Roller lifting safety assessment method and system based on load data

Abstract

This invention relates to the field of fault prediction and health management technology, specifically a method and system for safety assessment of road roller lifting based on load data. The method includes the following steps: constructing a tensor and rotation matrix based on tension and tilt parameters to extract deflection data; calculating the area energy of the hysteresis loop corresponding to synchronous force and deformation; extracting accumulated energy and comparing it with the fracture benchmark to establish degradation evaluation data; constructing an envelope surface by shrinking the anti-tipping shell based on oscillation fatigue parameters; judging the centroid separation state and deflection anomalies to generate a safety warning. In this invention, the spatial deflection angle is accurately calculated by constructing a matrix based on multidimensional tension and installation tilt; the energy dissipated by the single hysteresis loop in the closed area is captured by screening the force deformation; the alternating cumulative damage is aggregated and introduced into the fracture benchmark to quantify the depth of structural health degradation; dynamic oscillation fatigue is comprehensively used to promote the initial shell translation and reconstruction of the shrinkage safety boundary; and cross-analysis of centroid over-boundary displacement and damage is performed to accurately issue risk warning data.

Description

Technical Field

[0001] This invention relates to the field of fault prediction and health management technology, and in particular to a method and system for assessing the lifting safety of road rollers based on load data. Background Technology

[0002] The field of fault prediction and health management technology involves real-time monitoring of the operating status of mechanical equipment and prediction of structural fatigue degradation trends. Its core involves collecting physical stress and deformation values ​​of the mechanical structure during operation by arranging strain gauges and tension gauges at the stress points of the equipment. This data is then combined with the fatigue cumulative damage equation from materials mechanics to analyze the degree of physical degradation of internal components, and a preventative maintenance list is established based on the derived limit thresholds. Traditional methods for assessing the lifting safety of road rollers involve calculating the overall frame load-bearing capacity and lifting point stress of the road roller under suspended lifting conditions. Typically, on-site lifting personnel consult the manufacturer's manual to obtain the weight distribution ratio of the road roller before and after lifting, measure the physical geometric distance between each lifting lug hole using a tape measure, and read the single sling tension value displayed on the mechanical force measuring ring. Finally, the overall machine weight distribution ratio and the measured hole distances are directly substituted into the static moment balance equation for manual paper-and-pen calculation.

[0003] Traditional lifting operations rely on manual consultation of the manufacturer's instructions to obtain the weight distribution ratio. On-site personnel measure the geometric distance between the lifting lug holes with a tape measure and read the single sling tension value presented by the mechanical force measuring ring. They then substitute this value into the static moment balance equation for calculation and deduction. This manual operation mode based on a static mechanical model cannot capture the dynamic transient alternating characteristics of the lifting process. Single-dimensional observation is difficult to reflect the degree of fatigue accumulation damage under complex spatial stress conditions, which may lead to equipment overturning or structural fracture under alternating loads. Summary of the Invention

[0004] To achieve the above objectives, the present invention adopts the following technical solution: a method for safety assessment of road roller lifting based on load data, comprising the following steps:

[0005] S1: Based on the initial tension parameters and installation tilt parameters, construct local tensor matrices and spatial rotation matrices respectively. Calculate the product of the local tensor matrix and the spatial rotation matrix to generate a global tensor matrix. Extract the principal eigenvectors of the global tensor matrix and the deviation angle of the plumb line to establish hoisting spatial deflection feature data.

[0006] S2: Filter the lifting transient force parameters and transient deformation parameters of the synchronous time node of the lifting space deflection characteristic data, calculate the plane area parameter covered inside the closed geometric boundary of the paired plane coordinate point set, and establish single hysteresis loop strain energy data;

[0007] S3: Calculate the cumulative summation characteristics of the alternating cycle period corresponding to the single hysteresis loop strain energy data to establish the cumulative damage energy parameter, calculate the corresponding ratio of the cumulative damage energy parameter to the critical fracture absorption benchmark, and establish structural health degradation evaluation data.

[0008] S4: Based on the numerical values ​​of the structural health degradation evaluation data, calculate the centripetal offset distance of its product with the oscillation cycle statistical parameters and the single fatigue offset coefficient, calculate the set of reconstructed poles by moving the initial anti-overturning shell model edge pole coordinates to the spatial center coordinates centripetal offset distance, and establish the collapse anti-overturning envelope surface data.

[0009] S5: Analyze the free state of the real-time centroid spatial coordinates detached from the collapse and overturning envelope data, the abnormal deflection state of the lifting space deflection characteristic data exceeding the limit tilt benchmark, and the damage state of the structural health degradation evaluation data exceeding the boundary warning benchmark, calculate the joint risk attributes, and generate lifting safety early warning records.

[0010] As a further embodiment of the present invention, the hoisting space deflection characteristic data specifically includes deflection angle, attitude matrix, and spatial torque; the single hysteresis loop strain energy data includes area integral value, dissipated work, and damping internal loss; the structural health degradation evaluation data specifically refers to fatigue percentage, loss coefficient, and remaining life estimate; the collapse anti-overturning envelope surface data specifically includes reconstructed pole coordinates, contracted three-dimensional surface, and dynamic boundary mesh; and the hoisting safety early warning record includes risk level, alarm code, and timestamp.

[0011] As a further aspect of the present invention, the step of obtaining the hoisting space deflection characteristic data specifically includes:

[0012] S101: Obtain the axial tension parameters and transverse shear parameters at multiple lifting points of the lifting structure. Fill the spatial positions of the matrix elements in the orthogonal dimension for the axial tension parameters and transverse shear parameters. Establish an initial tensile parameter containing tension and shear information. Perform spatial expansion and tensor assembly operations on the initial tensile parameter based on the coordinate basis of the local coordinate system of the lifting structure. Fill the zero values ​​of the unstressed orthogonal axes. Calculate and generate a local tensor matrix with a three-dimensional spatial distribution.

[0013] S102: Collect physical assembly angle data of the surface curing of the lifting structure, extract the installation tilt parameter, extract the roll angle component, pitch angle component and yaw angle component inside the installation tilt parameter, and establish a three-dimensional orthogonal spatial rotation matrix. Perform matrix multiplication mapping operation on the local tensor matrix and the spatial rotation matrix to generate a global tensor matrix.

[0014] S103: Perform eigenvalue decomposition on the global tensor matrix, extract the principal eigenvector corresponding to the largest eigenvalue, obtain the plumb line vector corresponding to the natural gravity direction in the global coordinate system space, perform spatial vector dot product inverse cosine function operation on the principal eigenvector and the plumb line vector, calculate the deviation angle presented by the spatial angle formed by the principal eigenvector and the plumb line vector, and establish hoisting spatial deflection feature data.

[0015] As a further aspect of the present invention, the step of obtaining the single-cycle hysteresis loop strain energy data specifically comprises:

[0016] S201: Collect and analyze the transient force parameters and transient deformation parameters during the lifting process corresponding to the lifting space deflection characteristic data, extract the force timestamp associated with the transient force parameters, extract the deformation timestamp associated with the transient deformation parameters, perform time axis alignment processing, filter the transient force parameters and transient deformation parameters within the synchronous time node, and establish synchronous mechanical deformation data pairs;

[0017] S202: Extract the stress and strain components contained in the synchronous mechanical deformation data pair, map the stress and strain components to a two-dimensional rectangular coordinate system, establish a set of plane coordinate points composed of a continuous time series, perform connection and closure processing on the set of plane coordinate points, and establish a closed geometric boundary jointly enclosed by the lifting loading cycle and the unloading cycle.

