A folding machine structure reliability evaluation method based on multi-source load fusion

Abstract

The application relates to the field of reliability evaluation of printing machine structures, and discloses a folding machine structure reliability evaluation method based on multi-source load fusion, which comprises the following steps: establishing a global evaluation model of key structural parts of a folding machine; generating a phase condition global response base library under each main shaft phase interval; determining an anchor measuring point set; mapping to the main shaft phase interval to generate phase synchronization data; generating global response base coefficient prior values; obtaining posterior global response base coefficients; generating a global stress field in the global evaluation model; and outputting structure reliability evaluation results according to non-measuring point damage burst discrimination results. Through the establishment of the global evaluation model and the phase condition global response base library, the global response base coefficient prior values are generated through the phase condition load characteristic vectors, and the anchor measuring point response data is used for correction, so that the overall reliability evaluation of the key structural parts of the folding machine is realized.

Description

Technical Field

[0001] This invention relates to the field of printing machinery structural reliability assessment technology, specifically a method for assessing the structural reliability of a folding machine based on multi-source load fusion. Background Technology

[0002] Web-based printing equipment typically includes continuous operation units such as paper feeding, printing, drying, traction, cutting, and folding. The folding machine, used for cutting, guiding, folding, and outputting high-speed paper tape, is a key post-processing device in the web-based printing production line. During high-speed operation, the folding machine's folding knife mechanism, roller support structure, wall panel connection structure, bearing housing mounting structure, and transmission support structure continuously bear loads from paper tape tension, folding knife impact, roller clamping reaction force, spindle torque fluctuations, and frame vibration. Due to the periodicity and phase of the folding action, the timing, location, and transmission path of different loads vary within a single folding cycle. Therefore, the fatigue response of key structural components in the folding machine often exhibits multi-source load coupling and phase correlation characteristics.

[0003] Existing methods for assessing the structural reliability of folding machines typically involve placing strain gauges or vibration sensors at critical locations on key structural components to acquire stress response data from a single or a small number of measuring points. This data is then combined with material fatigue parameters, stress cycle statistics, or finite element analysis results to determine fatigue damage or fatigue life at the measuring points. While these methods can reflect the local stress state in the vicinity of the measuring points, the fatigue damage area of ​​key structural components in folding machines may spatially shift with the folding action under the combined effects of multi-source loads and spindle phase. Stress concentration and fatigue damage accumulation in non-measuring point areas are difficult to characterize using measuring point data, leading to the problem of missed assessment of fatigue damage at non-measuring points in the overall fatigue state assessment of the folding machine structure. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a method for assessing the structural reliability of folding machines based on multi-source load fusion. This method solves the problem that existing methods for assessing the structural reliability of folding machines often result in the omission of fatigue damage at non-measuring points during the overall fatigue state assessment of the folding machine structure.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a method for assessing the structural reliability of a folding machine based on multi-source load fusion, comprising the following steps:

[0006] A global evaluation model for key structural components of a folding machine is established. The global evaluation model includes evaluation nodes, structural regions, candidate measurement point mapping sets, material fatigue parameters, damage thresholds, and principal axis phase intervals.

[0007] Based on the global evaluation model, a global response base library under phase conditions is generated under each of the principal axis phase intervals;

[0008] Based on the candidate measurement point mapping set and the phase condition global response base library, the anchor measurement point set is determined;

[0009] The response data of the anchoring measurement point set and the multi-source load data of the folding machine are collected and mapped to the main shaft phase interval to generate phase synchronization data;

[0010] Based on the phase synchronization data, a phase conditional load feature vector is generated, and prior values ​​of global response basis coefficients are generated.

[0011] The prior values ​​of the global response basis coefficients are corrected using the response data to obtain the posterior global response basis coefficients.

[0012] Based on the phase condition global response basis library and the posterior global response basis coefficients, a global stress field is generated in the global evaluation model.

[0013] Based on the global stress field, a global fatigue damage field and a global reliability field are generated in the global evaluation model, and the structural reliability assessment results are output according to the non-measuring point damage protrusion discrimination results.

[0014] By adopting the above technical solution, and by establishing a global evaluation model and a phase-condition global response base library, the structural response of key structural components of the folding machine can be uniformly expressed according to the evaluation nodes and the main shaft phase. By determining the set of anchoring measurement points and collecting the response data of the anchoring measurement points and the multi-source load data of the folding machine, phase synchronization data is generated, so that the measurement point response and multi-source load information can be mapped to the same folding action stage. The prior value of the global response base coefficient is generated by the phase-condition load feature vector, and corrected by the anchoring measurement point response data to obtain the posterior global response base coefficient. Then, the global stress field, global fatigue damage field, and global reliability field are generated, realizing the overall reliability assessment of the key structural components of the folding machine. This solves the problem that existing folding machine structural reliability assessment methods are prone to missing non-measurement point fatigue damage in the overall fatigue state assessment of the folding machine structure.

[0015] Preferably, the step of establishing a global evaluation model for key structural components of the folding machine includes: acquiring structural geometric data, material parameter data, and connection relationship data of preset structural components of the folding machine; generating an evaluation node set based on the structural geometric data, and configuring node position data for each evaluation node in the evaluation node set; generating a structural region set based on the structural geometric data and the connection relationship data, and establishing a node region mapping set between the evaluation node set and the structural region set; generating a material fatigue parameter set and a damage threshold set based on the material parameter data; generating a main shaft phase interval set based on the main shaft phase signal of the folding machine; generating a candidate measurement point set based on the candidate measurement point position data, and establishing a candidate measurement point mapping set between the candidate measurement point set and the evaluation node set; and configuring the evaluation node set, the structural region set, the node region mapping set, the material fatigue parameter set, the damage threshold set, the main shaft phase interval set, and the candidate measurement point mapping set as the global evaluation model.

[0016] Preferably, generating a phase condition global response base library for each of the main spindle phase intervals includes: acquiring unit load response data for each of the main spindle phase intervals, wherein the unit load response data includes response data corresponding to paper tape tension load, folding knife impact load, roller support reaction force, main spindle torque load, frame vibration foundation excitation, and rotational inertial load; mapping the unit load response data to each evaluation node in the global evaluation model to generate phase global response data; performing principal component extraction processing on the phase global response data to obtain the phase condition global response base matrix corresponding to the main spindle phase interval; and archiving each of the phase condition global response base matrices according to the order of the main spindle phase intervals to form a phase condition global response base library.

[0017] Preferably, determining the set of anchoring measurement points includes: determining a set of candidate measurement points on the preset structural component of the folding machine; generating a candidate measurement point observation matrix based on the positional mapping relationship between the candidate measurement point set and the evaluation nodes in the global evaluation model; associating the candidate measurement point observation matrix with the phase condition global response base library to form reconstruction association data of candidate measurement points to the global fatigue damage field; generating measurement point reconstruction score data for each candidate measurement point based on the reconstruction association data, and forming a measurement point screening sequence according to the measurement point reconstruction score data; and selecting the set of anchoring measurement points from the set of candidate measurement points according to the measurement point screening sequence.

