An outdoor cooker compression resistance detection method based on intelligent sensors

Abstract

The application discloses an outdoor cooker pressure resistance detection method based on an intelligent sensor, and belongs to the technical field of outdoor cooker pressure resistance detection. The method comprises the following steps: arranging a multi-point array containing at least three circumferentially uniformly distributed piezoresistive strain sensor nodes on the pressure-bearing surface of an outdoor cooker to be detected; applying a standard incremental load sequence and recording the initial resistance value change curve of each node; identifying the local peak value of the curve to extract a voltage fluctuation characteristic parameter set to generate a piezoresistive response characteristic vector; inputting the characteristic vector into a pre-configured rigid body mechanics mapping model to calculate the pressure-bearing deformation amount, generate a deformation displacement distribution, and screen abnormal deformation regions with deformation displacement exceeding a first threshold value as pressure resistance strength weakening regions and output a detection result. The method can invert the full-field deformation of the pressure-bearing surface from the discrete resistance response and accurately locate the weak strength parts.

Description

Technical Field

[0001] This invention relates to the field of outdoor cookware pressure resistance testing technology, specifically to a method for testing the pressure resistance of outdoor cookware based on intelligent sensors. Background Technology

[0002] Outdoor cookware is frequently subjected to external pressures such as stacking, collisions, and stove supports during transport and outdoor use. Its compressive strength directly affects its safety and structural lifespan. Current methods for testing the compressive strength of outdoor cookware mostly employ universal testing machines to perform compression failure tests on the overall structure, obtaining the load-displacement curve to determine the ultimate bearing capacity and macroscopic stiffness. However, this method only reflects the overall mechanical behavior of the cookware and cannot reveal localized deformation differences and areas of strength weakening on the pressure-bearing surface.

[0003] To further obtain deformation distribution information, some solutions involve attaching resistance strain gauges to the surface of the cookware and using a data acquisition system to monitor local strain during loading. However, strain gauges have stringent requirements for the attachment process, poor adaptability to curved surfaces, and can only obtain the average strain in one direction at the attachment point, making it difficult to form dense circumferential multi-point synchronous sensing. For outdoor cookware with irregular shapes or curved surfaces, this method has blind spots and is easily affected by temperature drift and wire interference, leading to missing or distorted local deformation information. At the signal processing level, the traditional method of directly mapping resistance changes to strain mixes minute material discontinuous responses with noise, lacking effective capture of early damage characteristics such as local instability and micro-yielding, and failing to extract characteristic parameters reflecting differences in pressure conditions from the details of resistance changes.

[0004] Because it is impossible to reconstruct the displacement distribution across the entire pressure-bearing surface from resistance measurements at finite discrete points, the identification of weak areas in outdoor cookware has long relied on manual observation based on experience or visual inspection after destructive testing. The lack of a method that can combine distributed piezoresistive sensing with mechanical inversion to visualize the pressure-bearing deformation field and automatically calibrate abnormally weakened areas has become a prominent problem affecting the accuracy and efficiency of pressure resistance testing for outdoor cookware. Summary of the Invention

[0005] This paper provides a method for detecting the compressive strength of outdoor cookware based on intelligent sensors, in order to solve the problems in the prior art of making it difficult to obtain the full-field deformation displacement distribution of the pressure-bearing surface and the inability to accurately locate the area of ​​weakened compressive strength.

[0006] To achieve the above objectives, the present invention provides the following technical solution: The present invention provides a method for testing the compressive strength of outdoor cookware based on intelligent sensors. This involves deploying a multi-point piezoresistive strain sensor array on the pressure-bearing surface of the outdoor cookware to be tested. This sensor array includes at least three piezoresistive strain sensor nodes uniformly distributed circumferentially. A standard incremental load sequence is applied to the piezoresistive strain sensor array, and the initial resistance value change curve of each sensor node under each load level is recorded. Local peak identification processing is performed on the initial resistance value change curves, and the voltage fluctuation characteristic parameter set corresponding to each sensor node is extracted to generate a piezoresistive response feature vector. This piezoresistive response feature vector is input into a pre-configured rigid body mechanics mapping model for compressive deformation calculation, generating the deformation displacement distribution of the outdoor cookware under preset load conditions. Based on the deformation displacement distribution, abnormal deformation regions with deformation displacement exceeding a first threshold are selected and identified as regions with weakened compressive strength. The compressive strength test result of the outdoor cookware is then output. This method can reconstruct the deformation distribution state of the pressure-bearing surface from the resistance change, achieving accurate positioning and detection of regions with weakened compressive strength.

[0007] In a preferred embodiment of the present invention, when applying a standard incremental load sequence, a preset stepped incremental load sequence is applied to the piezoresistive strain sensor array. The difference between adjacent load levels in this stepped incremental load sequence is a constant value. During the application of each load level, the real-time resistance values ​​of all sensor nodes are synchronously acquired at a preset high-frequency sampling frequency, generating a time-series record of the resistance value of each sensor node under the complete load sequence. The time-series record of the resistance value is then subjected to sliding window smoothing filtering to filter out high-frequency noise interference, resulting in a smoothed resistance value curve for each sensor node. Baseline drift correction processing is then performed on the smoothed resistance value curve to eliminate resistance value shifts caused by temperature drift, generating an initial resistance value change curve. Through stepped loading and signal preprocessing, the interference of environmental noise and temperature drift on the resistance measurement results can be effectively suppressed, improving the accuracy of the resistance change curve in characterizing the micro-deformation of the structure.

