to a unified time scale. Then, the multi-dimensional mapping results in the nursing constraint sequence are traversed one by one according to the unified time index. At each time position, the force data within the corresponding time interval in the force data sequence is searched, and the unique corresponding force data is determined through interpolation, nearest neighbor selection, or window aggregation, thereby achieving point-by-point matching and forming a one-to-one corresponding data combination. For example, in rehabilitation training, the nursing constraint sequence is recorded using movement rhythms, while the force data is continuously collected by sensors. By establishing a unified time scale for both, each movement moment is mapped to the nearest force sampling point, ensuring that each set of constraint information corresponds synchronously with the actual force state. This avoids data mismatch problems caused by time misalignment.

[0024] Force data sequence refers to a set of force change data arranged in chronological order, reflecting the changes in the patient's mechanical state during movement; unified time index refers to establishing a unified time reference identifier for data from different sources, enabling various types of data to be aligned on the same time coordinate; force data refers to specific mechanical numerical information collected at a certain time position; point-by-point matching refers to establishing a one-to-one association between data elements in the nursing constraint sequence and the force data sequence according to the time index; time-aligned data combination refers to a data unit formed by combining nursing constraint information and force data at the same time position. This data unit contains both constraint semantics and mechanical information, providing a unified input basis for subsequent modeling and calculation.

[0025] After establishing the time correspondence, the multidimensional mapping results in the nursing constraint sequence are embedded into the corresponding force data. By combining the multidimensional mapping results and the force data, a semantically enhanced force data sequence containing nursing action constraint information and force information is formed.

[0026] After establishing the temporal correspondence between the nursing constraint sequence and the force data sequence, the multidimensional mapping result corresponding to each time position can be used as a constraint feature vector and jointly expressed with the force data at that time position. Specifically, the multidimensional mapping result can be directly attached to the force data to form an extended data unit through vector concatenation, feature expansion, or data structure reorganization. For example, at a certain time point, the multidimensional mapping result corresponding to the nursing constraint contains numerical expressions of the action restriction type, direction of action, and restriction intensity, while the force data contains the joint force value at that moment. The two are combined into a set of multidimensional data in a fixed order, so that each data point simultaneously contains constraint semantics and mechanical information. By performing this combination processing point by point on the entire time series, a continuous semantically enhanced force data sequence is formed, thereby ensuring that subsequent modeling calculations can synchronously perceive the nursing action constraint background when processing force data, avoiding the problem that a single force data cannot reflect the constraint conditions.

[0027] Combined processing refers to integrating data from different sources according to a unified structure, so that multiple data features can be expressed together in the same data unit; nursing action constraint information refers to the constraint feature vector after multi-dimensional mapping, reflecting the type, direction and intensity of action restrictions; force information refers to the mechanical data corresponding to the time position, used to describe the force state during the movement; semantically enhanced force data sequence refers to a sequence composed of multiple data units arranged in time sequence, where each data unit contains both nursing action constraint information and force information, so that the data not only has numerical features, but also has clear constraint semantic attributes, which are used as input for subsequent force modeling.

[0028] S2. Construct a semantic difference evolution sequence using the nursing constraint sequence. Calculate the difference between nursing constraint sequences at adjacent time points to determine whether nursing action constraints have switched. If nursing action constraints have switched, analyze the difference between the nursing constraint sequences before and after the switch to determine the degree of constraint difference and form a difference representation sequence. In this embodiment, S2 specifically includes the following steps: S201. Pair the multidimensional mapping results arranged in time order in the nursing constraint sequence with adjacent time points, perform dimension-wise difference calculation on each pair of adjacent time points to generate the corresponding difference vector, and arrange each difference vector in time order to form a semantic difference evolution sequence. In the implementation process, the multidimensional mapping results in the nursing constraint sequence can be traversed in chronological order. The multidimensional mapping results of adjacent time points are paired one by one, and a dimensional difference calculation is performed on each pair. This involves calculating the corresponding difference between the values ​​of the same dimension at one time point and the next, thus obtaining a set of difference values ​​reflecting the changes in each dimension. This set is then combined into a difference vector in a fixed dimensional order. This process is then repeated along the time axis for all adjacent time points, and the resulting difference vectors are arranged sequentially in chronological order to form a continuous semantic difference evolution sequence. For example, at two consecutive time points, the nursing constraints remain consistent in the type dimension but change in the intensity dimension. Through dimensional difference calculation, it can be seen that this change is concentrated only in the intensity dimension, thus being reflected in the difference vector as a change in the value of a specific dimension. This accurately depicts the trajectory of nursing constraints over time.

[0029] The multidimensional mapping results arranged chronologically in the nursing constraint sequence refer to the multidimensional numerical expression corresponding to each time point, forming a sequence structure according to time sequence; adjacent time pairing refers to selecting data from two adjacent time positions in the time series for pairing; dimensional difference calculation refers to calculating the numerical difference for each dimension in the multidimensional mapping results to obtain the changes in each dimension; difference vector refers to the vectorized expression composed of the difference values ​​of each dimension, used to describe the overall change state between adjacent time points; semantic difference evolution sequence refers to the sequence structure formed by arranging multiple difference vectors in chronological order, used to reflect the changing trend and evolution path of nursing constraints throughout the entire time process.