[0018] Preferably, the generation of phase synchronization data includes: collecting strain response data, spindle phase data, spindle speed data, motor current data, paper tape tension data, vibration acceleration data, and paper operating condition data from the anchored measuring point set, and adding time stamps to write them into the data buffer; performing outlier removal, trend term removal, filtering and noise reduction, and dimension normalization on the data in the data buffer to generate preprocessed data; and mapping the preprocessed data to the spindle phase coordinate system based on the spindle phase data to form the phase synchronization data.

[0019] Preferably, generating the phase condition load feature vector based on the phase synchronization data includes: extracting paper tape tension features, rotational speed features, motor current features, vibration features, and paper operating condition features from the phase synchronization data according to the main shaft phase interval, and performing normalization processing to generate normalized load feature data; and encapsulating the normalized load feature data according to a preset feature order to obtain the phase condition load feature vector corresponding to the main shaft phase interval.

[0020] Preferably, generating the prior values ​​of the global response basis coefficients includes: inputting the phase condition load feature vector and the corresponding principal axis phase interval into a preset prior mapping model, and outputting the prior values ​​of the global response basis coefficients corresponding to the phase condition global response basis matrix; writing the prior values ​​of the global response basis coefficients into a prior coefficient set in the order of the principal axis phase intervals; applying preset boundary constraints to the prior coefficient set to generate prior coefficient constraint data for posterior correction.

[0021] Preferably, obtaining the posterior global response basis coefficients includes: reading the response data of the anchored measurement point set, the prior values ​​of the global response basis coefficients, the anchored measurement point noise weight data and the spatial continuity constraint data, and inputting them into the weighted posterior correction model, outputting the posterior global response basis coefficients corresponding to each principal axis phase interval; and writing the posterior global response basis coefficients into the posterior coefficient set according to the principal axis phase interval.

[0022] Preferably, generating the global stress field in the global evaluation model includes: retrieving the corresponding phase-condition global response basis matrix from the phase-condition global response basis library according to the principal axis phase interval; reading the posterior global response basis coefficients corresponding to the phase-condition global response basis matrix according to the principal axis phase interval; performing matching processing on the phase-condition global response basis matrix and the corresponding posterior global response basis coefficients to generate phase-condition global stress data; writing the phase-condition global stress data into the corresponding evaluation nodes according to the evaluation nodes in the global evaluation model to form a set of node stress histories; and generating the global stress field according to the set of node stress histories.

[0023] Preferably, the step of outputting the structural reliability assessment result based on the non-measuring point damage protrusion discrimination result includes: reading the set of nodal stress histories in the global stress field according to the evaluation nodes, and generating a set of nodal fatigue damages in combination with the material fatigue parameters; writing the set of nodal fatigue damages into the global evaluation model to generate the global fatigue damage field; generating a set of nodal reliability based on the set of nodal fatigue damages and the damage threshold, and writing the set of nodal reliability into the global evaluation model to generate the global reliability field; distinguishing between anchored measuring point regions and non-anchored measuring point regions according to the candidate measuring point mapping set, and generating a non-measuring point damage protrusion discrimination result based on the global fatigue damage field; and encapsulating the global fatigue damage field, the global reliability field, and the non-measuring point damage protrusion discrimination result into a structural reliability assessment result.

[0024] This invention provides a method for assessing the structural reliability of a folding machine based on multi-source load fusion. It has the following beneficial effects:

[0025] 1. This invention establishes a global evaluation model and a phase-condition global response base library, enabling the structural response of key structural components of a folding machine to be uniformly expressed according to the evaluation nodes and the main shaft phase. By determining the set of anchoring measurement points and collecting the response data of the anchoring measurement points and the multi-source load data of the folding machine, phase synchronization data is generated, allowing the measurement point response and multi-source load information to correspond to the same folding action stage. Prior values ​​of global response base coefficients are generated through the phase-condition load feature vector, and corrected using the anchoring measurement point response data to obtain the posterior global response base coefficients. This, in turn, generates a global stress field, a global fatigue damage field, and a global reliability field, realizing the overall reliability assessment of key structural components of the folding machine. This solves the problem that existing folding machine structural reliability assessment methods are prone to missing non-measurement point fatigue damage in the overall fatigue state assessment of the folding machine structure.

[0026] 2. This invention constructs a global evaluation model and combines it with the phase interval of the main shaft to generate a global response base library of phase conditions. This unifies the structural regions, evaluation nodes, material fatigue properties, measurement point mapping relationships, and folding action phases of key structural components of the folding machine under the same evaluation framework. This allows the structural response under different folding action stages to be expressed according to the evaluation nodes, avoiding the need to rely solely on fixed measurement points or fixed response modes to judge the structural fatigue state. This improves the data consistency and phase adaptability of the structural reliability assessment of the folding machine.

[0027] 3. This invention generates phase synchronization data through anchored measuring point response data and multi-source load data of the folding machine, and further obtains the phase conditional load characteristic vector and the prior value of the global response basis coefficient. Then, the anchored measuring point response data is used to correct the prior value of the global response basis coefficient, so that multi-source load information such as paper tape tension, rotation speed, motor current, vibration acceleration and paper working conditions can participate in the global stress field reconstruction together with the measured structural response, thereby improving the matching degree between the global stress field and the actual operating state of the folding machine.

[0028] 4. This invention generates a global stress field, a global fatigue damage field, and a global reliability field in a global evaluation model, and outputs structural reliability assessment results by combining the non-measuring point damage protrusion discrimination results. This expands the structural reliability assessment of the folding machine from a local judgment of a single measuring point to a holistic judgment of the global evaluation nodes. It can identify hidden fatigue weak areas outside the anchored measuring point area and reduce the possibility of missing non-measuring point fatigue damage. Attached Figure Description

[0029] Figure 1 This is a flowchart of a folding machine structural reliability assessment method based on multi-source load fusion proposed in this invention;

[0030] Figure 2 This is an architecture diagram of a folding machine structural reliability assessment system based on multi-source load fusion proposed in an embodiment of the present invention;

[0031] Figure 3 This is a flowchart illustrating the establishment of a global evaluation model as proposed in an embodiment of the present invention;

[0032] Figure 4 This is a flowchart illustrating the generation of a phase-conditional global response base library proposed in an embodiment of the present invention;

[0033] Figure 5 This is a flowchart illustrating the process of determining the set of anchoring measurement points according to an embodiment of the present invention. Detailed Implementation

[0034] The technical solution of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0035] Example 1:

[0036] In a first embodiment of the present invention, the present invention provides a method for assessing the structural reliability of a folding machine based on multi-source load fusion, such as... Figure 1 As shown, it includes the following steps:

[0037] Please see the appendix Figure 3 A global evaluation model for key structural components of a folding machine is established. The global evaluation model includes evaluation nodes, structural regions, candidate measurement point mapping sets, material fatigue parameters, damage thresholds, and principal axis phase intervals.