[0008] Preferably, when generating the piezoresistive response feature vector, a first-order difference operation is performed on the initial resistance value change curve to obtain the time-series distribution of the resistance change rate. Zero-crossing point detection is then performed on the time-series distribution of the resistance change rate to identify local maxima where the resistance change rate changes from positive to negative. The resistance value change amplitude, change duration, and peak width parameters corresponding to each local maxima are extracted to form a subset of local waveform feature parameters for a single sensor node. These subsets of local waveform feature parameters for all sensor nodes are then sequentially concatenated according to their sensor node numbers to generate the piezoresistive response feature vector. This method can extract transient features reflecting local yielding and stress release of the material from the time-varying resistance signal, enabling the feature vector to sensitively identify stress concentration points.

[0009] As a further preferred embodiment of the present invention, when the piezoresistive response feature vector is input into the rigid body mechanics mapping model, the material property parameter extraction layer of the rigid body mechanics mapping model infers the local elastic modulus and Poisson's ratio of the pressure-bearing area of ​​the outdoor cookware based on the resistance change values ​​of each sensor node in the piezoresistive response feature vector. The local elastic modulus and Poisson's ratio are then input into the stress-strain field reconstruction layer of the rigid body mechanics mapping model. This stress-strain field reconstruction layer performs gridded stress distribution calculation on the pressure-bearing area based on the finite difference method, generating a stress score for each grid node. The stress scores are then integrated to obtain the displacement of each grid node in the direction perpendicular to the pressure-bearing surface. The displacements of all grid nodes are arranged according to spatial coordinates to generate a deformation displacement distribution. In this way, multi-source discrete piezoresistive signals can be converted into a continuous deformation displacement field, obtaining a high spatial resolution deformation distribution without relying on full-field optical measurement equipment.

[0010] In a preferred embodiment of the present invention, when screening abnormal deformation regions, the deformation displacement distribution is processed by two-dimensional contour plotting to generate a displacement contour map of the pressure-bearing surface. Closed contour regions where the displacement amplitude continuously exceeds a first threshold are identified in the displacement contour map and marked as candidate abnormal deformation regions. The area of ​​each candidate abnormal deformation region is calculated, and candidate abnormal deformation regions with an area greater than a second threshold are selected as abnormal deformation regions. The sensor node numbers and spatial coordinate data corresponding to the abnormal deformation regions are extracted to generate location identification information for the weakened compressive strength region, and the compressive strength detection result is output. Through contour analysis and area constraints, false anomalies caused by isolated noise points can be eliminated, making the determination of the range and location of the weakened region more accurate.

[0011] As a further improvement to the above scheme, before applying the standard incremental load sequence, the outdoor cooker is fixed to a rigid support base, and a circular rigid pressure head is placed in the center area of ​​the pressure-bearing surface of the outdoor cooker. A servo motor drives the circular rigid pressure head to move vertically downwards, applying an initial preload to the contact point between the circular rigid pressure head and the pressure-bearing surface of the outdoor cooker. This stabilizes the initial resistance values ​​of all sensor nodes within a preset baseline resistance value range, and the baseline resistance value range is recorded as a zero-load reference for subsequent resistance value change curves. This unifies the loading boundary conditions and eliminates differences in initial contact states, ensuring the comparability of test results.

[0012] As a specific scheme for extracting material property parameters in this invention, when back-calculating the local elastic modulus and Poisson's ratio, the linear regression slope between the change in resistance value of each sensor node during the elastic deformation stage and the applied load is extracted from the piezoresistive response feature vector. The linear regression slope is multiplied by a preset piezoresistive sensitivity coefficient to obtain the local stress value at the location of each sensor node. The local stress value is divided by the corresponding local strain value to obtain the local elastic modulus at the location of each sensor node. Then, based on the ratio between the stress and strain values ​​of adjacent sensor nodes, Poisson's ratio is calculated through difference. This method can simultaneously acquire the distribution of mechanical parameters in the pressure-bearing region during a single loading process, providing a spatially varying material property field for subsequent stress-strain field reconstruction.

[0013] In a preferred embodiment of the present invention, after generating the deformation displacement distribution, a three-dimensional surface fitting process is performed on the deformation displacement distribution to generate a three-dimensional surface model of the deformation displacement of the pressure-bearing surface of the outdoor cookware. Curvature calculation is then performed on the three-dimensional surface model of the deformation displacement to extract curvature anomaly regions where the rate of curvature change exceeds a third threshold. Spatial overlap analysis is then performed between the curvature anomaly regions and the high displacement regions in the deformation displacement distribution to obtain a spatial overlap index. The value of the first threshold is adjusted based on the spatial overlap index, and the adjusted first threshold is used for screening abnormal deformation regions in the next detection process. By introducing curvature analysis and a feedback adjustment mechanism, the abnormal deformation criteria can be adaptively optimized, reducing the probability of false alarms and false negatives.

[0014] Preferably, the sliding window smoothing filtering process uses the Hanning window function, and the window length is adaptively adjusted according to the high-frequency sampling frequency, thereby maximizing the removal of random noise while preserving signal characteristics. In the local peak identification processing, the validity of the identified local maxima is also verified. False peaks caused by noise are eliminated by comparing whether the slope of the resistance value change within a preset time window before and after the local maxima exceeds a slope threshold, ensuring that the extracted waveform features reflect the true mechanical response process.

[0015] The technical effects and advantages provided by the present invention in the above technical solution are as follows:

[0016] This method extracts a set of voltage fluctuation characteristic parameters from the initial resistance value change curve and generates a piezoresistive response feature vector through local peak identification processing. Instead of directly using the resistance value amplitude as a feature, it obtains the time-series distribution of the resistance change rate through first-order difference operations, detects zero-crossing points to identify local maxima where the resistance change rate changes from positive to negative, and extracts the resistance value change amplitude, duration of change, and peak width parameters corresponding to each local maxima, forming a subset of local waveform characteristic parameters. These parameters can characterize the transient resistance response morphology generated by local yielding, micro-slippage, or minor instability of the material during loading, transforming discontinuous damage signs originally submerged in the smooth resistance curve into explicit waveform features. The piezoresistive response feature vector formed by splicing the feature parameter subsets of all sensor nodes provides a sharper and more comprehensive description of the differences in the mechanical state at various points on the pressure-bearing surface, making it possible to perceive abnormal deformation precursors from discrete resistance signals. This provides a highly discriminative input representation for subsequent accurate calculation of the deformation field, avoiding the problem of local weakening information being masked by the overall trend in conventional single-value resistance mapping methods.

