Nonlinear ultrasonic based safety assessment system and method for pressure-containing components

Abstract

This invention discloses a safety assessment system and method for pressure-bearing components based on nonlinear ultrasound, belonging to the field of component safety assessment technology. The invention controls a robotic arm to drive an automatic probe to acquire nonlinear signals; establishes a mapping benchmark library and calibrates the background nonlinear threshold; divides the detection area of ​​the pressure-bearing component under test into a grid, and acquires the original field detection signal corresponding to each grid unit; removes nonlinear interference from each original field detection signal and calculates effective nonlinear parameters; based on the effective nonlinear parameters and the background nonlinear threshold, filters out abnormal candidate regions; for all abnormal candidate regions, inversely calculates the equivalent emission source parameters of the abnormal nonlinear signals; based on the three-dimensional spatial coordinates of the equivalent emission source parameters, determines the abnormal location region and outputs the location characteristics of the abnormal location region; trains a safety assessment model, inputs the location characteristics into the safety assessment model, and calculates the safety assessment result of the pressure-bearing component.

Description

Technical Field

[0001] This invention relates to the field of component safety assessment technology, specifically to a system and method for assessing the safety of pressure-bearing components based on nonlinear ultrasound. Background Technology

[0002] Pressure-bearing components of thermal power units, such as boiler headers, main steam pipelines, and pressure vessels, are subjected to complex operating conditions of high temperature, high pressure, alternating loads, and media corrosion for extended periods. They are highly susceptible to early defects such as fatigue cracks, intergranular corrosion, and microscopic damage. These defects can gradually expand over time, directly leading to safety accidents such as leaks and pipe ruptures. Therefore, conducting accurate and reliable safety assessments of pressure-bearing components is a core element in ensuring the safe and stable operation of thermal power units.

[0003] In existing testing processes, fluctuations in the clamping force between the probe and the component surface generate additional nonlinear interference signals due to clamping force coupling. These interference signals superimpose with the actual nonlinear damage signals, leading to distortion in the baseline nonlinear threshold calibration and a significant reduction in the detection signal-to-noise ratio. Current technologies can only mark abnormal locations through threshold comparison, failing to invert equivalent emission source parameters, making precise defect location and boundary delineation difficult, and the positioning accuracy cannot meet engineering requirements. After locating the abnormal area, current technologies can only achieve qualitative anomaly judgment, unable to output core results such as damage level, defect type, and remaining safe operating life; therefore, the assessment results lack engineering guidance value. Summary of the Invention

[0004] The purpose of this invention is to provide a safety assessment system and method for pressure-bearing components based on nonlinear ultrasound, in order to solve the problems raised in the prior art.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] In a first aspect, this application provides a safety assessment method for pressure-bearing components based on nonlinear ultrasound, comprising the following steps:

[0007] The robotic arm is controlled to drive an automatic probe to collect inherent nonlinear signals and additional nonlinear signals; a mapping benchmark library between clamping force and corresponding additional nonlinear signals is established, and the background nonlinear threshold is calibrated.

[0008] The testing area of ​​the pressure-bearing component under test is divided into grids, and the original field test signal corresponding to each grid cell is collected. Nonlinear interference is removed from each original field test signal, and the effective nonlinear parameter is calculated. Based on the effective nonlinear parameter, combined with the background nonlinear threshold, abnormal candidate areas are screened.

[0009] Specifically, for each original on-site detection signal, nonlinear interference is eliminated, and effective nonlinear parameters are calculated. Preferably, the inherent nonlinear signal of the system under zero clamping force, which was calibrated earlier, is retrieved. This signal is an inherent nonlinear component generated by the hardware and circuitry of the detection system itself and is unrelated to the pressure-bearing component under test or the clamping force of the probe. From the calculated total nonlinear parameters, the nonlinear component corresponding to this inherent nonlinear signal is completely deducted to eliminate the inherent interference brought by the detection system itself, resulting in a first-corrected nonlinear parameter. The real-time clamping force value synchronously locked during the signal acquisition of this grid cell is retrieved, and the mapping benchmark library of clamping force and additional nonlinear signal constructed earlier is accessed. Through an interpolation fitting matching algorithm, the additional nonlinear interference signal corresponding to this real-time clamping force value is obtained. From the first-corrected nonlinear parameter, the additional nonlinear interference signal obtained by the matching is completely deducted to eliminate the coupled nonlinear interference caused by the fluctuation of the probe clamping force, resulting in a second-corrected nonlinear parameter, which is used as the effective nonlinear parameter.