[0038] Furthermore, a global evaluation model for key structural components of the folding machine is established, including: acquiring structural geometric data, material parameter data, and connection relationship data of the preset structural components of the folding machine; generating an evaluation node set based on the structural geometric data and configuring node position data for each evaluation node in the evaluation node set; generating a structural region set based on the structural geometric data and connection relationship data, and establishing a node region mapping set between the evaluation node set and the structural region set; generating a material fatigue parameter set and a damage threshold set based on the material parameter data; generating a main shaft phase interval set based on the main shaft phase signal of the folding machine; generating a candidate measurement point set based on the candidate measurement point position data, and establishing a candidate measurement point mapping set between the candidate measurement point set and the evaluation node set; configuring the evaluation node set, structural region set, node region mapping set, material fatigue parameter set, damage threshold set, main shaft phase interval set, and candidate measurement point mapping set into a global evaluation model.

[0039] Specifically, in this step, the global evaluation model is used as a unified data carrier for the subsequent generation of response bases, determination of anchoring measurement points, writing of global stress fields, generation of global fatigue damage fields, and output of global reliability fields for key structural components of the folding machine. The load application locations and response regions of key structural components of the folding machine are not the same under different principal axis phases. Therefore, before performing multi-source load fusion, the structural geometry, material fatigue properties, structural regions, candidate measurement point locations, and principal axis phase intervals are uniformly configured into the same global evaluation model so that all subsequent data can be correlated according to the evaluation nodes.

[0040] In this embodiment, the structural geometry data, material parameter data, and connection relationship data of the key structural components preset by the folding machine are obtained. The structural geometry data includes the outer contour, hole positions, connection boundaries, and force transmission paths of the key structural components. The material parameter data includes the elastic modulus, Poisson's ratio, density, fatigue strength coefficient, and fatigue damage critical value. The connection relationship data includes the node adjacency relationships of bolt connections, bearing seat connections, wall panel connections, and folding knife connecting rod connections.

[0041] Based on the structural geometry data, the key structural components of the folding machine are discretized to form an evaluation node set, and node position data is configured for each evaluation node. The evaluation node set is represented as follows: ,in, Represents the set of evaluation nodes. Indicates the first Each evaluation node In this embodiment, the number of evaluation nodes is indicated. The optimal value is 1200. Each evaluation node is configured with a node position vector. ,in, Indicates the first The position of each evaluation node in the coordinate system of the key structural components of the folding machine. , , These represent the coordinates of the evaluation node along the length, width, and height directions, respectively. In this embodiment, the preferred unit for the coordinates is millimeters.

[0042] A set of structural regions is generated based on structural geometric data and connectivity data. This set of structural regions is represented as... ,in, Represents a set of structural regions. Indicates the first Each structural region Indicates the number of structural regions. In this embodiment, Preferably, there are six structural areas, including the root area of ​​the folding blade connecting rod, the roller support area, the edge area of ​​the wall panel opening, the bearing seat mounting area, the transmission connection area, and the support mounting area.

[0043] To establish the correspondence between evaluation nodes and structural regions, the locations of evaluation nodes are matched with the boundaries of structural regions, resulting in a set of node-region mappings. The node-region mapping relationship is represented as follows: ,in, Indicates evaluation node The structural region to which it belongs, Representation and Evaluation Nodes The corresponding structural region. Through this mapping relationship, the subsequently generated nodal stress data, nodal fatigue damage data, and nodal reliability data can be further aggregated into the corresponding structural region.

[0044] A set of material fatigue parameters and a set of damage thresholds are generated based on material parameter data. The set of material fatigue parameters is represented as follows: ,in, Represents the set of material fatigue parameters. Indicates the first The material fatigue parameters corresponding to each evaluation node. In this embodiment, the elastic modulus of the key structural component material is preferably 210 gigapascals, and the Poisson's ratio is preferably 0.3. The damage threshold set is represented as... ,in, Represents the set of damage thresholds. Indicates the first In this embodiment, the fatigue damage critical threshold for each evaluation node is... The preferred option is one.

[0045] In this embodiment, a set of spindle phase intervals is generated based on the spindle phase signal of the folding machine. The set of spindle phase intervals is represented as follows: ,in, Represents the set of principal axis phase intervals. Indicates the first One principal axis phase interval, Indicates the number of principal axis phase intervals; in this embodiment, The preferred value is eight. By using the set of main axis phase intervals, the subsequent phase condition global response base library can be generated and invoked according to the page-folding action stage.

[0046] A candidate measurement point set is generated based on the candidate measurement point location data, and a candidate measurement point mapping set is established between the candidate measurement point set and the evaluation node set. The candidate measurement point set is represented as follows: ,in, Represents the set of candidate measurement points. Indicates the first One candidate measurement point, Indicates the number of candidate measurement points. In this embodiment, The preferred number is thirty. Candidate measurement points. The corresponding evaluation node is determined by distance matching. ,in, Indicates candidate measurement points Mapped to the evaluation node, Indicates candidate measurement points The installation location, Indicates evaluation node The node position, This represents the spatial distance between the candidate measurement point and the evaluation node. In this embodiment, the maximum allowable mapping distance between the candidate measurement point and the evaluation node is preferably five millimeters. The evaluation node set, structural region set, node region mapping set, material fatigue parameter set, damage threshold set, principal axis phase interval set, and candidate measurement point mapping set are associated and stored to form a global evaluation model. Therefore, the subsequently generated phase condition global response base library can be called according to the evaluation node, the anchored measurement point can be determined according to the candidate measurement point mapping set, and the global stress field, global fatigue damage field, and global reliability field can be written into the same global evaluation model.

[0047] Please see the appendix Figure 4 Based on the global evaluation model, a global response base library for phase conditions is generated under each principal axis phase interval;

[0048] Furthermore, a phase condition global response base library is generated for each spindle phase interval, including: acquiring unit load response data for each spindle phase interval, including response data corresponding to paper tape tension load, folding knife impact load, roller support reaction force, spindle torque load, frame vibration foundation excitation, and rotational inertial load; mapping the unit load response data to each evaluation node in the global evaluation model to generate phase global response data; performing principal component extraction on the phase global response data to obtain the phase condition global response basis matrix corresponding to the spindle phase interval; and archiving each phase condition global response basis matrix according to the spindle phase interval order to form a phase condition global response base library.