[0030] S202. Perform dimension-by-dimensional analysis on the difference vectors in the semantic difference evolution sequence. Calculate the change in the difference vectors in each dimension and compare the change with the preset benchmark change in the corresponding dimension. When the change in any dimension is greater than the preset benchmark change in the corresponding dimension, determine the nursing action constraint switch at the corresponding time position and record the switch position. In the implementation process, each difference vector in the semantic difference evolution sequence can be parsed dimension by dimension, breaking down the difference vector into change data corresponding to multiple dimensions, and obtaining the corresponding change for each dimension. Then, a baseline change is pre-set for each dimension. This baseline change can be obtained statistically based on the change amplitude of nursing constraints under historical stable conditions, or by aggregating and analyzing the change range of difference vectors over a continuous time period. During the specific calculation, the change in each dimension at the current time position is compared one by one with the baseline change for the corresponding dimension. When the change in a certain dimension exceeds the baseline change for the corresponding dimension, it is determined that a nursing action constraint switch has occurred at that time position, and this time position is recorded as the switch position. For example, in a continuous time point, if the nursing constraint remains unchanged in the type dimension but shows a significant jump in the intensity dimension, by comparing the change in the intensity dimension with the baseline change, this time point can be identified as the constraint switch position, thus accurately capturing the moment of constraint change.

[0031] Dimensional analysis refers to the process of splitting the difference vector according to each dimension and processing them separately, so that the changes in each dimension can be independently identified; the change in the difference vector in each dimension refers to the numerical difference in each dimension of the multidimensional mapping results at adjacent time points, used to describe the magnitude of constraint change; the preset benchmark change in the corresponding dimension refers to the preset reference range of change for each dimension, used as a benchmark to determine whether the change is significant; nursing action constraint switching refers to the process of nursing action constraints changing from one constraint state to another in the time series; the switching position refers to the specific time point in the time series when a change in nursing action constraints is detected, used to identify the time position where the constraint change occurs.

[0032] S203. When nursing action constraints change, select the multidimensional mapping results in the nursing constraint sequence before and after the change position for dimension-by-dimensional comparison calculation, combine and express the differences in each dimension to determine the degree of constraint difference, and arrange the degree of constraint difference in time order to generate a difference representation sequence.

[0033] In this embodiment, S203 specifically refers to: When nursing action constraints change, the multidimensional mapping results in the nursing constraint sequence within the corresponding time interval are extracted according to the switching position. The sets of multidimensional mapping results for several consecutive moments before the switching position and the sets of multidimensional mapping results for several consecutive moments after the switching position are obtained respectively, and the correspondence between the two is established in chronological order. Once a nursing action constraint switch is determined at a specific time point through difference calculation, a bidirectional expansion of the data range can be performed based on the time index of that switch point in the nursing constraint sequence. Specifically, this can be achieved using a fixed-length window or by extracting continuous intervals based on time spans. This extracts a set of multidimensional mapping results for several consecutive moments before and after the switch point from the nursing constraint sequence, and arranges them chronologically. Then, using the switch point as a time boundary, the last moment in the set before the switch is paired with the first moment in the set after the switch, and the results are processed sequentially according to time distance. The hierarchical correspondence establishes a correspondence between the preceding and following data points, ensuring consistency in the temporal structure of each pair of data. For example, the three most recent moments before the switch and the three earliest moments after the switch are matched one-to-one in chronological order to form multiple pairs of data. This ensures that subsequent dimension-by-dimensional comparison calculations are performed on the same time scale. The set of multidimensional mapping results for several consecutive moments before the switch position represents the distribution of constraint states before the switch, and the set of multidimensional mapping results for several consecutive moments after the switch position represents the distribution of constraint states after the switch. The preceding and following correspondence represents the pairwise mapping relationship established under a unified time series structure, which is used to support the comparability of subsequent difference analysis.

[0034] The multidimensional mapping results before and after the position switch are compared and calculated dimension by dimension. The difference between the values ​​of the same dimension before and after the switch is calculated to obtain the difference of each dimension. The difference is arranged in the order of the dimensions to form a set of dimension differences. The set of dimension differences is weighted and combined to generate the degree of constraint difference of the corresponding time position. After establishing the correspondence between the multidimensional mapping results before and after the position switch, a dimensional comparison calculation is performed on the corresponding data before and after each set. Specifically, for each dimension, the values ​​at the corresponding time before and after the switch are extracted, and the differences in the values ​​of the same dimension before and after the switch are calculated to obtain the differences in each dimension. Then, all the differences in each dimension are arranged according to a preset dimensional order to form a set of dimensional differences. Dimensional weights are introduced into this set of differences for weighted combination expression. Different dimensions can be assigned different weights based on the importance of the constraint type, directional sensitivity, or intensity of influence. By weighting and accumulating the differences in each dimension, a single numerical expression is generated as the degree of constraint difference at the corresponding time position. For example, in nursing... When constraints change, the type dimension changes relatively little while the intensity dimension changes significantly. By using weighted combination, the difference in the intensity dimension can account for a higher proportion in the overall expression, thus accurately reflecting the actual degree of constraint change. The dimension-by-dimensional comparison calculation refers to performing numerical comparison processing on each dimension of the multidimensional mapping result. The value of the same dimension before and after the switch represents the value of the same constraint attribute under different time states. The difference calculation is used to obtain the change magnitude of the dimension. The difference of each dimension is expressed as a set of differences of each dimension. The set of dimension differences is a combination of difference data organized according to the dimension order. The weighted combination expression refers to the process of comprehensively processing the differences of different dimensions according to weights. The degree of constraint difference is a unified numerical expression reflecting the overall constraint change magnitude.