[0017] The piezoresistive response feature vector is input into a rigid body mechanics mapping model to calculate the bearing deformation. This model uses a material property parameter extraction layer to deduce the local elastic modulus and Poisson's ratio of the bearing region from the resistance changes of each sensor node in the feature vector. Specifically, the local stress value is obtained by multiplying the linear regression slope between the resistance change of the sensor node during the elastic deformation stage and the applied load by the piezoresistive sensitivity coefficient, and then divided by the local strain value to obtain the local elastic modulus. The Poisson's ratio is calculated based on the difference between the stress and strain ratios of adjacent nodes. Subsequently, the stress-strain field reconstruction layer performs gridded stress distribution calculation on the bearing region based on the finite difference method, obtaining the stress score of each grid node. After integration, the vertical displacement is obtained and the deformation displacement distribution is generated. This model establishes a nonlinear mapping channel from the resistance response of the sparse sensor array to the full-field displacement distribution of the bearing surface. It does not rely on pre-calibrated global material constants, but instead uses the local piezoresistive response of each node to invert differentiated mechanical parameters in real time, making the displacement calculation for heterogeneous or locally damaged areas more realistic. Based on this, the abnormal deformation areas selected can directly correspond to the locations of strength weakening, realizing the automated location of weak points in the compressive strength of outdoor cookware, and making up for the shortcomings of traditional testing which can only provide overall load data. Attached Figure Description

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

[0019] Figure 1 This is a flowchart of a method for testing the pressure resistance of outdoor cookware based on smart sensors;

[0020] Figure 2 This is a flowchart of the resistance signal processing for a piezoresistive strain gauge array;

[0021] Figure 3 This is a graph showing the relationship between the initial resistance change curve of the piezoresistive strain gauge and the applied load. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, 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.

[0023] See Figure 1 This invention provides a method for testing the compressive strength of outdoor cookware based on intelligent sensors. The method involves deploying a multi-point piezoresistive strain sensor array on the pressure-bearing surface of the cookware, with each array containing at least three piezoresistive strain sensor nodes evenly distributed circumferentially. A standard incremental load sequence is applied to the piezoresistive strain sensor array, and the initial resistance value change curve of each sensor node under each load level is recorded. Local peak identification processing is performed on the initial resistance value change curves to extract the voltage fluctuation characteristic parameter set corresponding to each sensor node, generating a piezoresistive response feature vector. The piezoresistive response feature vector is input into a pre-configured rigid body mechanics mapping model for compressive deformation calculation, generating the deformation displacement distribution of the outdoor cookware under preset load conditions. Based on the deformation displacement distribution, abnormal deformation regions exceeding a first threshold are selected, and these abnormal deformation regions are identified as areas of weakened compressive strength. The compressive strength test result of the outdoor cookware is then output.

[0024] Example 1

[0025] In specific implementation, please refer to Figure 2A preset stepped increasing load sequence is applied to the piezoresistive strain sensor array, with a constant difference between adjacent load levels. This constant value is preset based on the design load limit and required detection resolution of the outdoor cookware under test. For example, when the outdoor cookware under test is a portable cookware made of aluminum alloy, the constant value is set to 10 Newtons. The servo motor connected to the circular rigid pressure head outputs torque to gradually increase the vertical pressure applied to the center area of ​​the pressure-bearing surface of the outdoor cookware according to the stepped increasing load sequence. Each load level is maintained for a preset stable duration, which is set to 3 seconds, to ensure that the resistance response of each piezoresistive strain sensor node reaches a steady state.

[0026] During each load stage, real-time resistance values ​​of all sensor nodes are synchronously acquired at a preset high-frequency sampling rate. The high-frequency sampling rate is set to 1000 Hz and implemented through a multi-channel synchronous analog-to-digital converter (ADC) acquisition module. The signal lines of all piezoresistive strain gauge sensor nodes are connected to the same analog input channel of the acquisition module, triggered by the same sampling clock source, ensuring synchronized sampling times for each channel. The acquired real-time resistance values ​​are arranged in chronological order of sampling time, generating a time-series record of the resistance values ​​for each sensor node under a complete step-increasing load sequence. This resistance value time-series record is a discrete-time series, denoted as […]. ,in Number the sensor nodes with indexes. This is the sampling time point number.

[0027] A sliding window smoothing filter is applied to the time-series resistance values ​​to remove high-frequency noise interference, resulting in a smoothed resistance curve for each sensor node. The sliding window smoothing filter uses a Hanning window function as the weighting coefficients for each sensor node. Its sampling time point sequence number Smoothing resistance value at Calculated using the following formula:

[0028]

[0029] in, Indicating the sensor node in the resistance value timing record Sampling time point sequence number The original resistance value at that location; This is the offset of the relative sampling time point number, and its value is... arrive Integers; The half-length of the window represents the number of sampling points contained on one side of the filtering window. The system adaptively adjusts based on a preset high-frequency sampling frequency, and the adjustment method is as follows: ,in For high-frequency sampling, The half-length factor of the time window, Set to 0.01 seconds. This indicates a floor operation. When the sampling frequency is high... At 1000 Hz, The value is 10, and the total length of the filtering window is 21 sampling points. Through the above sliding window smoothing filtering process, the smoothed resistance curve of each sensor node is obtained. .