[0049] Specifically, in this step, the phase-conditional global response basis library is used to characterize the main stress response modes of the key structural components of the folding machine under different spindle phase intervals. The paper tape tension, folding knife impact, roller support reaction force, spindle torque, frame vibration foundation excitation, and rotational inertial load all act at different times within a folding cycle. If the same response basis is used to describe all phases, it can easily reduce the adaptability of subsequent global stress field reconstruction. Therefore, after establishing the global evaluation model, phase-conditional global response basis matrices are generated according to the spindle phase intervals.

[0050] Specifically, regarding the first Each principal axis phase interval The response data of the evaluation nodes under six types of unit loads were obtained respectively, and the unit load response vector is represented as follows: ,in, Indicates the first Within the first principal axis phase interval The global response vector corresponding to a unit load. This indicates that the unit load is at the evaluation node. The stress response value formed at that location, Indicates the number of evaluation nodes. This indicates the unit load type number. In this embodiment, The preferred values ​​are one to six, which correspond to paper tape tension load, folding knife impact load, roller support reaction force, spindle torque load, frame vibration foundation excitation, and rotational inertia load, respectively. The preferred unit load for paper tape tension is 100 Newtons, the preferred unit load for folding knife impact is 500 Newtons, and the preferred unit load for spindle torque is 20 Newton-meters.

[0051] The unit load response vectors within the same principal axis phase interval are used to form the phase global response matrix. ,in, Indicates the first Phase-wide response data for each principal axis phase interval In this embodiment, the quantity of a unit load type is indicated. The preferred method is six; mean removal is performed on the phase global response data to obtain centered response data. Then perform principal component extraction. ,in, This represents the global node response direction matrix. This represents the response contribution intensity matrix. This represents the load type correlation matrix; the top-ranked components are selected in descending order of response contribution intensity. The global node response directions form the phase-conditional global response basis matrix. ,in, Indicates the phase interval with the principal axis The corresponding phase conditional global response basis matrix, Indicates the first A global response basis; in this embodiment, the response contribution retention threshold is preferably 0.95. Preferably three to five. The phase intervals corresponding to each principal axis... Archived according to the principal axis phase order to form a phase condition global response base library, so that subsequent anchoring measurement point determination and global stress field reconstruction can call the corresponding response base data according to the same phase conditions.

[0052] Please see the appendix Figure 5 Based on the candidate measurement point mapping set and the phase condition global response base library, the anchor measurement point set is determined;

[0053] Furthermore, determining the set of anchor measurement points includes: determining a set of candidate measurement points on the preset structural components of the folding machine; generating a candidate measurement point observation matrix based on the positional mapping relationship between the candidate measurement point set and the evaluation nodes in the global evaluation model; associating the candidate measurement point observation matrix with the phase condition global response base library to form reconstructed correlation data of candidate measurement points to the global fatigue damage field; generating measurement point reconstruction score data for each candidate measurement point based on the reconstructed correlation data, and forming a measurement point screening sequence according to the measurement point reconstruction score data; and selecting the set of anchor measurement points from the set of candidate measurement points according to the measurement point screening sequence.

[0054] Specifically, in this step, the anchored measurement point set is used to support the reconstruction of the global fatigue state at a limited number of measured locations. If the measurement points are arranged only according to the local stress peak, the measurement points may be concentrated in a single structural region, making it difficult to reflect the damage changes in non-measurement point regions. Therefore, this embodiment evaluates the observation capability of each candidate measurement point to the global response basis based on the candidate measurement point mapping set and the phase condition global response basis library, and determines the anchored measurement point set accordingly.

[0055] For candidate measurement points The evaluation nodes corresponding to the candidate measurement points are obtained from the candidate measurement point mapping set, and observation row vectors are generated. In the principal axis phase interval The observation relationship between the candidate measurement points and the phase condition global response basis matrix is ​​as follows: ,in, Indicates candidate measurement points In the principal axis phase interval The response baseline observation data, This represents the extraction vector from the candidate measurement point to the evaluation node. This represents the phase condition global response basis matrix for the corresponding principal axis phase interval.

[0056] A successive screening method is used to form a sequence of selected measurement points. The set of selected measurement points is denoted as . Candidate measurement points After adding, the candidate combination observation matrix is ​​obtained. Its measurement point reconstruction scoring data is ,in, Indicates candidate measurement points The measurement points are reconstructed into scoring data. Indicates the principal axis phase interval The weight, Represents the noise covariance matrix at the measurement points. Represents the regularization coefficient. Represents the identity matrix; in this embodiment, The preferred option is the percentage of the duration of the principal axis phase interval. Preferably 10 -6 The preferred standard deviation of strain noise at the measuring point is 2. The preferred number of anchoring measurement points K is six; according to A sequence of measurement points is formed from high to low, and the first K candidate measurement points are selected as the anchor measurement point set so that the subsequent global stress field reconstruction has a stable observation input.

[0057] The response data of the anchor measurement point set and the multi-source load data of the folding machine are collected and mapped to the main spindle phase interval to generate phase synchronization data;

[0058] Furthermore, phase synchronization data is generated, including: collecting strain response data, spindle phase data, spindle speed data, motor current data, paper tape tension data, vibration acceleration data, and paper operating condition data from the anchored measuring point set, and adding time stamps to write them into the data buffer; performing outlier removal, trend term removal, filtering and noise reduction, and dimension normalization on the data in the data buffer to generate preprocessed data; using the spindle phase data as a reference, mapping the preprocessed data to the spindle phase coordinate system to form phase synchronization data.

[0059] Specifically, in this step, phase synchronization data is used to unify the anchoring point response and the multi-source load data of the folding machine under the same spindle phase reference. The sampling frequencies of anchoring point strain, spindle speed, motor current, paper tape tension, and vibration acceleration are different. If fusion is performed directly according to the sampling time, it can easily lead to data mismatch in the same folding action phase. Therefore, after obtaining the anchoring point set, various types of data are first time-stamped, preprocessed, and phase-mapped to ensure that the subsequent phase-conditional load feature vectors maintain a correspondence with the phase-conditional global response base library.

[0060] During the operation of the folding machine, strain response data, spindle phase data, spindle speed data, motor current data, paper tape tension data, vibration acceleration data, and paper condition data are simultaneously collected from the anchored measuring point set, and a time stamp is added to each collected data point. The collected data can be represented as follows: ,in, Indicates the sampling time The following data set Indicates the first Strain response at each anchored measuring point Indicates the principal axis phase. Indicates the spindle speed. Indicates motor current. Indicates paper tape tension. Indicates vibration acceleration. The data represents the paper's operating conditions. In this embodiment, the preferred sampling frequency for the anchoring point strain is 1 kHz, the preferred sampling frequency for the spindle phase is 1 kHz, the preferred sampling frequency for the motor current is 500 Hz, the preferred sampling frequency for the paper tape tension is 500 Hz, and the preferred sampling frequency for the vibration acceleration is 2 kHz.