[0018] S203: Extract the two-dimensional contour features defined by the closed geometric boundary, perform planar polygon numerical integration on the region inside the closed geometric boundary, calculate the planar area parameter covered by the region inside the closed geometric boundary, analyze the mechanical power dissipation associated with the planar area parameter, perform volume equivalent transformation calculation in combination with the stress section characteristics and deformation characteristics of the lifting structure, and establish single-cycle hysteresis loop strain energy data.

[0019] As a further aspect of the present invention, the process of analyzing the mechanical power dissipation associated with the plane area parameter specifically includes:

[0020] Access the historical test database of the lifting structure and extract the material specifications and material damping loss factor of the lifting structure.

[0021] Based on the material specifications, index matching is performed to extract the corresponding energy equivalence ratio coefficient.

[0022] Collect the geometric parameters of the stress section and the effective deformation step size data of the lifting structure under lifting loading state;

[0023] A preliminary dissipated energy density value is established by multiplying the plane area parameter and the energy equivalence ratio coefficient. A scalar product operation with different unit weights is then performed on the preliminary dissipated energy density value, the geometric parameter of the stressed section, and the effective deformation step size data to generate the physical energy level.

[0024] A weighted ratio compensation calculation is performed on the physical energy level and the material damping loss factor to establish the mechanical power dissipation.

[0025] As a further aspect of the present invention, the steps for obtaining the structural health degradation evaluation data are specifically as follows:

[0026] S301: Extract the timestamp sequence associated with the single hysteresis loop strain energy data, perform extreme point fluctuation scanning calculation on the timestamp sequence, extract the time span between adjacent extreme points in the same direction, define the alternating cycle period of the lifting load dynamic fluctuation over time according to the time span, and establish the load cycle time domain interval.

[0027] S302: Based on the load cycle time domain interval, range matching is performed to extract multiple single hysteresis loop strain energy data within the time period covered by the alternating cycle cycle. For multiple single hysteresis loop strain energy data, continuous accumulation and summation numerical calculation is performed along the time series axis to aggregate the energy loss share caused by each load fluctuation and establish the cumulative damage energy parameter.

[0028] S303: Call the accumulated damage energy parameter, read the critical fracture absorption benchmark in the factory material record of the metal component of the lifting structure, calculate the corresponding ratio of the accumulated damage energy parameter and the critical fracture absorption benchmark, and establish structural health degradation evaluation data.

[0029] As a further aspect of the present invention, the step of obtaining the collapse and overturning prevention envelope data specifically includes:

[0030] S401: Obtain the oscillation cycle statistical parameters within the time interval of the lifting operation environment, and use the structural health degradation evaluation data as an amplification factor to extract the single fatigue offset coefficient from the material parameter record table of the lifting equipment. Perform scalar multiplication numerical operation on the oscillation cycle statistical parameters and the single fatigue offset coefficient to calculate the mapping displacement of the fatigue accumulation effect on the geometric scale and generate the centripetal offset distance.

[0031] S402: Obtain the initial anti-overturning shell model generated by the spatial mapping of the standard drawings of the lifting structure, perform three-dimensional geometric contour point analysis and extraction operation on the initial anti-overturning shell model, obtain the coordinates of multiple edge poles on the outermost contour boundary of the shell model, extract the spatial center coordinates corresponding to the three-dimensional spatial volume geometric center point of the shell model, and establish three-dimensional pole space data that aggregates external features and internal features.

[0032] S403: Call the three-dimensional pole space data, read the edge pole coordinates and the space center coordinates, call the centripetal offset distance, perform linear algebraic coordinate translation operation on the edge pole coordinates along the spatial direction pointing to the space center coordinates, extract the reconstructed pole set generated after translation and shrinkage centripetal offset distance, and establish the collapse anti-overturning envelope surface data based on the surface enclosed by the spatial connection of all vertices in the reconstructed pole set.

[0033] As a further aspect of the present invention, the process of obtaining the initial anti-overturning shell model generated by spatial mapping of the standard drawings of the lifting structure is specifically as follows:

[0034] Access the lifting equipment design database and retrieve the standard drawings and physical dimension configuration table of the lifting structure;

[0035] A two-dimensional vector analysis operation is performed on the standard drawings of the lifting structure to extract the two-dimensional projection line segment data in the standard drawings and establish a set of the outer plane contour of the lifting structure.

[0036] Read the physical dimension configuration table and extract the tensile thickness parameter along the orthogonal spatial axis;

[0037] A three-dimensional basic solid geometric data is established by performing a stretching transformation operation based on the stretching thickness parameter along the orthogonal spatial axis on the lifting outer plane contour set.

[0038] The surface mesh is processed on the geometric data of the three-dimensional basic entity to extract polygonal facet elements exposed in the external space region;

[0039] The polygonal facet elements are subjected to coplanar merging and convex hull mesh assembly calculations to generate a three-dimensional topological surface, which is then identified as the initial anti-overturning shell model.

[0040] As a further aspect of the present invention, the step of obtaining the lifting safety early warning record specifically includes:

[0041] S501: Obtain the three-dimensional coordinate information of the lifting equipment at the current time node, extract the real-time centroid spatial coordinates, call the collapse and overturning envelope data, determine the spatial inclusion relationship between the real-time centroid spatial coordinates and the collapse and overturning envelope data, and establish a free state based on the relative position relationship of the real-time centroid spatial coordinates outside the spatial boundary.

[0042] S502: Read the limit tilt reference and warning reference in the preset configuration table, perform numerical comparison calculation on the lifting space deflection feature data and the limit tilt reference, determine the abnormal deflection state that exceeds the threshold range, perform the boundary value judgment operation on the structural health degradation evaluation data and the warning reference, generate the damage state, combine the free state, abnormal deflection state and damage state to establish multi-dimensional abnormal judgment data.

[0043] S503: Based on the multidimensional anomaly determination data, extract the free state, abnormal deflection state and damage state to construct a multidimensional evaluation matrix, read the preset risk weight configuration data, and perform weighted aggregation operation on the free state, abnormal deflection state and damage state in the multidimensional evaluation matrix to extract the joint risk attribute of multiple abnormal features, match the corresponding risk code and timestamp information according to the joint risk attribute, and generate a lifting safety early warning record.

[0044] A road roller lifting safety assessment system based on load data includes:

[0045] The deflection feature extraction module constructs a local tensor matrix and a spatial rotation matrix based on the initial tension parameter and the installation tilt parameter, respectively. It calculates the product of the local tensor matrix and the spatial rotation matrix to generate a global tensor matrix, and extracts the principal eigenvector of the global tensor matrix and the deviation angle of the plumb line to establish hoisting spatial deflection feature data.

[0046] The hysteresis energy calculation module filters the lifting transient force parameters and transient deformation parameters of the synchronous time node of the lifting space deflection characteristic data, calculates the plane area parameter covered inside the closed geometric boundary of the paired plane coordinate point set, and establishes single hysteresis loop strain energy data.

[0047] The degradation cumulative analysis module calculates the cumulative summation characteristics of the alternating cycle period corresponding to the single hysteresis loop strain energy data to establish the cumulative damage energy parameter, calculates the corresponding ratio of the cumulative damage energy parameter to the critical fracture absorption benchmark, and establishes structural health degradation evaluation data.