[0030] Subsequently, baseline drift correction was performed on the smoothed resistance curve to eliminate resistance value shifts caused by temperature drift and generate an initial resistance value change curve. The baseline drift correction process is as follows: before applying a stepped increasing load sequence, the resistance values ​​of each sensor node under no-load conditions are recorded, and the baseline drift curve of each sensor node is obtained through least-squares linear fitting. The fitting order is first. The smoothed resistance curve is then... Subtract the baseline drift curve of the corresponding sensor node The curve of the change in the initial resistance value was obtained. The initial resistance value change curve represents the net change in resistance value caused solely by the applied load.

[0031] See Figure 3 The figure illustrates the relationship between the initial resistance change curves of each sensor node and the time-series response of the applied load when a standard stepped increasing load sequence is applied to the piezoresistive strain sensor array in Example 1. The horizontal axis represents time in seconds, the left side of the vertical axis represents the initial resistance change ΔR in ohms, and the right side of the vertical axis represents the applied load in Newtons. The solid black line in the figure represents the stepped change curve of the applied load, starting from 0 Newtons and increasing by 10 Newtons every approximately 3 seconds until reaching 60 Newtons, which conforms to the constant load increment and the preset stable duration of 3 seconds for each load increment described in Example 1.

[0032] The three curves, each in a different color, represent the initial resistance changes ΔR at sensor nodes 1, 2, and 3, respectively. The resistance change curves at each sensor node exhibit a clear stepped upward trend, highly corresponding to the applied stepped load sequence, indicating stable and synchronized sensor responses. During each load plateau period, the resistance change tends to stabilize, demonstrating that the sliding window smoothing filter and baseline drift correction effectively eliminate high-frequency noise and temperature drift interference.

[0033] Specifically, the resistance change at sensor node 3 is the largest, reaching approximately 3.6 ohms, followed by node 1 at approximately 3.0 ohms, while node 2 shows the smallest change at approximately 2.4 ohms, reflecting the differences in stress distribution at each node's location. As the loading steps increase progressively, the resistance changes at all three nodes continue to increase without significant saturation or hysteresis, indicating that the material and sensor are operating in the elastic deformation stage.

[0034] Example 2

[0035] In practice, the initial resistance value change curve is processed using a first-order difference operation to obtain the time-series distribution of the resistance change rate. For each sensor node, the initial resistance value change curve is represented as a discrete sequence. ,in Used as a sensor node identifier. The sampling time point number. The value range is from 1 to the total number of sampling points. The integer. First-order difference operations are performed as follows: calculate the difference in resistance changes between two adjacent sampling points to obtain the time-series distribution of the resistance change rate. ,in The value range is from 1 to Integer values. Time series distribution of resistance change rate. It reflects the instantaneous rate of change of the initial resistance value at each sampling point.

[0036] Zero-crossing point detection is performed on the time-series distribution of the resistance change rate to identify local maxima where the resistance change rate changes from positive to negative. The zero-crossing point detection process involves traversing the time-series distribution of the resistance change rate. All time points are achieved. When detected... and Then, it is determined that at time point... At point [time], the sign of the rate of change of resistance changed from positive to negative. Points on the corresponding initial resistance value change curve Marked as a candidate local maximum point. If the condition is met continuously at multiple consecutive time points... In the case of zero value interval, the center point of this interval is determined as the sign change point of the resistance change rate, and the corresponding initial resistance value change curve point is marked as a candidate local maximum point.

[0037] The identified candidate local maxima are validated to remove spurious peaks caused by noise. Validity validation is achieved by comparing whether the slope of the resistance change within a preset time window before and after the candidate local maximum exceeds a slope threshold. The length of the preset time window is set to... Each sampling time point The value is an odd number determined based on the high-frequency sampling frequency and mechanical response characteristics. For example, The value is 7. For each candidate local maximum point, the corresponding time point... In its forward preset time window Internally, the slope of the forward resistance change is calculated through linear fitting. ; in its subsequent preset time window Internally, the slope of the backward resistance value change is calculated through linear fitting. The validity criterion formula is:

[0038]

[0039] in, The slope of the forward resistance value change represents the linear fitting slope of the initial resistance value change curve as a function of the sampling time point within a preset time window before the candidate local maximum point. The slope of the backward resistance value change represents the linear fitting slope of the curve of the initial resistance value change as a function of the sampling time point within a preset time window after the candidate local maximum point. This is the slope threshold. Slope threshold The acquisition method is as follows: Under a static state without any applied load, the basic resistance value signals of all sensor nodes are collected for a continuous period of time at the same high-frequency sampling frequency. The slope calculation process of the basic resistance value signals is performed in the same way as the validity verification to obtain the statistical distribution of the sum of the absolute values ​​of the slopes at all time points. The 95th percentile of the statistical distribution is taken as the slope threshold. This ensures that resistance fluctuations caused solely by noise cannot pass the validity check. If a candidate local maximum point satisfies the above discrimination formula, it is retained as a valid local maximum point; otherwise, it is discarded as a spurious peak.

[0040] For each valid local maximum point, the resistance change amplitude, duration of change, and peak width parameters are extracted to form a subset of local waveform characteristic parameters for a single sensor node. The resistance change amplitude is defined as the difference between the resistance value at the local maximum point and the resistance value at the adjacent previous valid local minimum point. If the local maximum point is the first valid local maximum point, it is compared with the nearest resistance baseline level, which is taken from the average resistance value during the zero-load reference stage before the application of the standard incremental load sequence. The duration of change is defined as the time length obtained by dividing the number of sampling time points from the start of the resistance value increase to the moment of the local maximum point by the high-frequency sampling frequency. The peak width parameter is defined as the time width between the two intersection points of the resistance change curve at half the height of the resistance change amplitude at the local maximum point; the time difference is calculated after obtaining the intersection point time through linear interpolation. For each sensor node, the above three characteristic parameters corresponding to all its valid local maximum points are combined according to the order in which the valid local maximum points appear to form a subset of local waveform characteristic parameters.