[0061] After the collected data is written to the data buffer, outlier removal, trend term removal, filtering and noise reduction, and dimension normalization are performed on the data. The preprocessed data is represented as follows: ,in, This indicates preprocessed data. This indicates outlier removal and trend removal processes, used to remove sudden jumps and slow drift values ​​from the sensor. This indicates filtering and noise reduction processing, used to preserve data within the effective frequency band of the folding machine's structural response. This indicates dimensional normalization processing, used to convert data with different dimensions into a unified scale. In this embodiment, the outlier removal threshold is preferably three times the absolute deviation of the median value in adjacent sampling windows, the filtering frequency band is preferably five Hz to three hundred Hz, and the normalization method is preferably based on the mean and standard deviation of the current stable operating segment.

[0062] Using the spindle phase data as a reference, the preprocessed data is mapped to the spindle phase coordinate system to obtain phase synchronization data. The phase mapping relationship is expressed as follows: ,in, Indicates the principal axis phase The phase synchronization data below, Indicates that the principal axis phase has reached The corresponding sampling time; when different channels are at When no direct sample value exists at a given moment, interpolation is performed using data from adjacent time markers. In this embodiment, the principal axis phase resolution is preferably one degree, and each complete page loop forms 360 phase sampling points. Through the above processing, the anchored measurement point response and multi-source load data can be grouped into corresponding data segments according to the principal axis phase interval, providing a unified input for the subsequent generation of phase conditional load feature vectors and the generation of prior values ​​for global response basis coefficients.

[0063] A phase conditional load feature vector is generated based on the phase synchronization data, and prior values ​​of the global response basis coefficients are generated.

[0064] Furthermore, a phase condition load feature vector is generated based on the phase synchronization data, including: extracting paper tape tension features, speed features, motor current features, vibration features, and paper operating condition features from the phase synchronization data according to the main shaft phase interval, and performing normalization processing to generate normalized load feature data; encapsulating the normalized load feature data according to a preset feature order to obtain the phase condition load feature vector corresponding to the main shaft phase interval.

[0065] Furthermore, generating prior values ​​of global response basis coefficients includes: inputting the phase condition load eigenvector and the corresponding principal axis phase interval into a preset prior mapping model, and outputting prior values ​​of global response basis coefficients corresponding to the phase condition global response basis matrix; writing the prior values ​​of global response basis coefficients into the prior coefficient set in the order of principal axis phase intervals; applying preset boundary constraints to the prior coefficient set to generate prior coefficient constraint data for posterior correction.

[0066] Specifically, in this step, the phase condition load feature vector is used to convert the phase synchronization data into a unified input that can characterize the load state of the current spindle phase interval. The prior value of the global response basis coefficient is used to provide an initial estimate of the global response mode of the key structural components of the folding machine before the anchor measurement point response correction. The influence paths of paper tape tension, spindle speed, motor current, vibration acceleration and paper working conditions on the structural response are different. Directly using the original data of each channel for global stress field reconstruction will cause inconsistencies in data scale and unclear phase meaning. Therefore, this step first extracts the load features according to the spindle phase interval, and then obtains the prior value of the global response basis coefficient corresponding to the phase condition global response basis matrix through a preset prior mapping model.

[0067] For the principal axis phase interval From phase synchronization data Extracting paper tape tension characteristics, rotational speed characteristics, motor current characteristics, vibration characteristics, and paper operating condition characteristics to form the original load feature vector. ,in, Indicates the first Load characteristic vectors corresponding to each principal axis phase interval This indicates the paper tape tension within the main shaft phase range. Stable representative value within, This indicates the amount of paper tape tension fluctuation. This represents the spindle speed. This indicates the amount of spindle speed fluctuation. This represents the motor current value. This indicates the amount of fluctuation in motor current. Indicates the energy characteristics of vibrational acceleration. The paper condition code is indicated; in this embodiment, the paper condition code is preferably determined by the paper weight and the paper width, with the paper weight preferably being 70 grams per square meter and the paper width preferably being 780 millimeters.

[0068] The original load feature vector is normalized to obtain normalized load feature data. ,in, This represents normalized load characteristic data. This represents the baseline mean of load characteristics obtained from stable operating samples. This represents the standard deviation of the load characteristics obtained from a stable operating sample. In this embodiment, the stable operating sample is preferably the sample data from a folding machine that has been running continuously for ten minutes without roll changes or downtime. The normalized load characteristic data is packaged according to a preset order of paper tape tension, spindle speed, motor current, vibration acceleration, and paper conditions to obtain the phase interval with the spindle. The corresponding phase condition load eigenvector.

[0069] The phase condition load eigenvector and the corresponding principal axis phase interval are input into a preset prior mapping model to generate prior values ​​of the global response basis coefficients. The preset prior mapping model adopts a linear mapping model with phase encoding. ,in, Indicates the first Prior values ​​of the global response basis coefficients under each principal axis phase interval. Indicates the principal axis phase interval The corresponding load feature mapping matrix, Indicates the principal axis phase interval The corresponding reference bias vector, Represents the phase condition load eigenvector; Each component corresponds to a global response basis in the phase-conditional global response basis matrix, used to represent the degree of participation of that response basis under the current load state; in this embodiment, The preferred dimension is three to five. It is obtained by pre-calibration using historical stable operating data and corresponding global response base coefficient samples. The reference response coefficient under no-load and low-speed conditions is preferred.

[0070] To prevent abnormal load characteristics from causing prior values ​​to exceed the physically feasible range, preset boundary constraints are applied to the set of prior coefficients to obtain prior coefficient constraint data for posterior correction. ,in, This represents the prior values ​​of the global response basis coefficients after applying boundary constraints. Describes the boundary constraint function. Indicates the principal axis phase interval The lower boundary of the global response basis coefficients. Indicates the principal axis phase interval The upper boundary of the global response basis coefficients; in this embodiment, Ideally, it should be 0.8 times the minimum value of the corresponding coefficient in the historical stable operating samples. The preferred value is 1.2 times the maximum value of the corresponding coefficient in the historical stable operating samples; the phase interval of each principal axis is... Write the coefficients into the prior coefficient set in the order of the principal axis phase interval, and use them as input for the subsequent weighted posterior correction model.

[0071] The prior values ​​of the global response basis coefficients are corrected using the response data to obtain the posterior global response basis coefficients;

[0072] Furthermore, the posterior global response basis coefficients are obtained by: reading the response data of the anchored measurement point set, the prior values ​​of the global response basis coefficients, the noise weight data of the anchored measurement points, and the spatial continuity constraint data, and inputting them into the weighted posterior correction model, outputting the posterior global response basis coefficients corresponding to each principal axis phase interval; and writing the posterior global response basis coefficients into the posterior coefficient set according to the principal axis phase interval.