[0035] The degree of constraint difference calculated at each time point is arranged in chronological order, and the degree of constraint difference is continuously organized to form a difference characterization sequence consistent with the time axis of the nursing constraint sequence.

[0036] In the implementation process, the constraint difference degree calculated at each time position is arranged according to the original time sequence, so that each constraint difference degree corresponds one-to-one with the corresponding time position. Then, the sequence of constraint difference degree is subjected to continuous organization processing. Specifically, time interpolation, neighborhood smoothing, or interval splicing can be used to transition the difference degree between adjacent time positions, so that the numerical changes maintain continuous change characteristics in the time dimension. For example, when there is an abrupt change in the constraint difference degree at a certain time point compared with the time points before and after, weighted smoothing can be performed on adjacent time points to make the value at that time point form a gradual trend in the time series, thereby avoiding the discontinuity problem caused by discrete jumps. Finally, a difference representation sequence consistent with the time axis of the nursing constraint sequence is formed for subsequent modeling calculations. Among them, continuous organization processing refers to the structural reorganization and numerical transition processing of the difference degree in the time series to maintain a continuous change relationship between adjacent time points. The difference representation sequence consistent with the time axis of the nursing constraint sequence refers to the sequence structure in which the difference degree is strictly aligned with the nursing constraint sequence in time order and changes synchronously, which is used to express the evolution of nursing constraint changes in the time dimension.

[0037] S3. Based on the difference representation sequence, the force modeling input data is segmented and isolated or processed according to the degree of constraint difference. The semantically enhanced force data sequence is reconstructed into multiple data intervals corresponding to different nursing action constraints, and a state label vector corresponding to the degree of constraint difference is embedded in each data interval. In this embodiment, S3 specifically includes the following steps: S301. Based on the difference characterization sequence, extract the degree of constraint difference corresponding to each time position in chronological order, compare the degree of constraint difference between adjacent time positions numerically, and when the difference between the degree of constraint difference between adjacent time positions is greater than the preset difference threshold, determine the corresponding time position as the segment boundary, and perform segment isolation processing on the force modeling input data based on the segment boundary to form multiple time-continuous data segments. In the implementation process, the degree of constraint difference at each time position can be extracted according to the difference representation sequence in chronological order, and the degree of constraint difference at adjacent time positions can be numerically compared. Specifically, the time series can be traversed point by point, and the difference between the degree of constraint difference at the current time position and the degree of constraint difference at the previous time position can be calculated to obtain the change amplitude at adjacent time positions. Then, the change amplitude is compared with a pre-set difference threshold. When the change amplitude exceeds the threshold, the current time position is marked as a segment boundary, and the force modeling input data is segmented and isolated based on the segment boundary, thereby dividing the original continuous data into multiple time-continuous data segments. For example, in the continuous training process, if the degree of constraint difference changes significantly before and after a certain moment, a segment boundary is formed at that moment, so that the data before and after that moment are assigned to different data segments, thereby avoiding the mixing of constraint data of different nursing actions in the same modeling interval.

[0038] Numerical comparison refers to the process of comparing the degree of constraint difference between two time points to determine the magnitude of change; the difference between the degree of constraint difference between adjacent time points refers to the amount of change in the degree of constraint difference between two consecutive time points to reflect the strength of constraint change; the preset difference threshold refers to the reference value set before processing to serve as the basis for determining whether to divide the segment boundary; the segment boundary refers to the time position used to divide different data intervals in the time series; the force modeling input data refers to the semantically enhanced force data participating in the force modeling calculation; the segment isolation processing refers to splitting continuous data according to the segment boundary to make the data in different intervals independent of each other; multiple time-continuous data segments refer to multiple time-continuous and separate data sets obtained by segment boundary division, which are used for subsequent modeling processing.

[0039] S302. Reconstruct the semantic enhancement force data sequence according to the segment boundaries, merge the semantic enhancement force data in the same segment interval according to the time order to form multiple data intervals corresponding to different nursing action constraints, and maintain the time order relationship between each data interval. S303. Embed state identifier vectors corresponding to the degree of constraint difference in each data interval, map the degree of constraint difference at each time position into a multi-dimensional numerical vector, and write it into the semantically enhanced force data in the corresponding data interval in chronological order, so that each data interval contains a state identifier vector corresponding to the degree of constraint difference.

[0040] The degree of constraint difference at each time point can be converted into a multi-dimensional numerical vector according to a preset mapping relationship. Specifically, this can be achieved by dividing the degree of constraint difference into intervals, encoding levels, or mapping proportions, transforming a single numerical value into a vector expression containing multiple dimensions. For example, the degree of constraint difference can be divided into different intervals and assigned corresponding vector codes, or it can be decomposed into amplitude components and trend components to form a multi-dimensional expression. Then, each data interval is traversed in chronological order, and at each time point, the corresponding multi-dimensional numerical vector is written into the semantically enhanced force data in the corresponding data interval. The state identifier vector is embedded into the original data structure through vector concatenation or feature expansion, so that the data at each time point simultaneously contains force information and constraint difference information. For example, at a certain time point, the degree of difference is mapped into a multi-dimensional identifier vector and appended to the back end of the corresponding force data, thereby forming a unified data expression and ensuring that different constraint change states can be distinguished and represented in the data.