[0048] The envelope surface collapse reconstruction module calculates the centripetal offset distance of the product of the structural health degradation evaluation data and the oscillation cycle statistical parameters and the single fatigue offset coefficient, calculates the set of reconstruction poles by moving the initial anti-overturning shell model edge pole coordinates to the spatial center coordinates by centripetal offset distance, and establishes the collapse anti-overturning envelope surface data.

[0049] The joint risk early warning module analyzes the free state of the real-time centroid spatial coordinates deviating from the collapse and overturning envelope data, the abnormal deflection state of the lifting space deflection characteristic data exceeding the limit tilt benchmark, and the damage state of the structural health degradation evaluation data exceeding the boundary warning benchmark. It calculates the joint risk attributes and generates lifting safety early warning records.

[0050] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0051] In this invention, the angle of gravity direction offset is analyzed by tensor rotation product operation, and the hysteresis loop energy area is calculated by closure integral of transient stress and deformation data. The accumulated damage model is constructed by aggregating the dissipated energy within the alternating cycle. The boundary of the anti-overturning three-dimensional envelope surface is reconstructed by centripetal contraction combined with fatigue degradation characteristics. Multidimensional dynamic early warning is completed by coordinating the centroid free deviation and damage over-limit judgment. The dynamic quantification of spatial force evolution is realized through mechanical mapping. The transient load fluctuation is transformed into long-term degradation index by using strain energy equation. The static accounting model is replaced by dynamic contraction boundary, which blocks the overturning chain of the whole machine hidden under continuous alternating oscillation in the single-dimensional force deduction. Attached Figure Description

[0052] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0053] Figure 1 This is a schematic diagram of the steps of the present invention;

[0054] Figure 2 This is a detailed schematic diagram of S1 of the present invention;

[0055] Figure 3 This is a detailed schematic diagram of S2 of the present invention;

[0056] Figure 4 This is a detailed schematic diagram of S3 of the present invention;

[0057] Figure 5 This is a detailed schematic diagram of S4 of the present invention;

[0058] Figure 6 This is a detailed schematic diagram of S5 of the present invention;

[0059] Figure 7 This is a system module diagram of the present invention. Detailed Implementation

[0060] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0061] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0062] Please see Figure 1 This invention provides a method for safety assessment of road roller lifting based on load data, including the following steps:

[0063] S1: Based on the initial tension parameters and installation tilt parameters, construct local tensor matrices and spatial rotation matrices respectively. Calculate the product of the local tensor matrix and the spatial rotation matrix to generate a global tensor matrix. Extract the principal eigenvectors of the global tensor matrix and the deviation angle of the plumb line to establish hoisting spatial deflection feature data.

[0064] S2: Filter the lifting space deflection characteristic data synchronous time node lifting transient force parameters and transient deformation parameters, calculate the plane area parameter covered inside the closed geometric boundary of the paired plane coordinate point set, and establish single hysteresis loop strain energy data;

[0065] S3: Calculate the cumulative summation characteristics of the alternating cycle period corresponding to the single hysteresis loop strain energy data to establish the cumulative damage energy parameter, calculate the corresponding ratio of the cumulative damage energy parameter to the critical fracture absorption benchmark, and establish structural health degradation evaluation data.

[0066] S4: Based on the numerical values ​​of the structural health degradation evaluation data, calculate the centripetal offset distance of its product with the oscillation cycle statistical parameters and the single fatigue offset coefficient, calculate the set of reconstructed poles by moving the initial anti-overturning shell model edge pole coordinates to the spatial center coordinates centripetal offset distance, and establish the collapse anti-overturning envelope surface data.

[0067] S5: Analyze the free state of real-time centroid spatial coordinates detached from the collapse and overturning protection envelope data, the abnormal deflection state of lifting space deflection characteristic data exceeding the limit tilt benchmark, and the damage state of structural health degradation evaluation data exceeding the boundary warning benchmark, calculate the joint risk attributes, and generate lifting safety early warning records.

[0068] The specific data on hoisting space deflection characteristics include deflection angle, attitude matrix, and spatial torque. The single-cycle hysteresis loop strain energy data includes area integral value, dissipated work, and damping internal loss. The specific data on structural health degradation evaluation refers to fatigue percentage, loss coefficient, and estimated remaining life. The specific data on collapse and overturning prevention envelope includes reconstructed pole coordinates, contracted three-dimensional surface, and dynamic boundary mesh. The hoisting safety early warning record includes risk level, alarm code, and timestamp.

[0069] Please see Figure 2 The specific steps for obtaining the lifting space deflection characteristic data are as follows:

[0070] S101: Obtain the axial tension parameters and transverse shear parameters at multiple lifting points of the lifting structure. Fill the spatial positions of the matrix elements in the orthogonal dimension for the axial tension parameters and transverse shear parameters. Establish an initial tensile parameter containing tension and shear information. Perform spatial expansion and tensor assembly operations on the initial tensile parameter based on the coordinate basis of the local coordinate system of the lifting structure. Fill the zero values ​​of the unstressed orthogonal axes. Calculate and generate a local tensor matrix with a three-dimensional spatial distribution.

[0071] By deploying a high-precision foil strain gauge network and triaxial force sensors at four key lifting points of the road roller's lifting structure, mechanical simulation electrical signals during the lifting operation are acquired in real time at a continuous sampling frequency of 500 Hz. These signals are then converted into a high-precision digital signal sequence via a transmitter with 24-bit analog-to-digital conversion resolution. From this sequence, the axial tensile and transverse shear parameters at each lifting point at the current moment are extracted. Once the specific stress values ​​are obtained, for example, if the measured axial tensile force at a primary load-bearing lifting point is 45,000 Newtons and the transverse shear force is 1,200 Newtons, the corresponding spatial positions of the matrix elements in the orthogonal dimension are immediately filled for these axial tensile and transverse shear parameters. Specifically, a 2x2 two-dimensional array space is first allocated in memory. The 45,000 Newtons are filled into the first row and first column of this array as the tensile normal stress component of the principal axis, and the 1,200 Newtons are filled into the first row and second column and the second row and first column as symmetrical shear stress components, respectively, thus establishing the initial tensile parameters containing underlying tensile and shear information. Next, the program reads the three orthogonal basis vectors of the pre-calibrated local coordinate system of the lifting structure. Based on these basis vectors, it performs spatial expansion and tensor assembly operations on the initial tension parameters, projecting the feature elements originally confined to the two-dimensional force plane into a complete three-dimensional Cartesian coordinate system. During the projection mapping operation, the program verifies the actual load distribution on each of the three spatial orthogonal axes and performs zero-value completion on orthogonal axes confirmed by sensors to be unaffected by external loads. For example, if the external force value along the local Z-axis is determined to be less than the dead zone threshold of 5 Newtons, all corresponding elements in the third row and third column of the expanded 3x3 matrix are assigned the value 0. After this standardized dimensional expansion and zero-filling calculation, a local tensor matrix containing the complete three-dimensional force situation and presenting a three-dimensional spatial distribution is finally generated.

[0072] S102: Collect physical assembly angle data of the surface curing of the lifting structure, extract the installation tilt parameter, extract the roll angle component, pitch angle component and yaw angle component inside the installation tilt parameter, and establish a three-dimensional orthogonal spatial rotation matrix. Perform matrix multiplication mapping operation on the local tensor matrix and the spatial rotation matrix to generate a global tensor matrix.