[0073] Specifically, in this step, the posterior global response basis coefficients are used to introduce the measured response constraints of the anchored measurement points based on the prior values ​​of the global response basis coefficients. This ensures that the global response mode is driven by the characteristics of multi-source loads and is consistent with the actual measured local structural response. The prior values ​​can reflect the overall load state under the current spindle phase interval, but they may still be affected by paper fluctuations, sensor noise, and model bias. Therefore, it is necessary to use the response data of the anchored measurement point set to perform weighted correction.

[0074] Specifically, regarding the principal axis phase interval Read the response data of the anchored measuring point set to form the measuring point response vector. ,in, Indicates the first The response data of the set of anchored measuring points under each principal axis phase interval. Indicates the first Each anchoring measuring point is in the principal axis phase interval The strain response or stress response under these conditions This indicates the number of anchoring measurement points; in this embodiment, Preferably, there are six, and the response data of the measuring points are preferably the representative values ​​of strain stability within the principal axis phase interval.

[0075] Reading the phase conditional global response basis matrix An anchored measuring point observation matrix is ​​formed based on the positional mapping relationship of the anchored measuring point set in the global evaluation model. Thus, the basic observation matrix of the measurement point response is obtained. ,in, Indicates the first Observational data of the phase condition global response basis matrix at anchored measurement points in each principal axis phase interval. This represents the observation mapping matrix between the set of anchored measurement points and the evaluation nodes. This represents the phase condition global response basis matrix for the corresponding principal axis phase interval.

[0076] The response vector of the measuring point global response basis coefficient prior value The anchored measuring point noise weight data and spatial continuity constraint data are input into the weighted posterior correction model to obtain the posterior global response basis coefficients. ,in, Indicates the first The posterior global response basis coefficients under each principal axis phase interval This represents the response basis coefficient to be corrected. This represents the prior values ​​of the global response basis coefficients after applying boundary constraints. This represents the noise covariance matrix of the anchored measurement points. Denotes the prior coefficient covariance matrix. This represents the spatial continuity constraint matrix between adjacent evaluation nodes in the global evaluation model. This represents the weight of the spatial continuity constraint; in this embodiment, the standard deviation of the strain noise at the anchoring measuring point is preferably... , The diagonal element is preferably 0.25. The preferred value is 0.01.

[0077] When the noise level at a certain anchoring measurement point increases within the current principal axis phase range, by increasing... The corresponding values ​​at the specified locations reduce the impact of that measurement point during the correction process. Specifically, strain spikes caused by motor current fluctuations or loose patches within the electromagnetic interference frequency band can be suppressed using anchored measurement point noise weight data. The weighted posterior correction model outputs... Write the values ​​into the posterior coefficient set in the order of the principal axis phase intervals. ,in, Denotes the set of posterior coefficients. Indicates the number of principal axis phase intervals; in this embodiment, The preferred value is eight; this posterior coefficient set is used to subsequently match the global response basis library under phase conditions to generate a global stress field.

[0078] Based on the phase-condition global response basis library and the posterior global response basis coefficients, a global stress field is generated in the global evaluation model.

[0079] Furthermore, generating a global stress field in the global evaluation model includes: retrieving the corresponding phase-condition global response basis matrix from the phase-condition global response basis library according to the principal axis phase interval; reading the posterior global response basis coefficients corresponding to the phase-condition global response basis matrix according to the principal axis phase interval; matching the phase-condition global response basis matrix with the corresponding posterior global response basis coefficients to generate phase-condition global stress data; writing the phase-condition global stress data into the corresponding evaluation nodes according to the evaluation nodes in the global evaluation model to form a set of node stress histories; and generating a global stress field based on the set of node stress histories.

[0080] Specifically, in this step, the global stress field is used to restore the aforementioned posterior global response basis coefficients to the stress state of each evaluation node of the key structural component of the folding machine under different principal axis phase intervals. The posterior global response basis coefficients only represent the degree of participation of each response basis under the current load state and cannot be directly used as nodal stress. Therefore, it is necessary to match them with the corresponding phase condition global response basis matrix in the phase condition global response basis library and write them into the evaluation nodes in the global evaluation model.

[0081] For the principal axis phase interval Retrieve the phase-conditional global response basis matrix from the phase-conditional global response basis library. And read the corresponding posterior global response basis coefficients from the posterior coefficient set. Generate phase-wide stress data ,in, Indicates the principal axis phase interval The phase-wide stress data below, This represents the global response basis matrix corresponding to the phase interval of the principal axis. This represents the basis coefficients of the posterior global response after correction based on the anchored measurement point response.

[0082] Will The nodal stress values ​​are written into the corresponding nodes according to the evaluation node order in the global evaluation model to obtain the nodal stress data. ,in, Indicates evaluation node In the principal axis phase interval The stress value below, This indicates the phase global stress data and the evaluation node. The corresponding data items; in this embodiment, the stress unit is preferably megapascal, and the number of principal axis phase intervals is preferably eight.

[0083] The stress values ​​of the same evaluation node in each principal axis phase interval are arranged in order of principal axis phase to form the nodal stress history. ,in, Indicates evaluation node The nodal stress history, This indicates the number of principal axis phase intervals; the set of nodal stress histories of all evaluation nodes is written into the global evaluation model to form a global stress field; thus, a global fatigue damage field and a global reliability field can be generated based on the same evaluation node system.

[0084] Based on the global stress field, a global fatigue damage field and a global reliability field are generated in the global evaluation model, and the structural reliability assessment results are output according to the non-measuring point damage protrusion discrimination results.

[0085] Furthermore, based on the non-measuring point damage provocation discrimination results, the structural reliability assessment results are output, including: reading the set of nodal stress histories in the global stress field according to the evaluation nodes, and generating a set of nodal fatigue damage according to the material fatigue parameters; writing the set of nodal fatigue damage into the global evaluation model to generate a global fatigue damage field; generating a set of nodal reliability according to the set of nodal fatigue damage and damage threshold, and writing the set of nodal reliability into the global evaluation model to generate a global reliability field; distinguishing between anchored and non-anchored measuring point regions according to the candidate measuring point mapping set, and generating non-measuring point damage provocation discrimination results based on the global fatigue damage field; and encapsulating the global fatigue damage field, the global reliability field, and the non-measuring point damage provocation discrimination results into a structural reliability assessment result.

[0086] Specifically, in this step, the global fatigue damage field and global reliability field are used to convert the aforementioned global stress field into evaluation results that can characterize the overall fatigue state of the key structural components of the folding machine. The stress change of a single anchored measuring point can only reflect the local response and cannot directly determine whether damage concentration has occurred in the non-anchored measuring point area. Therefore, after forming the set of nodal stress histories, fatigue damage data and reliability data are generated one by one according to the evaluation nodes, and the non-measuring point damage protrusion discrimination results are generated in combination with the candidate measuring point mapping set.