[0010] For all candidate anomaly regions, the equivalent emission source parameters of the anomalous nonlinear signal are inverted and calculated; based on the three-dimensional spatial coordinates of the equivalent emission source parameters, the anomaly location region is determined, and the location characteristics of the anomaly location region are output.

[0011] The safety assessment model is trained using artificial intelligence algorithms. The location features are input into the safety assessment model to calculate the safety assessment result of the pressure-bearing component.

[0012] In conjunction with the first aspect, in a first embodiment of the first aspect of this application, the control of the robotic arm to drive the automatic probe to collect inherent nonlinear signals and additional nonlinear signals includes:

[0013] Several sets of continuously gradient-distributed clamping forces are preset, and the range of the clamping forces covers the minimum effective coupling clamping force to the maximum safe clamping force required for on-site detection.

[0014] The robotic arm is controlled to move the automatic probe into a suspended, non-contact state, triggering the nonlinear ultrasonic detector to output the first fundamental excitation signal, acquiring the first full-wave signal, and calculating the inherent nonlinear signal under zero clamping force.

[0015] The robotic arm is controlled to clamp the actuator, the probe clamping stroke is adjusted, and the nonlinear ultrasonic detector is triggered to output the second fundamental wave excitation signal. The automatic probe is controlled to emit the ultrasonic fundamental wave into the interior of the undamaged standard test block, and the second full wave signal is collected. Coherent averaging is performed to obtain the original calibration signal under the current clamping force. The fundamental wave and each higher harmonic characteristic of the original calibration signal are extracted, and the total nonlinear parameter under the current clamping force is calculated. The inherent nonlinear signal is subtracted from the total nonlinear parameter to obtain the additional nonlinear signal corresponding to the current clamping force.

[0016] Preferably, a drive command is sent to the clamping actuator of the robotic arm to adjust the probe clamping stroke and control the automatic probe to gradually clamp the surface of the undamaged standard test block; the force feedback data of the clamping actuator is collected in real time, and the real-time clamping force value is compared with the preset target clamping force value in a closed loop, and the probe clamping stroke is dynamically fine-tuned until the real-time clamping force is stable within the allowable error range of the preset target clamping force, thus completing the stable coupling between the probe and the standard test block.

[0017] With the clamping force maintained stably, a synchronous trigger command is sent to the nonlinear ultrasonic detector to control the detector to output a second fundamental excitation signal that is completely consistent with the parameters of the inherent nonlinear signal acquisition stage. The excitation signal is then transmitted to the ultrasonic emission crystal of the automatic probe, which controls the probe to vertically emit an ultrasonic fundamental wave with the same parameters as the excitation signal into the interior of the undamaged standard test block.

[0018] Within the synchronous time window of the ultrasonic fundamental wave transmission, the ultrasonic receiving chip of the automatic probe acquires the full-time domain full-wave echo signal generated during the propagation of the ultrasonic fundamental wave inside the non-destructive standard test block. This covers the full waveform data of the fundamental wave echo, the echoes of each higher harmonic, and the interface reflection wave, completing a single full-wave signal acquisition. Simultaneously, the real-time clamping force data corresponding to this acquisition is locked, realizing a one-to-one correspondence between the signal and the clamping force parameters.

[0019] Under the current stable clamping force, the fundamental wave excitation and full-wave signal acquisition process is repeated a preset number of times to obtain multiple sets of full-wave echo signals under the same clamping force. Coherent averaging is performed on the multiple sets of full-wave echo signals to eliminate random noise and environmental electromagnetic interference components introduced by a single acquisition, and only retain the effective coherent signal that is strongly correlated with the ultrasonic excitation. After processing, the original calibration signal with a high signal-to-noise ratio under the current clamping force is obtained.