[0041] The local waveform feature parameters subsets of all sensor nodes are concatenated sequentially according to their sensor node numbers to generate a piezoresistive response feature vector. The sensor node numbers start from 1 and increment to the total number of sensor nodes. , It must be an integer of at least 3. For sensor node numbers... Its local waveform characteristic parameter subset is denoted as Piezoresistive response eigenvector Obtained by sequentially concatenating: If a sensor node does not detect any valid local maxima during the application of a standard incremental load sequence, its local waveform feature parameter subset is filled with zero vectors, and its length is consistent with the average length of the local waveform feature parameter subsets of other sensor nodes.

[0042] Example 3

[0043] In practice, the piezoresistive response feature vector is input into the rigid body mechanics mapping model. The rigid body mechanics mapping model consists of a material property parameter extraction layer and a stress-strain field reconstruction layer connected in series. The piezoresistive response feature vector first enters the material property parameter extraction layer.

[0044] In the material property parameter extraction layer, the linear regression slope between the resistance change of each sensor node during the elastic deformation stage and the applied load is extracted from the piezoresistive response feature vector. For each sensor node number... Sensor nodes are extracted from the piezoresistive response feature vector. A subset of local waveform characteristic parameters is used. Based on the resistance value change amplitude corresponding to each effective local maximum point recorded in the subset of local waveform characteristic parameters, and corresponding to the values ​​of each load level when applying a standard increasing load sequence, load level data points within the elastic deformation range of outdoor cookware materials are selected. The least squares method is used to fit the linear relationship between the resistance value change and the load to obtain the linear regression slope. Linear regression slope Represents sensor nodes The change in resistance value under a unit load increment.

[0045] The slope of the linear regression Multiply by the preset piezoresistive sensitivity coefficient Obtain sensor nodes Local stress value at the location Piezoresistive sensitivity coefficient It is the conversion factor of the sensor resistance change to the applied stress. It is obtained by performing a uniaxial stress calibration experiment on a standard specimen made of the same material as outdoor cooking utensils. A series of known uniaxial stresses are applied to the standard specimen using a universal testing machine, and the resistance changes at the sensor nodes are recorded simultaneously. The resistance change is then linearly fitted to the applied uniaxial stress, and the slope of the fitted line is the piezoresistive sensitivity coefficient. Piezoresistive sensitivity coefficient The value is located in Ohm per Pascal to Between ohms per pascal.

[0046] Local strain value The radial strain theoretical or finite element simulation values ​​at each sensor node location on the pressure surface under various levels of load in a standard incremental load sequence are pre-stored in the material property parameter extraction layer, based on the geometry of the outdoor cookware, the shape of the pressure-bearing surface, and the position of the circular rigid indenter. This is obtained by extracting the linear regression slope. Under synchronous load levels, the corresponding theoretical radial strain value is read as the local strain value. .

[0047] Local stress value Divide by the corresponding local strain value Obtain sensor nodes Local elastic modulus at the location .

[0048] Poisson's ratio is calculated using the difference method based on the ratio between the stress and strain values ​​of adjacent sensor nodes. In a circumferentially uniformly distributed array of sensor nodes, two sensor nodes with a 90-degree angle between them in the circumferential direction are selected, for example, sensor nodes... With sensor nodes Extract sensor nodes radial strain and sensor nodes Circumferential strain Poisson's ratio Calculated by difference using the following formula: If there are multiple pairs of sensor nodes that satisfy the vertical relationship, the average of the Poisson's ratios obtained from each pair is calculated as the Poisson's ratio of the entire outdoor cookware pressure-bearing area.

[0049] In the stress-strain field reconstruction layer, the local elastic modulus and Poisson's ratio are first mapped to the pressure-bearing region mesh. The pressure-bearing surface region of the outdoor cookware is discretized as follows: A rectangular grid, with grid node coordinates as ,in For row index, , For column indexes, The radial basis function interpolation method is used to obtain the local elastic modulus at all sensor node locations. Given points, interpolate to calculate the local elastic modulus at each grid node. Poisson's ratio A uniform value is taken throughout the pressure-bearing area.

[0050] The stress-strain field reconstruction layer calculates the gridded stress distribution in the pressure-bearing region using the finite difference method. The pressure-bearing region of the outdoor cookware is equivalent to an elastic thin plate bearing a concentrated load at the center. Based on classical thin plate bending theory, the biharmonic operator form of its deflection control equation is discretized using a finite difference operator. The local elastic modulus of each grid node is substituted into the discretized equation system. Poisson's ratio The equivalent nodal load applied by the circular rigid indenter, combined with fixed or simply supported boundary conditions, is used to solve for the bending moment components at each mesh node using an over-relaxation iterative method. Based on the bending moment components, the bending normal stress perpendicular to the bearing surface is calculated, and this bending normal stress is defined as the stress component. .

[0051] The stress values ​​are integrated to obtain the displacement of each grid node in the direction perpendicular to the bearing surface. The integration process employs a double numerical integration method, progressing from the edge of the bearing region towards the desired grid node. The calculation formula is as follows:

[0052]

[0053] in, Grid node coordinates The displacement perpendicular to the bearing surface, expressed in meters; Grid node coordinates The bending stiffness factor at the location, , The nominal wall thickness of the pressure-bearing area of ​​outdoor cooking utensils, in meters; This is a cumulative index along the row direction, with values ​​ranging from 1 to... Integers; To add an index in the column direction, the value range is from 1 to... Integers; Grid node coordinates The stress component at the point, in Pascals; For grid in The directional spacing dimension, in meters; For grid in The directional spacing dimension, in meters.

[0054] Displacement of all grid nodes According to the spatial coordinates of the grid nodes Arrange them to form a collection OK The matrix of columns represents the deformation displacement distribution. This distribution is then output to subsequent processing stages to filter for abnormal deformation regions.