[0087] For evaluation nodes Read its nodal stress history And combined with the corresponding material fatigue parameters in the material fatigue parameter set The stress history of nodes is cyclically identified to obtain the number of cycles and corresponding fatigue life under different stress cycle levels. The fatigue damage of nodes is represented as follows: ,in, Indicates evaluation node Fatigue damage at nodes, Indicates the first One stress cycle level, Indicates evaluation node The number of stress cycle levels, Indicates the first Number of cycles at each stress cycle level Indicates material fatigue parameters The corresponding allowable cycle life. In this embodiment, The preferred value is 10, and the preferred statistical time for node fatigue damage is one hour of working time. Obtained from the material stress-life curve.

[0088] The fatigue damage of all evaluation nodes is written into the global evaluation model to form a global fatigue damage field. ,in, Represents the global fatigue damage field. Indicates the number of evaluation nodes; based on the node damage thresholds in the damage threshold set. Generate node reliability data. The node limit state variables are represented as follows: ; Node reliability index is expressed as ;in, Indicates evaluation node The fatigue limit state quantity. Indicates evaluation node The damage threshold Indicates evaluation node Reliability index, This represents the mean of the limiting state quantity. This represents the standard deviation of the limiting state quantity. In this embodiment, The preferred option is one. The optimal coefficient of variation of the material fatigue parameters is determined by both the material fatigue parameter dispersion and the stress field reconstruction error, with the optimal coefficient of variation of the material fatigue parameters being 0.1.

[0089] Determine the set of anchored measuring point regions based on the candidate measuring point mapping set. Set of non-anchored measuring point areas Based on the global fatigue damage field, non-measuring point damage abruptness discrimination results are generated. ,in, This indicates the discriminant value for sudden damage at non-measuring points. This represents the set of anchored measuring points. This represents the set of non-anchored measuring point regions. This represents the correction amount for the denominator, preferably 10. -6 The optimal threshold for identifying sudden damage at non-measuring points is one point; when When the damage exceeds the threshold for non-measuring point damage protrusion, a non-measuring point damage protrusion discrimination result is generated, and the corresponding evaluation node and its corresponding structural region are written into the structural reliability assessment result. Thus, the structural reliability assessment result includes the global fatigue damage field, the global reliability field, and the non-measuring point damage protrusion discrimination result, which can provide data basis for judging the overall reliability of key structural components of the folding machine.

[0090] Example 2:

[0091] In a second embodiment of the present invention, the present invention provides a folding machine structural reliability assessment system based on multi-source load fusion, such as... Figure 2 As shown, it includes the following modules:

[0092] Modeling module: Used to establish a global evaluation model for key structural components of the folding machine. The global evaluation model includes evaluation nodes, structural regions, candidate measurement point mapping sets, material fatigue parameters, damage thresholds, and principal axis phase intervals.

[0093] Base Library Module: Used to generate a phase-condition global response base library based on the global evaluation model under each principal axis phase interval;

[0094] Measurement point module: used to determine the set of anchored measurement points based on the candidate measurement point mapping set and the phase condition global response base library;

[0095] Synchronization module: used to collect response data from the anchor measurement point set and multi-source load data of the folding machine, and map them to the main spindle phase interval to generate phase synchronization data;

[0096] Prior module: used to generate phase conditional load feature vectors based on phase synchronization data, and generate prior values ​​of global response basis coefficients;

[0097] Correction module: Used to correct the prior values ​​of the global response basis coefficients using response data to obtain the posterior global response basis coefficients;

[0098] Stress module: Used to generate a global stress field in a global evaluation model based on the phase-condition global response basis library and the posterior global response basis coefficients;

[0099] Evaluation module: Based on the global stress field, it generates a global fatigue damage field and a global reliability field in the global evaluation model, and outputs the structural reliability evaluation results based on the non-measuring point damage protrusion discrimination results.

[0100] In continuous production scenarios of roll paper folding machines, key structural components are subjected to the combined effects of paper belt tension, folding knife impact, roller support reaction force, spindle torque fluctuation, and frame vibration. The structural fatigue state changes with the phase of the folding action. Traditional single-point monitoring methods can only reflect localized stress changes and are unable to identify fatigue damage concentrations in non-measuring areas, easily leading to deviations in the structural reliability assessment results from the actual service condition. To address these issues, this invention employs a multi-source load fusion-based structural reliability assessment system for folding machines, the architecture of which is as follows: Figure 2 As shown. The specific implementation process of this system is as follows:

[0101] First, the modeling module acquires the structural geometry data, material parameter data, connection relationship data, candidate measurement point position data, and main shaft phase signal of the key structural components of the folding machine, and associates and configures the above data into a global evaluation model, so that the evaluation nodes, structural regions, candidate measurement point mapping sets, material fatigue parameters, damage thresholds, and main shaft phase intervals are within the same data framework;

[0102] Subsequently, the base library module calls the evaluation nodes and principal axis phase intervals in the global evaluation model, organizes the unit load response data corresponding to different load sources, and generates a phase condition global response base library under each principal axis phase interval;

[0103] Next, the measurement point module generates reconstructed correlation data of the candidate measurement points to the global fatigue damage field based on the candidate measurement point mapping set and the phase condition global response base library, and determines the anchoring measurement point set accordingly; the synchronization module collects the anchoring measurement point response data, spindle phase data, spindle speed data, motor current data, paper tape tension data, vibration acceleration data and paper working condition data, and maps them to the spindle phase coordinate system after preprocessing to generate phase synchronization data;

[0104] Meanwhile, the prior module extracts load features from the phase synchronization data according to the principal axis phase interval, and generates the phase condition load feature vector and the prior value of the global response basis coefficient; the correction module uses the anchored measuring point response data to perform weighted correction on the prior value of the global response basis coefficient to obtain the posterior global response basis coefficient.

[0105] Then, the stress module generates a global stress field in the global evaluation model based on the phase condition global response basis library and the posterior global response basis coefficients.

[0106] Finally, the evaluation module generates a global fatigue damage field and a global reliability field based on the global stress field, and outputs the structural reliability evaluation results by combining the non-measuring point damage protrusion discrimination results.

[0107] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for assessing the structural reliability of a folding machine based on multi-source load fusion, characterized in that, Includes the following steps: A global evaluation model for key structural components of a folding machine is established. The global evaluation model includes evaluation nodes, structural regions, candidate measurement point mapping sets, material fatigue parameters, damage thresholds, and principal axis phase intervals. Based on the global evaluation model, a global response base library under phase conditions is generated under each of the principal axis phase intervals; Based on the candidate measurement point mapping set and the phase condition global response base library, the anchor measurement point set is determined; The response data of the anchoring measurement point set and the multi-source load data of the folding machine are collected and mapped to the main shaft phase interval to generate phase synchronization data; Based on the phase synchronization data, a phase conditional load feature vector is generated, and prior values ​​of global response basis coefficients are generated. The prior values ​​of the global response basis coefficients are corrected using the response data to obtain the posterior global response basis coefficients. Based on the phase condition global response basis library and the posterior global response basis coefficients, a global stress field is generated in the global evaluation model. Based on the global stress field, a global fatigue damage field and a global reliability field are generated in the global evaluation model, and the structural reliability assessment results are output according to the non-measuring point damage protrusion discrimination results.