[0020] The original calibration signal is converted from time to frequency to separate the fundamental component and the higher harmonic components of the preset order. The amplitude and phase core features of the fundamental component and the amplitude and phase features of each higher harmonic component are extracted. Based on the extracted fundamental and higher harmonic features, the total nonlinear parameter under the current clamping force is calculated according to the standardized calculation logic of nonlinear ultrasonic parameters. This parameter includes three types of components: inherent nonlinearity, clamping force coupled additional nonlinearity, and background nonlinearity of the non-damaged standard test block material.

[0021] In conjunction with the first aspect, in the second embodiment of the first aspect of this application, the step of establishing a mapping reference library between clamping force and corresponding additional nonlinear signal, and calibrating the background nonlinear threshold, includes:

[0022] The corresponding data of full-gradient clamping force and additional nonlinear interference signal are retrieved. The clamping force value is used as the independent variable, and the amplitude and phase characteristics of the additional nonlinear interference signal are used as the dependent variable. An interpolation fitting algorithm is used to generate a continuous mapping function and construct a mapping benchmark library.

[0023] The robotic arm is controlled to move the automatic probe to the calibration area of ​​the undamaged base material of the pressure-bearing component to be tested. A standard clamping force value is selected, and the robotic arm is driven to adjust the probe clamping state, triggering ultrasonic excitation and signal acquisition to obtain the original base material signal of the current area. The inherent nonlinear signal is subtracted from the original base material signal, and the additional nonlinear interference signal corresponding to the standard clamping force value is retrieved from the mapping reference library and further subtracted to obtain the nonlinear signal of the base material. The fundamental and higher harmonic characteristics of the nonlinear signal of the base material are extracted, and the material background nonlinear parameters of the undamaged base material are calculated. Based on the material background nonlinear parameters superimposed with a preset error redundancy coefficient, the background nonlinear threshold is calibrated.

[0024] In conjunction with the first aspect, in the third embodiment of the first aspect of this application, the step of dividing the test area of ​​the pressure-bearing component under test into a grid and collecting the original field test signal corresponding to each grid cell includes:

[0025] Based on the geometry of the pressure-bearing component to be tested and the preset detection range, the detection area is divided into equally spaced grids, and the robotic arm motion trajectory, probe pose parameters and detection execution sequence corresponding to each grid unit are determined.

[0026] Following the movement trajectory of the robotic arm, the robotic arm is driven to move the automatic probe to each grid cell in sequence. The probe pressing stroke is monitored and adjusted in real time, triggering the nonlinear ultrasonic detector to output the third fundamental wave excitation signal. The automatic probe is then controlled to emit the ultrasonic fundamental wave into the pressure-bearing component under test. The full wave signal of the corresponding grid cell is collected, and coherent averaging processing is performed to obtain the original on-site detection signal corresponding to each grid cell.

[0027] In conjunction with the first aspect, in the fourth embodiment of the first aspect of this application, the step of filtering abnormal candidate regions based on effective nonlinear parameters and a background nonlinear threshold includes:

[0028] The effective nonlinear parameters corresponding to all grid cells are compared one by one with the background nonlinear threshold; grid cells whose effective nonlinear parameters exceed the background nonlinear threshold are selected and marked as anomalous candidate grid cells; the spatial coordinate information corresponding to all anomalous candidate grid cells is integrated and combined with the distribution boundary of the grid cells to delineate the anomalous candidate region.

[0029] In conjunction with the first aspect, in the fifth embodiment of the first aspect of this application, the step of inverting and calculating the equivalent emission source parameters of the anomalous nonlinear signal for all anomalous candidate regions includes:

[0030] By combining the isotropic properties of the metallic material of the pressure-bearing component under test and the propagation law of nonlinear ultrasound in solid media, a forward model of nonlinear ultrasound propagation is constructed.