[0073] Using a microelectromechanical inertial measurement unit (MEMS) with an IP68 protection rating, permanently installed on the steel surface of the roller lifting structure, physical assembly angle data, including gravitational acceleration components and geomagnetic reference vectors, is continuously acquired at a 100 Hz update rate. Quaternion-based algorithms are used to perform low-pass filtering and attitude fusion on the acquired raw attitude signals to extract the installation tilt parameters under the current spatial pose. Furthermore, from these installation tilt parameters, roll, pitch, and yaw components based on the geographic northeast coordinate system are extracted. Assuming the extracted roll angle is 5 degrees, pitch angle is 12 degrees, and yaw angle is 2 degrees, these three angular components are substituted into the classic 3D spatial coordinate rotation formula to calculate the cosine and sine trigonometric function values ​​around the X, Y, and Z axes, respectively. The calculated decimal results are then assembled into three independent single-axis rotation matrices according to the derived definition of the spatial rotation matrix. These three single-axis rotation matrices are then multiplied together in a fixed order of yaw, pitch, and roll to establish a 3D orthogonal spatial rotation matrix that integrates the pose changes of the three degrees of freedom. After obtaining this 3x3 spatial rotation matrix, the 3x3 local tensor matrix output from the previous step is immediately called. In the digital signal processor, a standard matrix multiplication mapping operation is performed between the local tensor matrix and the spatial rotation matrix. During the operation, the spatial rotation matrix is ​​used as the left multiplication operator and the local tensor matrix is ​​used as the right multiplication operator. The numerical reconstruction of the nine elements is completed by the multiplication and addition instructions of the row and column inner product. The mechanical tensor originally attached to the local reference system of the device is effectively transformed to the geodetic coordinate system, generating a global tensor matrix with global spatial reference significance.

[0074] S103: Perform eigenvalue decomposition on the global tensor matrix, extract the principal eigenvector corresponding to the largest eigenvalue, obtain the plumb line vector corresponding to the natural gravity direction in the global coordinate system space, perform spatial vector dot product inverse cosine function operation on the principal eigenvector and the plumb line vector, calculate the deviation angle presented by the spatial angle formed by the principal eigenvector and the plumb line vector, and establish hoisting space deflection feature data.

[0075] For the 3x3 global tensor matrix generated by the above calculation, the Jacobi iteration algorithm or QR decomposition algorithm is called to perform matrix eigenvalue decomposition. In the operation loop, off-diagonal elements of the matrix are continuously eliminated through orthogonal similarity transformations until the values ​​of all off-diagonal elements are less than the convergence accuracy threshold of 0.0001. At this point, the three values ​​on the main diagonal of the matrix are the eigenvalues ​​of the tensor matrix. These three eigenvalues ​​are compared, and the largest eigenvalue is identified. The normalized column vector corresponding to this largest eigenvalue is extracted and defined as the principal eigenvector. This vector physically represents the global geometric direction in which the lifting structure experiences the most concentrated force and the strongest deformation trend in the current global space. Simultaneously, the underlying geographic environment configuration constants are accessed to obtain the vertically downward vector corresponding to the direction of gravity pointing towards the Earth's center in the global coordinate system space. This vector is typically normalized in the Cartesian coordinate system as a direction vector composed of coordinate points 0, 0, and -1. Subsequently, in the three-dimensional vector space, the inverse cosine function operation of the spatial vector dot product is performed on the extracted principal feature vector and the standard perpendicular line vector. Specifically, the scalar algebraic sum of the corresponding coordinate components of the principal feature vector and the perpendicular line vector is first calculated. Then, this scalar algebraic sum is divided by the product of the magnitudes of the two vectors to obtain the cosine value of the angle. Finally, the inverse cosine function is used to nonlinearly solve this cosine value, calculating the specific deviation angle of the angle formed by the principal feature vector and the perpendicular line vector in the real three-dimensional space. This deviation angle value quantifies the severity of the lifting center of gravity shift and force imbalance. It is packaged and stored in a cache along with the corresponding sampling time points to establish lifting spatial deflection feature data.

[0076] Please see Figure 3 The specific steps for obtaining single-cycle hysteresis loop strain energy data are as follows:

[0077] S201: Collect and analyze the transient force parameters and transient deformation parameters during the lifting process corresponding to the spatial deflection characteristic data of the lifting process, extract the force timestamp associated with the transient force parameters, extract the deformation timestamp associated with the transient deformation parameters, perform time axis alignment processing, filter the transient force parameters and transient deformation parameters that are within the synchronous time nodes, and establish synchronous mechanical deformation data pairs;

[0078] The system monitors and intercepts the lifting status stream transmitted on the underlying data bus in real time, collects and analyzes various high-frequency sensor sequences during the lifting process corresponding to the lifting spatial deflection characteristic data, and focuses on extracting the lifting transient force parameters and transient deformation parameters. During the data parsing phase, the header of each incoming load data packet is decoded, and the 64-bit high-precision global positioning force timestamp attached to the end of the lifting transient force parameters is extracted. Similarly, the messages returned by the laser displacement sensor or strain grid are decoded, and the deformation timestamp associated with the transient deformation parameters is extracted. Since the hardware sampling clocks of the mechanical sensors and displacement sensors often have a physical time difference at the microsecond level, a bidirectional buffer queue is created in memory for time axis alignment. A time alignment sliding window with a tolerance of 10 milliseconds is set. Using the force timestamp as the reference axis, linear interpolation or nearest-neighbor matching is performed on the deformation timestamp sequence within the sliding window. Isolated data points that lag behind or precede the window are removed, and the lifting transient force parameters and transient deformation parameters within the synchronized time node are selected. For example, when the force parameter is recorded as a tensile force of 50,000 Newtons at timestamp 1625098700.150, a deformation parameter of 12 mm recorded at the same timestamp 1625098700.150 is found. These force and deformation data are encapsulated in a structure to construct synchronous mechanical deformation data pairs, providing a verification variable correspondence for subsequent hysteresis loop integral calculations.

[0079] S202: Extract the stress and strain components contained in the synchronous mechanical deformation data pair, map the stress and strain components to a two-dimensional rectangular coordinate system, establish a set of plane coordinate points composed of continuous time series, perform connection and closure processing on the set of plane coordinate points, and establish a closed geometric boundary jointly enclosed by the lifting loading cycle and the unloading cycle.

[0080] The memory encapsulation format of synchronous mechanical deformation data pairs is deconstructed to extract the stress components characterizing the internal stress of the structure and the strain components characterizing the relative rate of change of physical dimensions. In the software drawing engine or numerical analysis matrix, the extracted stress and strain components are mapped to a standard two-dimensional Cartesian coordinate system, explicitly specifying that the strain component values ​​are used as the horizontal X-axis coordinate and the stress component values ​​as the vertical Y-axis coordinate. As the lifting operation continues, the continuously extracted coordinate pairs are successively calibrated on the two-dimensional plane, establishing a set of planar coordinate points consisting of more than 1000 consecutive dense time series. After a complete cycle of lifting and hoisting by the hoist and subsequent hovering and unloading, the trajectory of the coordinate point set is monitored, and a convex hull algorithm or trajectory addressing logic is used to connect the planar coordinate point set to its beginning and end for closure. Specifically, the last coordinate point at the end of the unloading stage is connected to the first coordinate point at the beginning of the loading stage by linear interpolation, thus establishing a closed geometric boundary on the two-dimensional coordinate plane formed by the rising segment of the lifting loading cycle curve and the falling segment of the unloading cycle curve.