[0055] Example 4

[0056] In practice, the deformation displacement distribution is processed using two-dimensional contour plotting to generate a displacement contour map of the bearing surface. The deformation displacement distribution is a map containing... OK A matrix of columns, and the elements in the matrix. Represents grid node coordinates The displacement perpendicular to the bearing surface, where For row index, , For column indexes, Two-dimensional contour line drawing is performed in the following way: Within a rectangular planar region composed of grid nodes, a set of equally spaced displacement reference values ​​are defined as contour levels. The spacing of the displacement reference values ​​is determined by dividing the difference between the maximum and minimum displacement in the deformation displacement distribution by a predetermined number of levels, which is set to 20. A linear interpolation method is used to track the position coordinates of each contour level crossing the grid cell boundary. The position coordinates of the same displacement reference value are connected in adjacent order to form closed or open curve segments. The curve segments corresponding to all contour levels together constitute a displacement contour map.

[0057] Closed contour regions where the displacement amplitude continuously exceeds a first threshold are identified in the displacement contour map and marked as candidate anomalous deformation regions. First threshold The allowable displacement is determined based on the allowable stress corresponding to the allowable displacement of the outdoor cookware material. The method for setting this is: based on the nominal wall thickness of the pressure-bearing area of ​​the outdoor cookware. Multiply by the allowable strain value of the material Divide by the safety factor To obtain, that is Allowable strain value of material Take one-third of the material's yield strain as the safety factor. The threshold is set to 1.5. The identification process is as follows: traverse all closed contour lines in the displacement contour map. For each closed contour line, check the displacement of all grid nodes within its enclosed area. If the reference value of the displacement corresponding to the closed contour line is greater than the first threshold... Furthermore, there is no displacement within the closed contour line that is lower than the first threshold. If a concave region is identified, the area enclosed by the closed contour line is determined to be a closed contour line region where the displacement amplitude continuously exceeds the first threshold, and it is marked as a candidate abnormal deformation region. If multiple nested closed contour lines satisfy the conditions, the area enclosed by the outermost closed contour line is taken as the candidate abnormal deformation region.

[0058] For each candidate abnormal deformation region, area calculation is performed, and regions with areas greater than a second threshold are selected as abnormal deformation regions. The area calculation uses a polygon area formula to calculate the boundary coordinate sequence of the candidate abnormal deformation regions. For the... There are 1 candidate abnormal deformation regions, and their boundary coordinate sequence is as follows: ,in For the first The total number of boundary vertices of each candidate anomalous deformation region For the first The first candidate abnormal deformation region Each boundary vertex is in The coordinates of the direction. For the first The first candidate abnormal deformation region Each boundary vertex is in The coordinates of the direction. The boundary vertex index, Area of ​​candidate anomalous deformation regions The calculation formula is:

[0059]

[0060] in, Defined as , Defined as , This indicates the absolute value operation. Second threshold. The area is set at 5% of the total pressure-bearing surface area of ​​the outdoor cookware, which is directly obtained from the geometric design parameters of the pressure-bearing surface. The area of ​​each candidate abnormal deformation region is compared one by one. With the second threshold , will satisfy Candidate abnormal deformation regions are retained as abnormal deformation regions; candidate abnormal deformation regions that do not meet the area condition are removed and not included in subsequent processing.

[0061] The sensor node numbers and spatial coordinate data corresponding to the abnormal deformation areas are extracted to generate location identification information for the weakened compressive strength areas. The extraction process is as follows: the boundary coordinates of the abnormal deformation areas are mapped to the layout coordinate system of the piezoresistive strain gauge array. All sensor nodes located within the boundary range of the abnormal deformation area or within a preset distance range of the boundary's neighborhood are retrieved. The preset distance range is half the circumferential spacing between adjacent sensor nodes. The sensor node numbers are obtained to form a sensor node number list; simultaneously, the spatial coordinate data of the retrieved sensor nodes is obtained. The spatial coordinate data represents the two-dimensional installation position coordinates of the sensor nodes on the pressure-bearing surface of the outdoor cookware. The sensor node number list and spatial coordinate data are combined to generate location identification information for the abnormal deformation area. The location identification information for all abnormal deformation areas is summarized as the location result of the weakened compressive strength areas, and output as a component of the compressive strength test result. The compressive strength test result also includes the maximum displacement value, area value, and corresponding load level information for each abnormal deformation area.

[0062] Example 5

[0063] In practice, the following steps are performed before applying a standard incremental load sequence to the piezoresistive strain gauge array. The outdoor cookware is fixed to a rigid support base. The rigid support base is made of cast iron, with a positioning groove on its upper surface that matches the bottom contour of the outdoor cookware. The bottom of the outdoor cookware is embedded in the positioning groove. Four manually operated locking clamps distributed circumferentially around the rigid support base hold the edge flange of the outdoor cookware, keeping it stationary relative to the rigid support base during the test. A circular rigid indenter is placed in the center of the pressure-bearing surface of the outdoor cookware. The circular rigid indenter is made of hardened tool steel, with a spherical or planar lower end face. The radius of the sphere or the diameter of the planar face is selected according to the shape of the pressure-bearing surface and the standard test specifications. The area of ​​the lower end face of the circular rigid indenter covers the center area of ​​the pressure-bearing surface, and the axis of the circular rigid indenter coincides with the normal direction of the pressure-bearing surface.