[0041] The state identifier vector corresponding to the degree of constraint difference refers to the multidimensional numerical expression obtained by mapping the degree of constraint difference, which is used to characterize the state of constraint change; the multidimensional numerical vector refers to a data structure composed of multiple numerical dimensions, each dimension being used to express different attributes of constraint difference; the semantically enhanced force data in the corresponding data interval refers to a data set that simultaneously contains nursing action constraint information and force information within a specific time interval; embedding refers to writing the state identifier vector into the data structure of the semantically enhanced force data, so that the two types of information coexist in the same data unit, which is used for subsequent modeling calculations to identify and utilize the state of constraint difference.

[0042] In this embodiment, S302 specifically refers to: The semantic enhancement force data sequence is reconstructed according to the segment boundaries. The semantic enhancement force data sequence is segmented according to the time index corresponding to the segment boundaries. Continuous semantic enhancement force data between each segment boundary is extracted to form multiple initial data segments. The semantically enhanced data sequence can be restructured based on the time index corresponding to the segment boundaries. By locating the segment boundaries on the time axis, the semantically enhanced data sequence is divided into multiple continuous intervals. Specifically, a continuous data segment is extracted from the starting position to the first segment boundary in chronological order, and then corresponding data intervals are extracted sequentially based on the time range between adjacent segment boundaries until the complete sequence is covered, thus forming multiple independent and temporally continuous data intervals. For example, when there are multiple segment boundaries in the time series, the original semantically enhanced data can be divided into several continuous data segments, each corresponding to a stable nursing action constraint interval, providing a basis for subsequent processing. Reconstruction refers to the process of reorganizing the original data sequence according to the segment boundaries to change the data structure. Segmentation refers to the operation of dividing continuous data into multiple intervals based on the time index. The continuous semantically enhanced data between each segment boundary refers to the data set arranged in chronological order between two adjacent segment boundaries. The initial data segment refers to the basic data interval unit obtained through segmentation, which is used for subsequent merging processing.

[0043] The semantic enhancement force data within the same segment interval are merged in chronological order. The semantic enhancement force data in each initial data segment are reordered and continuously spliced ​​according to the time index to form multiple time-continuous data intervals defined by segment boundaries, and each data interval corresponds to a single nursing action constraint. The semantic enhancement data within the same segment interval can be traversed according to the time index. After extracting the semantic enhancement data from each initial data segment, it is uniformly sorted according to the time index to ensure that the data is monotonically increasing on the time axis. Then, it is continuously spliced ​​according to the sorting results to connect the data at adjacent time positions to form a continuous data stream, thus forming a complete data interval. For example, when there are time index intersections or local disorder in multiple initial data segments, they are adjusted to a unified time order through sorting, and then spliced ​​to form an uninterrupted data sequence. This ensures that each data interval is within the time range defined by the segment boundaries and corresponds to a single nursing action constraint. Here, merging refers to the aggregation of data belonging to the same segment interval, re-sorting and continuous splicing refers to adjusting the order of data according to the time index and connecting them by time to form a continuous sequence. Multiple time-continuous data intervals defined by segment boundaries refer to a set of data that is time-continuous between adjacent segment boundaries and whose interval range is determined by the segment boundaries. A single nursing action constraint means that the nursing action constraint remains consistent and does not switch within the data interval.

[0044] The data intervals corresponding to different nursing action constraints are arranged according to the time order of the segment boundaries. The start time position and end time position of each data interval are sequentially associated to establish the time order relationship between the data intervals.

[0045] Multiple data intervals corresponding to different nursing action constraints can be sorted according to the sequential position of their segment boundaries on the time axis. By extracting the start and end times of each data interval, and using the start time as the primary sorting criterion, all data intervals are sorted. Then, adjacent data intervals are sequentially associated, establishing a connection between the end time of the previous data interval and the start time of the next data interval, thus forming a continuous time chain structure. For example, if the end time of one data interval is earlier than the start time of another, the two are arranged in chronological order and a sequential connection is established, so that all data intervals form a complete and orderly arrangement on the time axis. The start and end times of each data interval refer to the start and end indices of each data interval in the time series, used to define the time range of the data interval. The chronological order refers to the structure of arranging multiple data intervals according to the time index and establishing a sequential connection, used to ensure the continuity and traceability of the data intervals in the time dimension.