[0081] S203: Extract the two-dimensional contour features defined by the closed geometric boundary, perform numerical integration of the planar polygons for the region inside the closed geometric boundary, calculate the planar area parameter covered by the region inside the closed geometric boundary, analyze the mechanical power dissipation associated with the planar area parameter, perform volume equivalent transformation calculation by combining the stress section characteristics and deformation characteristics of the lifting structure, and establish single hysteresis loop strain energy data.

[0082] A polygon vertex traversal algorithm is used to extract the two-dimensional contour features defined by the closed geometric boundary, obtaining the coordinate sequence of all discrete vertices constituting the closed loop. For the region inside the closed geometric boundary, a planar polygon numerical integration operation based on Green's theorem is employed. This involves calculating the planar area parameter covered by the region inside the closed geometric boundary by taking half the sum of the cross products of the coordinates of two adjacent vertices. The mechanical energy dissipation associated with this planar area parameter in the laws of mechanics and thermodynamics is analyzed. This area is numerically equivalent to the energy lost as heat due to internal microcrystalline lattice friction during a single load cycle. To convert this two-dimensional mathematical area into a physical energy value with practical engineering guidance, a volume equivalent transformation calculation is performed, combining the stress-bearing cross-sectional characteristics and deformation characteristics of the lifting structure. Specifically, the solid cross-sectional area of ​​the lifting connector (e.g., 0.05 square meters) and the initial stress length (e.g., 2 meters) are read. The calculated two-dimensional planar area parameter is multiplied by the volume of the structural component (the product of the cross-sectional area and length, 0.1 cubic meters), and a dimensional scaling and volume equivalent transformation calculation is performed. This multiplication operation converts the dissipated energy per unit volume into the total macroscopic energy loss of the entire lifting load-bearing component, ultimately establishing single-cycle hysteresis loop strain energy data.

[0083] Please see Figure 4 The specific steps for obtaining structural health degradation assessment data are as follows:

[0084] S301: Extract the timestamp sequence associated with the single hysteresis loop strain energy data, perform extreme point fluctuation scanning calculation on the timestamp sequence, extract the time span between adjacent extreme points in the same direction, define the alternating cycle period of the lifting load dynamic fluctuation over time based on the time span, and establish the load cycle time domain interval.

[0085] Continuously monitor and store multiple energy data accumulated over time, and extract the continuous timestamp sequence associated with the single hysteresis loop strain energy data in the database. Perform extreme point fluctuation scanning calculations based on the first derivative on the energy fluctuation curve corresponding to this timestamp sequence, detecting local maxima and minima in each data sequence. A peak is marked when the derivative sign changes from positive to negative, and a trough when it changes from negative to positive. Extract the time span between two adjacent extreme points in the same direction (e.g., from one peak to the next, or from one trough to the next). Based on this calculated time span value (e.g., an interval of 4.5 seconds), define the alternating cycle period of the lifting load dynamically fluctuating over time due to wind load disturbance or winch vibration while the roller is suspended in a lifting state. Anchor this 4.5-second cycle interval, which includes complete loading and unloading actions, on the operating time axis to establish the load cycle time domain interval used to define the single fatigue failure domain.

[0086] S302: Based on the load cycle time domain interval, range matching is performed to extract multiple single hysteresis loop strain energy data within the time period covered by the alternating cycle cycle. For multiple single hysteresis loop strain energy data, continuous accumulation and summation numerical calculations are performed along the time series axis to aggregate the energy loss share caused by each load fluctuation and establish the cumulative damage energy parameter.

[0087] Based on the load cycle time domain interval determined above, the database query range is matched, the time constraint conditions of the SQL retrieval command are set, and dozens or even hundreds of single hysteresis loop strain energy data within the time period covered by the alternating cycle are extracted.

[0088] Table 1. Record of Cumulative Energy Loss During Cycles

[0089]

[0090] For the extracted strain energy data of multiple single-cycle hysteresis loops listed in Table 1, a continuous cumulative summation operation along the time series axis is performed. In the processor's accumulation register, the energy loss values ​​corresponding to each load cycle (e.g., 450.2 joules for cycle 1, 465.8 joules for cycle 2, etc.) are linearly superimposed, representing the energy loss share caused by each load fluctuation during the continuous lifting and suspension process of the deep-aggregate roller. This accumulated value shows a monotonically increasing trend as the lifting time increases. Finally, the total value containing all historical operational damage memories is persistently stored, establishing a cumulative damage energy parameter characterizing the driving force of microcrack initiation and propagation within the metal component, thereby transforming the abstract fatigue degradation process into a quantified macroscopic energy scale.

[0091] S303: Call the cumulative damage energy parameter, read the critical fracture absorption benchmark in the factory material record of the metal components of the lifting structure, calculate the corresponding ratio of the cumulative damage energy parameter and the critical fracture absorption benchmark, and establish structural health degradation evaluation data.

[0092] The system retrieves the accumulated damage energy parameter residing in the memory stack, accesses the equipment manufacturer's cloud database via an Industrial Internet of Things (IIoT) interface, and reads the critical fracture absorption benchmark specified in the factory material record of the specific metal component of the corresponding road roller lifting structure. This benchmark value represents the maximum total energy that a specific grade of special steel can absorb before macroscopic fracture failure, for example, a factory-specified value of 500,000 joules. In the controller's floating-point arithmetic unit, the accumulated damage energy parameter obtained in real time (e.g., 150,000 joules) is used as the numerator, and the obtained critical fracture absorption benchmark is used as the denominator. A division instruction is executed to calculate the corresponding ratio between the two. The calculated ratio (i.e., 150,000 divided by 500,000 equals 0.3, or 30%) is assigned to engineering evaluation significance, establishing structural health degradation evaluation data.

[0093] Please see Figure 5 The specific steps for obtaining the collapse and overturning protection envelope data are as follows:

[0094] S401: Obtain the oscillation cycle statistical parameters within the time interval of the lifting operation environment, and use the structural health degradation evaluation data as an amplification factor to extract the single fatigue offset coefficient from the material parameter record table of the lifting equipment. Perform scalar multiplication numerical calculation on the oscillation cycle statistical parameters and the single fatigue offset coefficient to calculate the mapping displacement of the fatigue accumulation effect on the geometric scale and generate the centripetal offset distance.

[0095] Based on the structural health degradation evaluation data calculated above, the statistical parameters of oscillation cycles recorded within the time interval of the road roller lifting operation environment are further extracted. These parameters specifically record the number of micro-oscillations caused by wind and mechanical resonance affecting the equipment in the air, for example, a cumulative oscillation cycle of 8500. Simultaneously, the single-cycle fatigue offset coefficient from the material parameter record table of the lifting equipment is retrieved via a file reading protocol. This coefficient, determined by the materials mechanics laboratory, characterizes the permanent geometric creep rate caused by a single micro-oscillation on material deformation, for example, calibrated to 0.005 mm per cycle. In the arithmetic logic unit of the central processing unit, a standard scalar multiplication numerical operation is performed on the extracted statistical parameters of 8500 oscillation cycles and the 0.005 mm single-cycle fatigue offset coefficient. Through this multiplication analysis, the mapped displacement continuously generated by the long-term fatigue accumulation effect on the overall geometric scale of the equipment is calculated, i.e., 8500 multiplied by 0.005 equals 42.5 mm. The value of 42.5 mm is defined and a centripetal offset distance is generated. This distance physically means that due to long-term fatigue damage, the original initial safety boundary of the equipment needs to be retreated by 42.5 mm towards the core area to ensure lifting safety.