[0064] A servo motor drives a circular rigid pressure head to move vertically downwards. The output shaft of the servo motor is connected to the pressure head mounting base via a ball screw mechanism, and the circular rigid pressure head is fixed to the lower end of the mounting base. The servo motor receives commands from the motion controller and feeds downwards at a low speed of 0.1 mm / s. Just as the lower end face of the circular rigid pressure head is about to contact the pressure-bearing surface, the real-time resistance values ​​of all sensor nodes are monitored. When a identifiable change in the resistance value of any sensor node occurs, it is determined that the circular rigid pressure head has made contact with the pressure-bearing surface. The servo motor continues to drive, applying an initial preload to the contact point between the circular rigid pressure head and the pressure-bearing surface of the outdoor cookware. The initial preload is controlled as follows: the target resistance change is set to the force value corresponding to two percent of the full scale. Closed-loop control is achieved through force sensor feedback or current estimation. When the resistance changes of at least two-thirds of all sensor nodes fall within the preset range, the initial preload is considered to be applied correctly. The preset baseline resistance range is defined as the range within ±0.5 percent of the average resistance value of all sensor nodes during a 30-second stable period. Record the baseline resistance value range and use it as the zero-load reference for subsequent resistance value change curves. That is, when calculating the subsequent initial resistance value change curve, subtract the corresponding average baseline resistance value from the real-time resistance value to obtain the net change.

[0065] After generating the deformation displacement distribution of the outdoor cookware under preset load conditions, a three-dimensional surface fitting process is performed on the deformation displacement distribution to generate a three-dimensional surface model of the deformation displacement of the pressure-bearing surface of the outdoor cookware. The three-dimensional surface fitting process uses a bicubic spline interpolation method. The mesh node coordinates in the deformation displacement distribution are shown below. and displacement As known control points, among which For row index, , For column indexes, In the direction of travel and column direction As the independent variable, displacement As the dependent variable, respectively in direction and The direction is used to construct a cubic spline function, and the coefficients of the spline function are determined by solving a system of linear equations, thus forming a continuous spatial surface expression. The three-dimensional surface model of deformation displacement is derived from the spatial surface expression. It consists of all the values ​​taken in the rectangular area formed by the grid nodes.

[0066] Curvature calculations are performed on a three-dimensional surface model subjected to deformation and displacement to extract curvature anomaly regions where the rate of curvature change exceeds a third threshold. The curvature calculation process includes calculating the coefficients of the first and second fundamental forms of the surface, and then solving for the Gaussian curvature at each point on the surface. and mean curvature Based on the Gaussian curvature value at each point on the surface, the rate of change of curvature is defined as the gradient magnitude of the Gaussian curvature, i.e. The Gaussian curvature gradient magnitude is calculated at each point on the deformed displacement three-dimensional surface model to obtain the curvature change rate distribution. Third threshold. The method for obtaining this value is as follows: calculate the mean value of the Gaussian curvature gradient magnitude at all points on the entire bearing surface. and standard deviation Set the third threshold to Continuous regions where the magnitude of the Gaussian curvature gradient exceeds the third threshold are marked as curvature anomaly regions.

[0067] Spatial overlap analysis is performed between the curvature anomaly region and the high displacement region in the deformation displacement distribution to obtain a spatial overlap index. The high displacement region is defined as the area in the deformation displacement distribution where the displacement exceeds 80% of the maximum displacement. The spatial overlap index is defined as the ratio of the intersection area of ​​the curvature anomaly region and the high displacement region to the area of ​​the curvature anomaly region. The spatial overlap index is calculated as follows: ,in This represents the area of ​​the intersection region between the curvature anomaly region and the high displacement region. This represents the area of ​​the curvature anomaly region. The value of the first threshold is adjusted based on the spatial overlap index. The adjustment method is as follows: if the spatial overlap index... If the value is greater than 0.7, the first threshold is multiplied by a reduction factor of 0.8 to improve the sensitivity of abnormal deformation region screening in the next detection process; if the spatial overlap index If the value is less than 0.3, the first threshold is multiplied by an amplification factor of 1.2 to reduce the probability of false screening; if the spatial overlap index If the value is between 0.3 and 0.7, the first threshold remains unchanged. The adjusted first threshold is then used for screening abnormal deformation regions in the next detection process, thus completing the adaptive update of the first threshold.

[0068] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A method for detecting the pressure resistance of outdoor cookware based on intelligent sensors, characterized in that, The method includes: A multi-point piezoresistive strain sensor array is deployed on the pressure-bearing surface area of ​​the outdoor cookware to be tested. The sensor array includes at least three piezoresistive strain sensor nodes that are uniformly distributed along the circumference. A standard incremental load sequence was applied to the piezoresistive strain sensor array, and the initial resistance value change curve of each sensor node under each load level was recorded. The initial resistance value change curve is processed for local peak identification, and the voltage fluctuation feature parameter set corresponding to each sensor node is extracted to generate a piezoresistive response feature vector. The piezoresistive response feature vector is input into a pre-configured rigid body mechanics mapping model for compressive deformation calculation, thereby generating the deformation displacement distribution of the outdoor cookware under a preset load condition. Based on the deformation displacement distribution, abnormal deformation regions where the deformation displacement exceeds the first threshold are selected, and these abnormal deformation regions are locked as regions with weakened compressive strength. The compressive strength test results of the outdoor cookware are then output.

2. The method for detecting the pressure resistance of outdoor cooking utensils based on intelligent sensors according to claim 1, characterized in that, The process of applying a standard incremental load sequence to the piezoresistive strain sensor array and recording the initial resistance value change curve of each sensor node under each load level includes: A preset stepped increasing load sequence is applied to the piezoresistive strain sensor array, wherein the difference between adjacent load levels in the stepped increasing load sequence is a constant value; During each load level application process, the real-time resistance values ​​of all sensor nodes are synchronously collected at a preset high-frequency sampling frequency to generate a time-series record of the resistance values ​​of each sensor node under the complete load sequence. The time-series resistance value records are subjected to sliding window smoothing filtering to remove high-frequency noise interference, resulting in a smoothed resistance value curve for each sensor node. The smoothed resistance curve is subjected to baseline drift correction to eliminate the resistance value shift caused by temperature drift, thereby generating the initial resistance value change curve.