[0046] S4. Perform joint calculations on each data interval and status identifier vector, extract the force adaptation parameters and nursing constraint consistency parameters, generate evaluation coefficients through combined calculations, and classify the data intervals according to preset intervals. In this embodiment, S4 specifically includes the following steps: S401. Perform joint calculation on each data interval and status identifier vector, extract semantically enhanced force data and corresponding status identifier vector in each data interval in chronological order, perform intra-interval numerical aggregation processing on semantically enhanced force data to obtain force change, and perform intra-interval vector aggregation processing on status identifier vector to obtain constraint change, and use force change and constraint change as force adaptation parameter and nursing constraint consistency parameter, respectively. S402. The stress adaptation parameters and nursing constraint consistency parameters are weighted and superimposed to generate evaluation coefficients corresponding to each data interval, and the one-to-one correspondence between the evaluation coefficients and the data intervals is maintained. The stress adaptation parameters and nursing constraint consistency parameters corresponding to each data interval can be jointly calculated and processed. By setting weight coefficients for the stress adaptation parameters and nursing constraint consistency parameters respectively, and performing weighted superposition calculation according to the parameter pairing relationship within the same data interval, a single value is generated as the evaluation coefficient. Specifically, the stress adaptation parameter can be multiplied by a first weight value, and the nursing constraint consistency parameter can be multiplied by a second weight value, and then the values ​​are superimposed to obtain an evaluation coefficient that comprehensively reflects the degree of matching between the stress state and the nursing constraint. For example, in a certain data interval, the stress adaptation parameter reflects the trend of stress change, and the nursing constraint consistency parameter reflects the stability of the constraint. Through weighted superposition, the two can be unified under the same evaluation scale, and the generated evaluation coefficient is bound to the corresponding data interval, so that each data interval corresponds to a unique evaluation coefficient. The weighted superposition process refers to the process of combining the values ​​of different parameters according to preset weights. The evaluation coefficient corresponding to each data interval refers to the unique comprehensive evaluation value generated within each data interval. The one-to-one correspondence between the evaluation coefficient and the data interval means that each data interval corresponds to only one evaluation coefficient, and the evaluation coefficient corresponds to only one data interval. This mapping relationship is used for subsequent classification and control processing.

[0047] S403. The evaluation coefficient is compared with the preset interval one by one. The evaluation coefficient is determined by the upper and lower boundary values ​​of multiple preset intervals. The numerical interval to which the evaluation coefficient belongs is determined. The corresponding data interval is classified according to the numerical interval to which the evaluation coefficient belongs. Each data interval is divided into different categories and each data interval is assigned a classification label corresponding to the numerical interval to which it belongs.

[0048] After obtaining the evaluation coefficients corresponding to each data interval, the evaluation coefficients are compared one by one according to multiple pre-divided numerical intervals. Specifically, multiple continuous and non-overlapping numerical interval ranges are set, each numerical interval is limited by upper and lower boundary values. Then, each evaluation coefficient is compared with the upper and lower boundary values ​​of each numerical interval in turn. When the value of the evaluation coefficient is greater than the lower boundary value of a certain interval and less than the corresponding upper boundary value, it is determined that the evaluation coefficient falls into that numerical interval, and the corresponding data interval is classified into the category corresponding to that interval. For example, when the evaluation coefficient is in a lower range, it is classified into one category, when it is in a middle range, it is classified into another category, and when it is in a higher range, it is classified into a third category. This enables the differentiation of data intervals with different force and nursing constraint matching states and provides a classification basis for subsequent processing.

[0049] Preset intervals refer to multiple numerical ranges defined before processing based on evaluation requirements, each numerical range being defined by a set of upper and lower boundary values; sequential comparison refers to the process of comparing each evaluation coefficient with each numerical interval in turn; the upper and lower boundary values ​​of multiple preset intervals refer to the minimum and maximum values ​​used to define the range of each numerical interval; interval determination refers to the process of determining the interval to which an evaluation coefficient belongs based on the numerical relationship between the evaluation coefficient and the upper and lower boundary values; the numerical interval to which an evaluation coefficient belongs refers to the specific interval position in which the evaluation coefficient falls within the numerical range; the classification label corresponding to the numerical interval to which it belongs refers to the category label pre-set for each numerical interval, used to mark and distinguish each data interval after classification.

[0050] In this embodiment, S401 specifically refers to: The semantically enhanced force data and corresponding state identifier vectors of each data interval are jointly calculated. The semantically enhanced force data and corresponding state identifier vectors of each data interval are extracted in chronological order. The semantically enhanced force data are traversed according to the time index and the force values ​​at each time position are extracted. The force values ​​at each time position are accumulated and averaged within the interval to obtain the force change. The semantically enhanced force data and corresponding state identifier vectors in each data interval can be jointly processed. By traversing the semantically enhanced force data in each data interval in chronological order, the force value at each time position is extracted one by one. The force values ​​at all time positions within the same interval are accumulated, and the accumulated result is divided by the number of time positions within the data interval to obtain the average value. This forms the force change that reflects the overall force level of the data interval. For example, if there are force values ​​at multiple time points within a certain data interval, by accumulating these values ​​and calculating the average, the comprehensive force characteristics of the interval can be obtained. The characteristics are used for subsequent joint calculations with the status identifier vector; where each data interval and status identifier vector refers to the data set divided according to nursing action constraints and the corresponding constraint difference expression; joint calculation refers to the process of simultaneously processing force data and status identifier vector within the same data interval; the force value at each time position refers to the mechanical numerical information collected under each time index; the interval accumulation and averaging calculation refers to the process of summing and averaging multiple force values ​​within the same interval; the force change refers to the numerical result obtained by accumulation and averaging to characterize the overall force change characteristics of the data interval.