[0096] S402: Obtain the initial anti-overturning shell model generated by the spatial mapping of the standard drawings of the lifting structure, perform three-dimensional geometric contour point analysis and extraction operation on the initial anti-overturning shell model, obtain the coordinates of multiple edge poles on the outermost contour boundary of the shell model, extract the spatial center coordinates corresponding to the three-dimensional spatial volume geometric center point of the shell model, and establish three-dimensional pole space data that aggregates external features and internal features.

[0097] By reading the standard 3D point cloud file exported by computer-aided design, an initial anti-overturning shell model was generated by spatial mapping of the standard drawings for the road roller lifting structure. For this initial anti-overturning shell model, a polygonal mesh 3D geometric contour point extraction operation was performed. All triangular facet vertices on the model surface were traversed, and coordinate extreme value comparison algorithms (such as finding the maximum and minimum points in the X, Y, and Z axes) were used to obtain the coordinates of multiple key edge poles on the outermost contour boundary of the shell model. Simultaneously, based on the 3D volume step-by-step integration formula, the centroid of the entire initial model was solved under the assumption of uniform mass, extracting the effective spatial center coordinates corresponding to the 3D spatial volume geometric center point of the shell model. The extracted outer edge pole coordinate sequence and the unique spatial center coordinates were packaged and bound in a data structure to establish 3D pole spatial data aggregating external polygonal boundary features and internal geometric core features, providing spatial topological reference anchor points for subsequent coordinate centripetal translation.

[0098] S403: Call the three-dimensional pole space data, read the coordinates of the edge poles and the coordinates of the spatial center, call the centripetal offset distance, perform linear algebraic coordinate translation operation on the edge pole coordinates along the spatial direction pointing to the spatial center coordinates, extract the reconstructed pole set generated after translation and shrinking centripetal offset distance, and establish the collapse anti-overturning envelope surface data based on the surface enclosed by the spatial connection of all vertices in the reconstructed pole set.

[0099] The system retrieves the 3D pole space data from memory, reading the coordinates of each edge pole and the unique coordinates of the spatial center line by line, while simultaneously retrieving the previously calculated centripetal offset distance of 42.5 mm. For each edge pole coordinate, the system calculates the 3D direction vector between it and the spatial center coordinates, and then normalizes this direction vector to a unit.

[0100] Table 2. Centripetal Reconstruction of Pole Coordinates

[0101]

[0102] For the edge pole coordinates, a standard coordinate translation operation is performed along the line connecting the normalized coordinates pointing to the spatial center coordinates using linear algebra matrix addition, as shown in Table 2. This moves the initial pole 42.5 mm towards the center point along the line. This translation logic is executed iteratively, extracting all poles after translation, shrinkage, and centripetal offset distances to generate a new 3D coordinate array, thus establishing a reconstructed pole set. Subsequently, a 3D convex hull reconstruction algorithm is called. Based on the 3D spatial connections of all vertices in the reconstructed pole set, facet stitching and triangulation are performed again to enclose and generate a smaller, more conservative surface structure, which is then persistently established as the collapse-resistant overturning envelope data.

[0103] Please see Figure 6 The specific steps for obtaining lifting safety warning records are as follows:

[0104] S501: Obtain the three-dimensional coordinate information of the lifting equipment at the current time node, extract the real-time centroid spatial coordinates, call the collapse and overturning envelope data, determine the spatial inclusion relationship between the real-time centroid spatial coordinates and the collapse and overturning envelope data, and establish the free state based on the relative position relationship of the real-time centroid spatial coordinates outside the spatial boundary.

[0105] The system acquires the global coordinates of the main control board of the road roller lifting equipment at the current millisecond time point, filters out environmental noise, and extracts the real-time centroid spatial coordinates, which characterize the core of the machine's gravity distribution. Next, it retrieves the collapse-prevention envelope data with retreat boundary attributes generated in the previous step from the solid-state drive. In the graphics processor's compute shader, a high-concurrency ray penetration algorithm is used to determine the spatial inclusion relationship between the real-time centroid spatial coordinates (containing only X, Y, and Z dimensions) and the collapse-prevention envelope data composed of patches. A virtual ray is emitted from the real-time centroid coordinates to any spatial direction, and the total number of intersection points between the ray and the envelope is counted. If the number of intersection points is odd, the main control unit determines that the centroid is still inside the safe envelope; if the number of intersection points is even (or 0), the relative positional relationship based on the real-time centroid spatial coordinates being outside the spatial boundary is established. Once the external out-of-bounds condition is triggered, the main control program immediately sets the relevant flag to 1 in the status register, formally establishing the free state in which the device may become unstable and overturn at any time.

[0106] S502: Read the limit tilt reference and warning reference in the preset configuration table, perform numerical comparison calculation on the lifting space deflection characteristic data and the limit tilt reference, determine the abnormal deflection state that exceeds the threshold range, perform out-of-bounds numerical judgment operation on the structural health degradation evaluation data and the warning reference, generate damage state, combine free state, abnormal deflection state and damage state, and establish multi-dimensional abnormal judgment data.

[0107] The system reads the limit tilt reference (e.g., 10 degrees) and warning reference (e.g., 60% fatigue consumption rate) from the preset configuration table of the industrial control host. In the main process's logical judgment branch, a numerical comparison calculation is performed on the previously continuously updated hoisting space deflection characteristic data (e.g., measured deviation angle of 12 degrees) and the set limit tilt reference. When it is determined that the measured 12 degrees has exceeded the threshold range of 10 degrees, an alarm logic is triggered, indicating an abnormal deflection state. Under synchronous parallel operation, an out-of-bounds numerical judgment operation is performed on the acquired structural health degradation evaluation data (e.g., the current ratio is 65%) and the 60% warning reference. Since 65% is greater than 60%, the processor generates a damage state indicating that the material is at the edge of high-risk fracture. Subsequently, on the data aggregation bus, the three Boolean variables representing the centroid exceeding the limit, the abnormal deflection state representing attitude imbalance, and the damage state representing the material being endangered are combined bit by bit and packaged into a binary data frame containing 3 feature bits, establishing multidimensional anomaly judgment data for subsequent high-dimensional adjudication.

[0108] S503: Based on multidimensional anomaly judgment data, extract free state, abnormal deflection state and damage state to construct a multidimensional evaluation matrix, read the preset risk weight configuration data, and perform weighted aggregation operation on free state, abnormal deflection state and damage state in the multidimensional evaluation matrix to extract the joint risk attribute of multiple abnormal features, match the corresponding risk code and timestamp information according to the joint risk attribute, and generate a lifting safety early warning record.

[0109] Based on the multidimensional anomaly determination data in the buffer queue, the execution unit extracts discrete quantities representing the detached state, the anomaly deflection state, and the damage state, and constructs a multidimensional evaluation matrix in the form of a 1x3 row vector in memory space. Subsequently, it reads the preset risk weight configuration data from the security expert database. This data consists of a 3x1 column vector, which corresponds to the hazard penalty weights of the three states (e.g., detached state weight 40, deflection state weight 35, and damage state weight 25).