2. The method for assessing the structural reliability of a folding machine based on multi-source load fusion as described in claim 1, characterized in that: The establishment of a global evaluation model for key structural components of the folding machine includes: acquiring structural geometric data, material parameter data, and connection relationship data of preset structural components of the folding machine; generating an evaluation node set based on the structural geometric data and configuring node position data for each evaluation node in the evaluation node set; generating a structural region set based on the structural geometric data and the connection relationship data, and establishing a node region mapping set between the evaluation node set and the structural region set; generating a material fatigue parameter set and a damage threshold set based on the material parameter data; generating a main shaft phase interval set based on the main shaft phase signal of the folding machine; generating a candidate measurement point set based on the candidate measurement point position data, and establishing a candidate measurement point mapping set between the candidate measurement point set and the evaluation node set; and configuring the evaluation node set, the structural region set, the node region mapping set, the material fatigue parameter set, the damage threshold set, the main shaft phase interval set, and the candidate measurement point mapping set as the global evaluation model.

3. The method for assessing the structural reliability of a folding machine based on multi-source load fusion as described in claim 1, characterized in that: The step of generating a phase-condition global response base library for each of the main shaft phase intervals includes: acquiring unit load response data for each of the main shaft phase intervals, wherein the unit load response data includes response data corresponding to paper tape tension load, folding knife impact load, roller support reaction force, main shaft torque load, frame vibration foundation excitation, and rotational inertial load; mapping the unit load response data to each evaluation node in the global evaluation model to generate phase-condition global response data; performing principal component extraction processing on the phase-condition global response data to obtain the phase-condition global response base matrix corresponding to the main shaft phase interval; and archiving each of the phase-condition global response base matrices according to the order of the main shaft phase intervals to form a phase-condition global response base library.

4. The method for assessing the structural reliability of a folding machine based on multi-source load fusion as described in claim 1, characterized in that: The determination of the anchoring measurement point set includes: determining a candidate measurement point set on the preset structural component of the folding machine; generating a candidate measurement point observation matrix based on the positional mapping relationship between the candidate measurement point set and the evaluation nodes in the global evaluation model; associating the candidate measurement point observation matrix with the phase condition global response base library to form reconstruction association data of the candidate measurement points to the global fatigue damage field; generating measurement point reconstruction score data for each candidate measurement point based on the reconstruction association data, and forming a measurement point screening sequence according to the measurement point reconstruction score data; and selecting the anchoring measurement point set from the candidate measurement point set according to the measurement point screening sequence.

5. The method for assessing the structural reliability of a folding machine based on multi-source load fusion as described in claim 1, characterized in that: The generation of phase synchronization data includes: collecting strain response data, spindle phase data, spindle speed data, motor current data, paper tape tension data, vibration acceleration data, and paper operating condition data from the anchored measuring point set, and adding time stamps to write them into the data buffer; performing outlier removal, trend term removal, filtering and noise reduction, and dimension normalization on the data in the data buffer to generate preprocessed data; and mapping the preprocessed data to the spindle phase coordinate system based on the spindle phase data to form the phase synchronization data.

6. The method for assessing the structural reliability of a folding machine based on multi-source load fusion according to claim 1, characterized in that: The step of generating a phase condition load feature vector based on the phase synchronization data includes: extracting paper tape tension features, rotational speed features, motor current features, vibration features, and paper operating condition features from the phase synchronization data according to the main shaft phase interval, and performing normalization processing to generate normalized load feature data; and encapsulating the normalized load feature data according to a preset feature order to obtain a phase condition load feature vector corresponding to the main shaft phase interval.

7. The method for assessing the structural reliability of a folding machine based on multi-source load fusion as described in claim 1, characterized in that: The step of generating prior values ​​of global response basis coefficients includes: inputting the phase condition load feature vector and the corresponding principal axis phase interval into a preset prior mapping model, and outputting prior values ​​of global response basis coefficients corresponding to the phase condition global response basis matrix; writing the prior values ​​of global response basis coefficients into a prior coefficient set in the order of principal axis phase intervals; applying preset boundary constraints to the prior coefficient set to generate prior coefficient constraint data for posterior correction.

8. The method for assessing the structural reliability of a folding machine based on multi-source load fusion according to claim 1, characterized in that: The process of obtaining the posterior global response basis coefficients includes: reading the response data of the anchored measurement point set, the prior values ​​of the global response basis coefficients, the anchored measurement point noise weight data and the spatial continuity constraint data, and inputting them into the weighted posterior correction model, outputting the posterior global response basis coefficients corresponding to each principal axis phase interval; and writing the posterior global response basis coefficients into the posterior coefficient set according to the principal axis phase interval.

9. The method for assessing the structural reliability of a folding machine based on multi-source load fusion according to claim 1, characterized in that: The step of generating a global stress field in the global evaluation model includes: retrieving the corresponding phase-condition global response basis matrix from the phase-condition global response basis library according to the principal axis phase interval; reading the posterior global response basis coefficients corresponding to the phase-condition global response basis matrix according to the principal axis phase interval; performing matching processing on the phase-condition global response basis matrix and the corresponding posterior global response basis coefficients to generate phase-condition global stress data; writing the phase-condition global stress data into the corresponding evaluation nodes according to the evaluation nodes in the global evaluation model to form a set of node stress histories; and generating the global stress field according to the set of node stress histories.

10. The method for assessing the structural reliability of a folding machine based on multi-source load fusion according to claim 1, characterized in that: The step of outputting structural reliability assessment results based on the non-measuring point damage protrusion discrimination results includes: reading the set of nodal stress histories in the global stress field according to the evaluation nodes, and generating a set of nodal fatigue damages in combination with the material fatigue parameters; writing the set of nodal fatigue damages into the global evaluation model to generate the global fatigue damage field; generating a set of nodal reliability based on the set of nodal fatigue damages and the damage threshold, and writing the set of nodal reliability into the global evaluation model to generate the global reliability field; distinguishing between anchored and non-anchored measuring point regions according to the candidate measuring point mapping set, and generating non-measuring point damage protrusion discrimination results based on the global fatigue damage field; and encapsulating the global fatigue damage field, the global reliability field, and the non-measuring point damage protrusion discrimination results into a structural reliability assessment result.

Images