[0051] Based on the obtained force change, the state identifier vectors in each data interval are traversed in chronological order. The state identifier vectors at each time position are accumulated and normalized one dimension at a time position to form a vector aggregation result within the interval. This vector aggregation result is used as the constraint change. After obtaining the force change, the state identifier vectors in each data interval are traversed in chronological order. The state identifier vector at each time position is split into multiple numerical components according to dimensions, and the values ​​of the same dimension are accumulated dimension by dimension throughout the entire data interval to obtain the cumulative value of each dimension. Then, the cumulative value is normalized so that the values ​​of different dimensions are expressed within a unified scale range, and finally, the vector aggregation result within the interval is formed. For example, in a certain data interval, each time position corresponds to a multi-dimensional state identifier vector. By accumulating each dimension and unifying the scale, a vector expression reflecting the overall constraint change characteristics of the interval can be obtained, and this vector is used as the constraint change quantity in subsequent calculations. The dimension-by-dimensional numerical accumulation and normalization process refers to the process of accumulating the values ​​of the state identifier vector in each dimension and unifying the scale of the result. The vector aggregation result within the interval refers to the multi-dimensional vector expression formed by summarizing all state identifier vectors within a data interval. The constraint change quantity refers to the vectorized numerical result used to characterize the constraint change characteristics of nursing actions within the data interval.

[0052] The changes in force and the changes in constraints are used as force adaptation parameters and nursing constraint consistency parameters, respectively. A correspondence between the changes in force and the changes in constraints is established so that each data interval corresponds to a set of force adaptation parameters and nursing constraint consistency parameters.

[0053] After obtaining the force and constraint changes, the two types of data are processed to establish a correspondence based on the data interval division results. The force changes calculated within the same data interval are paired with the constraint changes formed within the corresponding time interval. Using the data interval as the association unit, the force changes are identified as force adaptation parameters, and the constraint changes are identified as nursing constraint consistency parameters. This ensures that each data interval corresponds to a set of parameter combinations. For example, within a certain data interval, the overall force change is calculated from the force data, and the constraint change is obtained by aggregating the state identifier vector. Pairing the two together forms the parameter expression for that interval, which is used for subsequent evaluation calculations. The force adaptation parameter refers to the numerical parameter transformed from the force change to characterize the force state and the degree of motion adaptation, while the nursing constraint consistency parameter refers to the vector parameter transformed from the constraint change to characterize the stability and consistency of nursing action constraints. Establishing a correspondence refers to the process of matching the two types of parameters within the same data interval one by one and establishing an association structure, so that each data interval has a complete parameter expression for subsequent calculations.

[0054] S5. In response to the classification results, adjust the organization and calculation participation relationships of the input data corresponding to the data intervals to generate updated force modeling input data.

[0055] In this embodiment, S5 specifically refers to: In response to the classification results, the organization of the input data corresponding to the data intervals is adjusted. The data intervals are grouped according to the classification labels of each data interval, and the semantic enhancement data in each data interval is rearranged according to the order corresponding to the classification labels. The data intervals with the same classification label are continuously spliced ​​according to the time index. At the same time, the data connection relationship between data intervals with different classification labels is re-established to form a classification-driven input data organization structure. After obtaining the classification results, the classification labels corresponding to each data interval are parsed, and the data intervals are grouped according to the classification labels, with data intervals with the same classification labels grouped into the same data set. Then, according to the order of the classification labels, the semantic enhancement data in each data set is rearranged. Specifically, the time index of each data interval is extracted and reordered so that data of the same category forms a continuous structure on the time axis. Then, data intervals with the same classification label are continuously spliced ​​according to the time index to form multiple internally continuous data blocks. At the same time, for data sets with different classification labels, new data connection relationships are established by adjusting the connection order between data blocks. For example, data blocks of one category are prioritized and connected to data blocks of another category in a predetermined order, thereby forming a new input data organization structure, so that the data arrangement order is consistent with the classification results, providing structured input for subsequent modeling calculations.

[0056] The input data organization relationship corresponding to the data intervals refers to the arrangement and connection relationship of each data interval in the overall data sequence; the classification label refers to the category label obtained according to the evaluation coefficient, which is used to distinguish different types of data intervals; grouping refers to the process of classifying data intervals according to the classification label; rearrangement refers to the process of rearranging the semantically enhanced force data in the data intervals according to the time index or classification order; continuous splicing refers to connecting multiple time-adjacent data intervals in sequence to form a continuous data structure; data connection relationship refers to the temporal or logical connection method between different data intervals; the classification-driven input data organization structure refers to the data arrangement structure formed after reorganizing the data according to the classification label, which is used to support subsequent force modeling calculations.

[0057] Based on the adjustment of the input data organization, the calculation participation relationship of the corresponding data intervals is adjusted. The corresponding participation coefficients are assigned to each data interval according to the classification identifier, and the participation coefficients are written into the semantically enhanced force data in the corresponding data intervals. By uniformly integrating the participation coefficients of each data interval, updated force modeling input data is generated.

[0058] After adjusting the organization of the input data, the classification labels corresponding to each data interval are parsed, and participation coefficients are assigned to each data interval based on the classification labels. Specifically, the classification labels of each data interval are mapped to the corresponding participation coefficients by pre-setting the numerical weights corresponding to different classification labels. Then, each data interval is traversed in chronological order, and the participation coefficients are written into the semantically enhanced force data in the corresponding data interval. The participation coefficients are embedded into the data structure through vector concatenation or field expansion. Subsequently, the participation coefficients in all data intervals are uniformly integrated. For example, the data with embedded participation coefficients are rearranged in chronological order to form a complete data sequence, thereby generating updated force modeling input data. For example, under different classification categories, higher or lower participation coefficients can be assigned to make the participation degree of different categories of data different in subsequent modeling calculations.