[0031] An objective function for inversion is constructed, which is logically defined as the deviation between the simulated received signal characteristics output by the forward model and the anomalous nonlinear signal. The goal of inversion is to minimize this deviation by adjusting the parameters of the source to be determined. Reasonable value ranges that meet the constraints are defined for the parameters of the equivalent source. The value ranges of each parameter are then divided into equally spaced grids to form a discrete set of initial parameter grid points. Each initial parameter grid point is substituted into the nonlinear ultrasonic propagation forward model to calculate the corresponding objective function deviation. Several sets of initial parameter grid points with the smallest deviations are selected. These initial parameter grid points are then substituted into the adapted global optimization algorithm to initiate the iterative solution process, thereby calculating the equivalent source parameters of the anomalous nonlinear signal.

[0032] Preferably, a forward model is used as the forward calculation carrier to construct an objective function for inversion solution. The logic of the objective function is the quantitative calculation of multi-dimensional comprehensive deviation: the parameters of the source to be determined are input into the forward model, and the corresponding full-dimensional features of the simulated received signal are output; the features of the simulated received signal are compared with the features of the actual acquired abnormal nonlinear incremental signal in each dimension, and the amplitude deviation, phase deviation, propagation delay deviation, and spatial distribution matching deviation between multiple acquisition points are quantified respectively; combined with the sensitivity of different harmonics to material micro-damage, differentiated weights are assigned to the deviations in each dimension, and the multi-dimensional deviation values ​​are integrated into a single comprehensive deviation value. The objective function finally outputs this comprehensive deviation value, realizing the quantification of the difference between the simulated signal and the actual abnormal signal.

[0033] Rigid constraints are set throughout the solution process of the objective function to avoid the ill-posedness of the inversion problem and ensure that the solution results conform to physical laws and engineering realities. The constraints include: the three-dimensional spatial coordinates of the emission source must fall within the solid geometric boundaries and wall thickness range of the pressure-bearing component under test; the lower limit of the emission source radiation intensity must not be lower than the calibrated background nonlinear threshold, and the upper limit must not exceed the physical limit of the nonlinear response of the tested metallic material; the ultrasonic signal propagation delay must conform to the causal correspondence between the material's sound velocity and propagation distance; and the frequency band characteristics of the simulated received signal must match the frequency band range of the actual excitation signal. Any combination of emission source parameters that exceeds the constraints will be directly assigned a large deviation value and excluded from the effective solution range.

[0034] For each transmitter parameter to be determined, a reasonable range of values ​​that meets the constraints is defined by combining the spatial range of the anomaly candidate region and the characteristic amplitude of the actual anomaly signal, forming a multi-dimensional parameter optimization space. The range of values ​​of each parameter in the optimization space is discretized at equal intervals to generate several sets of discrete initial transmitter parameter combinations. Each set of initial parameter combinations is sequentially input into the forward model to calculate the corresponding simulated received signal characteristics. The comprehensive deviation value corresponding to each set of parameters is calculated through the objective function. Several sets of parameter combinations with the smallest comprehensive deviation values ​​are selected.

[0035] The optimized initial parameter combination is substituted into the adapted global optimization algorithm to initiate iterative optimization. Throughout the iteration process, the comprehensive deviation value output by the objective function is used as the core judgment criterion. A single iteration proceeds as follows: the algorithm adjusts the values ​​of the transmitter parameters according to the preset optimization logic, inputs the adjusted parameter combination into the forward model, and calculates the corresponding simulated received signal characteristics; the comprehensive deviation value corresponding to this set of parameters is calculated through the objective function, and the change in deviation value between the current iteration and the previous iteration is compared to determine the parameter adjustment direction and step size for the next iteration; the parameter combination is checked throughout the process to ensure it meets the constraints, and invalid parameter combinations are directly eliminated. The iteration stops when the preset termination conditions are met. Termination conditions include the comprehensive deviation value being lower than the preset convergence accuracy threshold and the number of iterations reaching the preset maximum safe iteration number. If either condition is met, the iteration terminates, and the transmitter parameter combination with the globally minimum comprehensive deviation value is output as the initial optimal solution for inversion.