[0110] Table 3 Joint Risk Attribute Determination Table

[0111]

[0112] As shown in Table 3, the matrix operation unit performs standard weighted aggregation operations (i.e., matrix multiplication and addition) on the free state, abnormal deflection state, and damage state (with values ​​of 0 or 1) within the multidimensional evaluation matrix and the weight column vector. This results in a scalar score ranging from 0 to 100, which is then extracted as a joint risk attribute, aggregating the severity of multiple abnormal features. Finally, using a lookup command, the corresponding severity level risk code is matched in the database dictionary based on the specific numerical range of this joint risk attribute. The international standard timestamp information of the currently running program is appended, and the data is concatenated into a 128-byte standardized message to generate a lifting safety warning record. This record is then pushed to the handheld terminal interface of the on-site operator via industrial Ethernet.

[0113] Please see Figure 7 A road roller lifting safety assessment system based on load data includes:

[0114] The deflection feature extraction module constructs a local tensor matrix and a spatial rotation matrix based on the initial tension parameter and the installation tilt parameter, respectively. It calculates the product of the local tensor matrix and the spatial rotation matrix to generate a global tensor matrix, and extracts the principal eigenvector of the global tensor matrix and the deviation angle of the plumb line to establish hoisting spatial deflection feature data.

[0115] The hysteresis energy calculation module filters the lifting transient force parameters and transient deformation parameters of the synchronous time node of the lifting space deflection characteristic data, calculates the plane area parameter covered inside the closed geometric boundary of the paired plane coordinate point set, and establishes single hysteresis loop strain energy data.

[0116] The degradation cumulative analysis module calculates the cumulative summation characteristics of the alternating cycle period corresponding to the strain energy data of a single hysteresis loop to establish the cumulative damage energy parameter, calculates the corresponding ratio of the cumulative damage energy parameter to the critical fracture absorption benchmark, and establishes structural health degradation evaluation data.

[0117] The envelope surface collapse reconstruction module calculates the centripetal offset distance of the product of the structural health degradation evaluation data and the oscillation cycle statistical parameters and the single fatigue offset coefficient, calculates the set of reconstruction poles by moving the initial anti-overturning shell model edge pole coordinates to the spatial center coordinates by centripetal offset distance, and establishes the collapse anti-overturning envelope surface data.

[0118] The joint risk early warning module analyzes the free state of real-time centroid spatial coordinates detached from the collapse and overturning prevention envelope data, the abnormal deflection state of lifting space deflection characteristic data exceeding the limit tilt benchmark, and the damage state of structural health degradation evaluation data exceeding the boundary warning benchmark. It calculates joint risk attributes and generates lifting safety early warning records.

[0119] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for evaluating the safety of a road roller hoisting based on load data, characterized by, Includes the following steps: S1: Based on the initial tension parameters and installation tilt parameters, construct local tensor matrices and spatial rotation matrices respectively. Calculate the product of the local tensor matrix and the spatial rotation matrix to generate a global tensor matrix. Extract the principal eigenvectors of the global tensor matrix and the deviation angle of the plumb line to establish hoisting spatial deflection feature data. S2: Filter the lifting transient force parameters and transient deformation parameters of the synchronous time node of the lifting space deflection characteristic data, calculate the plane area parameter covered inside the closed geometric boundary of the paired plane coordinate point set, and establish single hysteresis loop strain energy data; S3: Calculate the cumulative summation characteristics of the alternating cycle period corresponding to the single hysteresis loop strain energy data to establish the cumulative damage energy parameter, calculate the corresponding ratio of the cumulative damage energy parameter to the critical fracture absorption benchmark, and establish structural health degradation evaluation data. S4: Based on the numerical values ​​of the structural health degradation evaluation data, calculate the centripetal offset distance of its product with the oscillation cycle statistical parameters and the single fatigue offset coefficient, calculate the set of reconstructed poles by moving the initial anti-overturning shell model edge pole coordinates to the spatial center coordinates centripetal offset distance, and establish the collapse anti-overturning envelope surface data. S5: Analyze the free state of the real-time centroid spatial coordinates detached from the collapse and overturning envelope data, the abnormal deflection state of the lifting space deflection characteristic data exceeding the limit tilt benchmark, and the damage state of the structural health degradation evaluation data exceeding the boundary warning benchmark, calculate the joint risk attributes, and generate lifting safety early warning records.

2. The method of claim 1, wherein, The hoisting space deflection characteristic data specifically includes deflection angle, attitude matrix, and spatial torque; the single hysteresis loop strain energy data includes area integral value, dissipated work, and damping internal loss; the structural health degradation evaluation data specifically refers to fatigue percentage, loss coefficient, and estimated remaining life; the collapse anti-overturning envelope data specifically includes reconstructed pole coordinates, contracted three-dimensional surface, and dynamic boundary mesh; and the hoisting safety early warning record includes risk level, alarm code, and timestamp.

3. The method for safety assessment of roller lifting based on load data according to claim 1, characterized in that, The specific steps for obtaining the hoisting space deflection characteristic data are as follows: S101: Obtain the axial tension parameters and transverse shear parameters at multiple lifting points of the lifting structure. Fill the spatial positions of the matrix elements in the orthogonal dimension for the axial tension parameters and transverse shear parameters. Establish an initial tensile parameter containing tension and shear information. Perform spatial expansion and tensor assembly operations on the initial tensile parameter based on the coordinate basis of the local coordinate system of the lifting structure. Fill the zero values ​​of the unstressed orthogonal axes. Calculate and generate a local tensor matrix with a three-dimensional spatial distribution. S102: Collect physical assembly angle data of the surface curing of the lifting structure, extract the installation tilt parameter, extract the roll angle component, pitch angle component and yaw angle component inside the installation tilt parameter, and establish a three-dimensional orthogonal spatial rotation matrix. Perform matrix multiplication mapping operation on the local tensor matrix and the spatial rotation matrix to generate a global tensor matrix. S103: Perform eigenvalue decomposition on the global tensor matrix, extract the principal eigenvector corresponding to the largest eigenvalue, obtain the plumb line vector corresponding to the natural gravity direction in the global coordinate system space, perform spatial vector dot product inverse cosine function operation on the principal eigenvector and the plumb line vector, calculate the deviation angle presented by the spatial angle formed by the principal eigenvector and the plumb line vector, and establish hoisting spatial deflection feature data.

4. The method for safety assessment of roller lifting based on load data according to claim 3, characterized in that, The specific steps for obtaining the single-cycle hysteresis loop strain energy data are as follows: S201: Collect and analyze the transient force parameters and transient deformation parameters during the lifting process corresponding to the lifting space deflection characteristic data, extract the force timestamp associated with the transient force parameters, extract the deformation timestamp associated with the transient deformation parameters, perform time axis alignment processing, filter the transient force parameters and transient deformation parameters within the synchronous time node, and establish synchronous mechanical deformation data pairs; S202: Extract the stress and strain components contained in the synchronous mechanical deformation data pair, map the stress and strain components to a two-dimensional rectangular coordinate system, establish a set of plane coordinate points composed of a continuous time series, perform connection and closure processing on the set of plane coordinate points, and establish a closed geometric boundary jointly enclosed by the lifting loading cycle and the unloading cycle. S203: Extract the two-dimensional contour features defined by the closed geometric boundary, perform planar polygon numerical integration on the region inside the closed geometric boundary, calculate the planar area parameter covered by the region inside the closed geometric boundary, analyze the mechanical power dissipation associated with the planar area parameter, perform volume equivalent transformation calculation in combination with the stress section characteristics and deformation characteristics of the lifting structure, and establish single-cycle hysteresis loop strain energy data.