[0059] The calculation participation relationship refers to the degree of participation and influence weight of each data interval in the modeling calculation process; the participation coefficient refers to the numerical weight assigned to each data interval according to the classification label, which is used to adjust its contribution ratio in the calculation; writing refers to embedding the participation coefficient into the data structure of the semantically enhanced force data, making it part of the data; unified integration processing refers to the process of organizing and arranging the data containing the participation coefficient in all data intervals as a whole; the updated force modeling input data refers to the new data sequence formed after the participation coefficient embedding and integration are completed. This data sequence contains force information, nursing constraint information and calculation participation relationship information, which are used for subsequent force modeling calculations.

[0060] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. A computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions according to the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. Computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired or wireless means (e.g., infrared, wireless, microwave, etc.). A computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. Available media can be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., DVDs), or semiconductor media. Semiconductor media can be solid-state drives.

[0061] It should be understood that in the various embodiments of this application, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0062] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0063] In the several embodiments provided in this application, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection through some interfaces, devices, or units, and may be electrical, mechanical, or other forms.

[0064] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0065] In addition, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0066] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for modeling stress adaptation in sports medicine nursing, characterized in that, Specifically, the following steps are included: S1. Perform multidimensional mapping processing on the nursing action constraint information in the postoperative sports medicine nursing training sequence to generate a nursing constraint sequence, and bind it with the corresponding force data sequence according to the time correspondence to form a semantically enhanced force data sequence. S2. Construct a semantic difference evolution sequence using the nursing constraint sequence. Calculate the difference between nursing constraint sequences at adjacent time points to determine whether nursing action constraints have switched. If nursing action constraints have switched, analyze the difference between the nursing constraint sequences before and after the switch to determine the degree of constraint difference and form a difference representation sequence. S3. Based on the difference representation sequence, the force modeling input data is segmented and isolated or processed according to the degree of constraint difference. The semantically enhanced force data sequence is reconstructed into multiple data intervals corresponding to different nursing action constraints, and a state label vector corresponding to the degree of constraint difference is embedded in each data interval. S4. Perform joint calculations on each data interval and status identifier vector, extract the force adaptation parameters and nursing constraint consistency parameters, generate evaluation coefficients through combined calculations, and classify the data intervals according to preset intervals. S5. In response to the classification results, adjust the organization and calculation participation relationships of the input data corresponding to the data intervals to generate updated force modeling input data.

2. The method for force adaptation modeling in sports medicine nursing according to claim 1, characterized in that, S1 specifically includes the following steps: Multidimensional mapping processing is performed on the nursing action constraint information in the postoperative sports medicine nursing training sequence. The nursing action constraint information is split according to the constraint type dimension, constraint direction dimension, and constraint intensity dimension, and numerical mapping is performed on each dimension to generate multidimensional mapping results. The multidimensional mapping results are then arranged in chronological order to generate a nursing constraint sequence. The nursing constraint sequence is bound to the corresponding force data sequence according to the time correspondence. By constructing a unified time index, the multidimensional mapping results in the nursing constraint sequence are matched point by point with the force data at the corresponding time position in the force data sequence to form a time-aligned data combination. After establishing the time correspondence, the multidimensional mapping results in the nursing constraint sequence are embedded into the corresponding force data. By combining the multidimensional mapping results and the force data, a semantically enhanced force data sequence containing nursing action constraint information and force information is formed.

3. The method for force adaptation modeling in sports medicine nursing according to claim 1, characterized in that, S2 specifically includes the following steps: S201. Pair the multidimensional mapping results arranged in time order in the nursing constraint sequence with adjacent time points, perform dimension-wise difference calculation on each pair of adjacent time points to generate the corresponding difference vector, and arrange each difference vector in time order to form a semantic difference evolution sequence. S202. Perform dimension-by-dimensional analysis on the difference vectors in the semantic difference evolution sequence. Calculate the change in the difference vectors in each dimension and compare the change with the preset benchmark change in the corresponding dimension. When the change in any dimension is greater than the preset benchmark change in the corresponding dimension, determine the nursing action constraint switch at the corresponding time position and record the switch position. S203. When nursing action constraints change, select the multidimensional mapping results in the nursing constraint sequence before and after the change position for dimension-by-dimensional comparison calculation, combine and express the differences in each dimension to determine the degree of constraint difference, and arrange the degree of constraint difference in time order to generate a difference representation sequence.

4. The method for force adaptation modeling in sports medicine nursing according to claim 3, characterized in that, S203 specifically refers to: When nursing action constraints change, the multidimensional mapping results in the nursing constraint sequence within the corresponding time interval are extracted according to the switching position. The sets of multidimensional mapping results for several consecutive moments before the switching position and the sets of multidimensional mapping results for several consecutive moments after the switching position are obtained respectively, and the correspondence between the two is established in chronological order. The multidimensional mapping results before and after the position switch are compared and calculated dimension by dimension. The difference between the values ​​of the same dimension before and after the switch is calculated to obtain the difference of each dimension. The difference is arranged in the order of the dimensions to form a set of dimension differences. The set of dimension differences is weighted and combined to generate the degree of constraint difference of the corresponding time position. The degree of constraint difference calculated at each time point is arranged in chronological order, and the degree of constraint difference is continuously organized to form a difference characterization sequence consistent with the time axis of the nursing constraint sequence.