3. The method for detecting the pressure resistance of outdoor cookware based on intelligent sensors according to claim 1, characterized in that, The process of performing local peak identification processing on the initial resistance value change curve, extracting the voltage fluctuation feature parameter set corresponding to each sensor node, and generating a piezoresistive response feature vector includes: The initial resistance value change curve is processed by first-order difference operation to obtain the time distribution of the resistance change rate; Zero-crossing point detection processing is performed on the time-series distribution of the resistance change rate to identify the local maximum point where the resistance change rate changes from positive to negative. Extract the resistance value change amplitude, change duration, and peak width parameters corresponding to each local maximum point to form a subset of local waveform feature parameters for a single sensor node; The local waveform feature parameter subsets of all sensor nodes are sequentially concatenated according to the sensor node numbers to generate the piezoresistive response feature vector.

4. The method for detecting the pressure resistance of outdoor cooking utensils based on intelligent sensors according to claim 1, characterized in that, The step of inputting the piezoresistive response feature vector into a pre-configured rigid body mechanics mapping model for compressive deformation calculation to generate the deformation displacement distribution of the outdoor cookware under a preset load condition includes: The piezoresistive response feature vector is input into the material property parameter extraction layer of the rigid body mechanics mapping model. The material property parameter extraction layer infers the local elastic modulus and Poisson's ratio of the pressure-bearing area of ​​the outdoor cookware based on the resistance change values ​​of each sensor node in the piezoresistive response feature vector. The local elastic modulus and Poisson's ratio are input into the stress-strain field reconstruction layer of the rigid body mechanics mapping model. The stress-strain field reconstruction layer performs gridded stress distribution calculation on the pressure-bearing region based on the finite difference method, generating stress scores for each grid node. The stress score is integrated to obtain the displacement of each grid node in the direction perpendicular to the bearing surface. The displacements of all grid nodes are arranged according to spatial coordinates to generate the deformation displacement distribution.

5. The method for detecting the pressure resistance of outdoor cookware based on intelligent sensors according to claim 1, characterized in that, The process involves filtering out abnormal deformation regions whose deformation displacement exceeds a first threshold based on the deformation displacement distribution, locking these abnormal deformation regions as areas of weakened compressive strength, and outputting the compressive strength test results of the outdoor cookware, including: The deformation displacement distribution is plotted using two-dimensional contour lines to generate a displacement contour map of the pressure-bearing surface; In the displacement contour map, closed contour regions where the displacement amplitude continuously exceeds the first threshold are identified and marked as candidate abnormal deformation regions. The area of ​​each candidate abnormal deformation region is calculated, and the candidate abnormal deformation regions with an area greater than the second threshold are selected as the abnormal deformation regions. Extract the sensor node number and spatial coordinate data corresponding to the abnormal deformation area, generate the positioning identification information of the weakened compressive strength area, and output the compressive strength detection result.

6. The method for detecting the pressure resistance of outdoor cooking utensils based on intelligent sensors according to claim 1, characterized in that, Before applying the standard incremental load sequence to the piezoresistive strain sensor array, the method further includes: The outdoor cooker is fixed on a rigid support base, and a circular rigid pressure head is placed in the center area of ​​the pressure-bearing surface of the outdoor cooker. The circular rigid pressure head is driven vertically downward by a servo motor to apply an initial preload to the contact point between the circular rigid pressure head and the pressure-bearing surface of the outdoor cookware, so that the initial resistance value of all sensor nodes is stabilized within the preset baseline resistance value range. Record the baseline resistance value range and use it as a zero-load reference for subsequent resistance value change curves.

7. The method for detecting the pressure resistance of outdoor cooking utensils based on intelligent sensors according to claim 4, characterized in that, The step of inputting the piezoresistive response feature vector into the material property parameter extraction layer of the rigid body mechanical mapping model, wherein the material property parameter extraction layer deduces the local elastic modulus and Poisson's ratio of the pressure-bearing area of ​​the outdoor cookware based on the resistance change values ​​of each sensor node in the piezoresistive response feature vector, includes: Extract the linear regression slope between the resistance change of each sensor node during the elastic deformation stage and the applied load from the piezoresistive response feature vector; Multiplying the linear regression slope by a preset piezoresistive sensitivity coefficient yields the local stress value at the location of each sensor node. Divide the local stress value by the corresponding local strain value to obtain the local elastic modulus at the location of each sensor node; The Poisson's ratio is obtained by differential calculation based on the ratio between the stress and strain values ​​of adjacent sensor nodes.

8. The method for detecting the pressure resistance of outdoor cookware based on intelligent sensors according to claim 1, characterized in that, After generating the deformation displacement distribution of the outdoor cooking appliance under a preset load condition, the method further includes: The deformation displacement distribution is subjected to three-dimensional surface fitting to generate a three-dimensional surface model of the deformation displacement of the pressure-bearing surface of the outdoor cookware. The curvature calculation process is performed on the three-dimensional surface model of the deformation displacement to extract the curvature anomaly region where the rate of curvature change exceeds the third threshold; The spatial overlap of the curvature anomaly region and the high displacement region in the deformation displacement distribution is analyzed to obtain a spatial overlap index. The value of the first threshold is adjusted according to the spatial overlap index, and the adjusted first threshold is used for screening abnormal deformation areas in the next detection process.

9. A method for detecting the pressure resistance of outdoor cooking utensils based on intelligent sensors according to claim 2, characterized in that, The sliding window smoothing filtering process uses the Hanning window function, and the window length is adaptively adjusted according to the high-frequency sampling frequency.

10. A method for detecting the pressure resistance of outdoor cookware based on intelligent sensors according to claim 3, characterized in that, The local peak identification process also includes validating the identified local maxima and removing false peaks caused by noise. The validity check is achieved by comparing whether the slope of the resistance value change within a preset time window before and after the local maxima exceeds the slope threshold.

Images