5. The method for safety assessment of roller lifting based on load data according to claim 4, characterized in that, The process of analyzing the mechanical power dissipation associated with the plane area parameter is as follows: Access the historical test database of the lifting structure and extract the material specifications and material damping loss factor of the lifting structure. Based on the material specifications, index matching is performed to extract the corresponding energy equivalence ratio coefficient. Collect the geometric parameters of the stress section and the effective deformation step size data of the lifting structure under lifting loading state; A preliminary dissipated energy density value is established by multiplying the plane area parameter and the energy equivalence ratio coefficient. A scalar product operation with different unit weights is then performed on the preliminary dissipated energy density value, the geometric parameter of the stressed section, and the effective deformation step size data to generate the physical energy level. A weighted ratio compensation calculation is performed on the physical energy level and the material damping loss factor to establish the mechanical power dissipation.

6. The method for safety assessment of roller lifting based on load data according to claim 4, characterized in that, The specific steps for obtaining the structural health degradation evaluation data are as follows: S301: Extract the timestamp sequence associated with the single hysteresis loop strain energy data, perform extreme point fluctuation scanning calculation on the timestamp sequence, extract the time span between adjacent extreme points in the same direction, define the alternating cycle period of the lifting load dynamic fluctuation over time according to the time span, and establish the load cycle time domain interval. S302: Based on the load cycle time domain interval, range matching is performed to extract multiple single hysteresis loop strain energy data within the time period covered by the alternating cycle cycle. For multiple single hysteresis loop strain energy data, continuous accumulation and summation numerical calculation is performed along the time series axis to aggregate the energy loss share caused by each load fluctuation and establish the cumulative damage energy parameter. S303: Call the accumulated damage energy parameter, read the critical fracture absorption benchmark in the factory material record of the metal component of the lifting structure, calculate the corresponding ratio of the accumulated damage energy parameter and the critical fracture absorption benchmark, and establish structural health degradation evaluation data.

7. The method for safety assessment of roller lifting based on load data according to claim 6, characterized in that, The specific steps for obtaining the collapse-prevention overturning envelope data are as follows: S401: Obtain the oscillation cycle statistical parameters within the time interval of the lifting operation environment, and use the structural health degradation evaluation data as an amplification factor to extract the single fatigue offset coefficient from the material parameter record table of the lifting equipment. Perform scalar multiplication numerical operation on the oscillation cycle statistical parameters and the single fatigue offset coefficient to calculate the mapping displacement of the fatigue accumulation effect on the geometric scale and generate the centripetal offset distance. S402: Obtain the initial anti-overturning shell model generated by the spatial mapping of the standard drawings of the lifting structure, perform three-dimensional geometric contour point analysis and extraction operation on the initial anti-overturning shell model, obtain the coordinates of multiple edge poles on the outermost contour boundary of the shell model, extract the spatial center coordinates corresponding to the three-dimensional spatial volume geometric center point of the shell model, and establish three-dimensional pole space data that aggregates external features and internal features. S403: Call the three-dimensional pole space data, read the edge pole coordinates and the space center coordinates, call the centripetal offset distance, perform linear algebraic coordinate translation operation on the edge pole coordinates along the spatial direction pointing to the space center coordinates, extract the reconstructed pole set generated after translation and shrinkage centripetal offset distance, and establish the collapse anti-overturning envelope surface data based on the surface enclosed by the spatial connection of all vertices in the reconstructed pole set.

8. The method for safety assessment of roller lifting based on load data according to claim 7, characterized in that, The process of obtaining the initial anti-overturning shell model generated by the spatial mapping of the standard drawings of the lifting structure is as follows: Access the lifting equipment design database and retrieve the standard drawings and physical dimension configuration table of the lifting structure; A two-dimensional vector analysis operation is performed on the standard drawings of the lifting structure to extract the two-dimensional projection line segment data in the standard drawings and establish a set of the outer plane contour of the lifting structure. Read the physical dimension configuration table and extract the tensile thickness parameter along the orthogonal spatial axis; A three-dimensional basic solid geometric data is established by performing a stretching transformation operation based on the stretching thickness parameter along the orthogonal spatial axis on the lifting outer plane contour set. The surface mesh is processed on the geometric data of the three-dimensional basic entity to extract polygonal facet elements exposed in the external space region; The polygonal facet elements are subjected to coplanar merging and convex hull mesh assembly calculations to generate a three-dimensional topological surface, which is then identified as the initial anti-overturning shell model.

9. The method for safety assessment of roller lifting based on load data according to claim 7, characterized in that, The specific steps for obtaining the lifting safety early warning record are as follows: S501: Obtain the three-dimensional coordinate information of the lifting equipment at the current time node, extract the real-time centroid spatial coordinates, call the collapse and overturning envelope data, determine the spatial inclusion relationship between the real-time centroid spatial coordinates and the collapse and overturning envelope data, and establish a free state based on the relative position relationship of the real-time centroid spatial coordinates outside the spatial boundary. S502: Read the limit tilt reference and warning reference in the preset configuration table, perform numerical comparison calculation on the lifting space deflection feature data and the limit tilt reference, determine the abnormal deflection state that exceeds the threshold range, perform the boundary value judgment operation on the structural health degradation evaluation data and the warning reference, generate the damage state, combine the free state, abnormal deflection state and damage state to establish multi-dimensional abnormal judgment data. S503: Based on the multidimensional anomaly determination data, extract the free state, abnormal deflection state and damage state to construct a multidimensional evaluation matrix, read the preset risk weight configuration data, and perform weighted aggregation operation on the free state, abnormal deflection state and damage state in the multidimensional evaluation matrix to extract the joint risk attribute of multiple abnormal features, match the corresponding risk code and timestamp information according to the joint risk attribute, and generate a lifting safety early warning record.

10. A road roller lifting safety assessment system based on load data, characterized in that, The system is used to implement the roller lifting safety assessment method based on load data as described in any one of claims 1-9, and the system includes: The deflection feature extraction module constructs a local tensor matrix and a spatial rotation matrix based on the initial tension parameter and the installation tilt parameter, respectively. It calculates the product of the local tensor matrix and the spatial rotation matrix to generate a global tensor matrix, and extracts the principal eigenvector of the global tensor matrix and the deviation angle of the plumb line to establish hoisting spatial deflection feature data. The hysteresis energy calculation module filters the lifting transient force parameters and transient deformation parameters of the synchronous time node of the lifting space deflection characteristic data, calculates the plane area parameter covered inside the closed geometric boundary of the paired plane coordinate point set, and establishes single hysteresis loop strain energy data. The degradation cumulative analysis module calculates the cumulative summation characteristics of the alternating cycle period corresponding to the single hysteresis loop strain energy data to establish the cumulative damage energy parameter, calculates the corresponding ratio of the cumulative damage energy parameter to the critical fracture absorption benchmark, and establishes structural health degradation evaluation data. The envelope surface collapse reconstruction module calculates the centripetal offset distance of the product of the structural health degradation evaluation data and the oscillation cycle statistical parameters and the single fatigue offset coefficient, calculates the set of reconstruction poles by moving the initial anti-overturning shell model edge pole coordinates to the spatial center coordinates by centripetal offset distance, and establishes the collapse anti-overturning envelope surface data. The joint risk early warning module analyzes the free state of the real-time centroid spatial coordinates deviating from the collapse and overturning envelope data, the abnormal deflection state of the lifting space deflection characteristic data exceeding the limit tilt benchmark, and the damage state of the structural health degradation evaluation data exceeding the boundary warning benchmark. It calculates the joint risk attributes and generates lifting safety early warning records.

Images