5. The method for force adaptation modeling in sports medicine nursing according to claim 1, characterized in that, S3 specifically includes the following steps: S301. Based on the difference characterization sequence, extract the degree of constraint difference corresponding to each time position in chronological order, compare the degree of constraint difference between adjacent time positions numerically, and when the difference between the degree of constraint difference between adjacent time positions is greater than the preset difference threshold, determine the corresponding time position as the segment boundary, and perform segment isolation processing on the force modeling input data based on the segment boundary to form multiple time-continuous data segments. S302. Reconstruct the semantic enhancement force data sequence according to the segment boundaries, merge the semantic enhancement force data in the same segment interval according to the time order to form multiple data intervals corresponding to different nursing action constraints, and maintain the time order relationship between each data interval. S303. Embed state identifier vectors corresponding to the degree of constraint difference in each data interval, map the degree of constraint difference at each time position into a multi-dimensional numerical vector, and write it into the semantically enhanced force data in the corresponding data interval in chronological order, so that each data interval contains a state identifier vector corresponding to the degree of constraint difference.

6. The method for force adaptation modeling in sports medicine nursing according to claim 5, characterized in that, S302 specifically refers to: The semantic enhancement force data sequence is reconstructed according to the segment boundaries. The semantic enhancement force data sequence is segmented according to the time index corresponding to the segment boundaries. Continuous semantic enhancement force data between each segment boundary is extracted to form multiple initial data segments. The semantic enhancement force data within the same segment interval are merged in chronological order. The semantic enhancement force data in each initial data segment are reordered and continuously spliced ​​according to the time index to form multiple time-continuous data intervals defined by segment boundaries, and each data interval corresponds to a single nursing action constraint. The data intervals corresponding to different nursing action constraints are arranged according to the time order of the segment boundaries. The start time position and end time position of each data interval are sequentially associated to establish the time order relationship between the data intervals.

7. The method for force adaptation modeling in sports medicine nursing according to claim 1, characterized in that, S4 specifically includes the following steps: S401. Perform joint calculation on each data interval and status identifier vector, extract semantically enhanced force data and corresponding status identifier vector in each data interval in chronological order, perform intra-interval numerical aggregation processing on semantically enhanced force data to obtain force change, and perform intra-interval vector aggregation processing on status identifier vector to obtain constraint change, and use force change and constraint change as force adaptation parameter and nursing constraint consistency parameter, respectively. S402. The stress adaptation parameters and nursing constraint consistency parameters are weighted and superimposed to generate evaluation coefficients corresponding to each data interval, and the one-to-one correspondence between the evaluation coefficients and the data intervals is maintained. S403. The evaluation coefficient is compared with the preset interval one by one. The evaluation coefficient is determined by the upper and lower boundary values ​​of multiple preset intervals. The numerical interval to which the evaluation coefficient belongs is determined. The corresponding data interval is classified according to the numerical interval to which the evaluation coefficient belongs. Each data interval is divided into different categories and each data interval is assigned a classification label corresponding to the numerical interval to which it belongs.

8. The method for force adaptation modeling in sports medicine nursing according to claim 7, characterized in that, S401 specifically refers to: The semantically enhanced force data and corresponding state identifier vectors of each data interval are jointly calculated. The semantically enhanced force data and corresponding state identifier vectors of each data interval are extracted in chronological order. The semantically enhanced force data are traversed according to the time index and the force values ​​at each time position are extracted. The force values ​​at each time position are accumulated and averaged within the interval to obtain the force change. Based on the obtained force change, the state identifier vectors in each data interval are traversed in chronological order. The state identifier vectors at each time position are accumulated and normalized one dimension at a time position to form a vector aggregation result within the interval. This vector aggregation result is used as the constraint change. The changes in force and the changes in constraints are used as force adaptation parameters and nursing constraint consistency parameters, respectively. A correspondence between the changes in force and the changes in constraints is established so that each data interval corresponds to a set of force adaptation parameters and nursing constraint consistency parameters.

9. The method for force adaptation modeling in sports medicine nursing according to claim 1, characterized in that, S5 specifically refers to: In response to the classification results, the organization of the input data corresponding to the data intervals is adjusted. The data intervals are grouped according to the classification labels of each data interval, and the semantic enhancement data in each data interval is rearranged according to the order corresponding to the classification labels. The data intervals with the same classification label are continuously spliced ​​according to the time index. At the same time, the data connection relationship between data intervals with different classification labels is re-established to form a classification-driven input data organization structure. Based on the adjustment of the input data organization, the calculation participation relationship of the corresponding data intervals is adjusted. The corresponding participation coefficients are assigned to each data interval according to the classification identifier, and the participation coefficients are written into the semantically enhanced force data in the corresponding data intervals. By uniformly integrating the participation coefficients of each data interval, updated force modeling input data is generated.

Images