[0036] If multiple independent anomalous nonlinear signal emission sources exist within the anomaly candidate region, a step-by-step separation solution based on the objective function is performed. The specific process is as follows: After obtaining the optimal solution for the first emission source, input the parameters of that emission source into the forward model to calculate the signal contribution of that emission source at each acquisition point; deduct the full amount of this signal contribution from the total actual acquired anomalous nonlinear incremental signal to obtain the remaining anomalous signal components; using the remaining anomalous signal components as the new comparison benchmark, reconstruct the objective function, and repeat the entire process of parameter optimization, iterative solution, and verification until the amplitude of the remaining signal components is lower than the calibrated background nonlinear threshold, thus completing the parameter inversion solution for all independent anomalous nonlinear signal emission sources.

[0037] In conjunction with the first aspect, in the sixth embodiment of the first aspect of this application, the step of constructing a nonlinear ultrasonic propagation forward model by combining the isotropic properties of the metallic material of the pressure-bearing component under test and the propagation law of nonlinear ultrasound in a solid medium includes:

[0038] Based on the component's physical structure, a spatial grid partitioning algorithm is used to discretize the ultrasonic propagation region, assigning isotropic material parameters to the grid cells and setting medium boundary constraints. Based on the linear ultrasonic dynamics, a numerical iterative solution algorithm is used to simulate the propagation process of the fundamental wave in an isotropic medium, obtaining the global fundamental wave sound field distribution characteristics. Combining the nonlinear interaction mechanism of the medium, a nonlinear iterative algorithm is used to calculate the initial amplitude and phase distribution of higher harmonics generated by the interaction between the fundamental wave and the anomalous emission source. Following the grid discretization format, a wave numerical transfer algorithm is used to solve for the sound field characteristics of higher harmonics propagating in the medium to the probe receiving surface. Through an acoustic-electric response fitting algorithm, the sound field characteristics of the receiving surface are converted into the probe output electrical signal, extracting the fundamental wave and higher harmonic characteristic quantities.

[0039] Using the parameters of the abnormal nonlinear signal source as independent variables, the characteristics of the global fundamental sound field distribution, the initial amplitude and phase distribution of higher harmonics, the sound field characteristics of the probe receiving surface, and the characteristic quantities of the fundamental and higher harmonics as full-link mapping nodes, and the signal characteristics of the probe receiving end as the final dependent variable, a multivariate mapping fitting algorithm is used to establish a stable positive mapping relationship between the independent variables and the final dependent variable through each mapping node, thus completing the construction of the forward model.

[0040] In conjunction with the first aspect, in the seventh embodiment of the first aspect of this application, determining the abnormal location region based on the three-dimensional spatial coordinates of the equivalent emission source parameters and outputting the location features of the abnormal location region includes:

[0041] Based on the three-dimensional spatial coordinates of the equivalent emission source parameters and the grid distribution of the anomaly candidate region, the anomaly location region is determined. Centered on the anomaly location region, the detection area is divided into secondary grids according to the preset encrypted grid size. The robotic arm drives the automatic probe to perform supplementary sampling and detection along the encrypted grid path. The nonlinear interference elimination process is repeated to obtain the effective nonlinear parameters to be measured corresponding to each encrypted grid unit. Based on the distribution of the effective nonlinear parameters to be measured and combined with the radiation intensity attenuation characteristics of the equivalent emission source, the boundary of the anomaly location region is delineated, and the location characteristics of the anomaly location region are output, including three-dimensional spatial coordinate information and boundary size parameters.

[0042] In conjunction with the first aspect, in the eighth embodiment of the first aspect of this application, the step of using an artificial intelligence algorithm to train a safety assessment model, inputting location features into the safety assessment model, and calculating the safety assessment result of the pressure-bearing component includes:

[0043] The safety assessment model employs the LightGBM lightweight gradient boosting decision tree AI algorithm for supervised training. The model is trained and validated using historical defect detection data of pressure-bearing components in thermal power units, shutdown dissection verification results, and full lifecycle operation data. Based on the location characteristics of abnormal areas, the model matches the corresponding region's effective nonlinear parameters, equivalent emission source radiation intensity, pressure-bearing component material parameters, design parameters, and historical operating condition data to complete feature normalization preprocessing, forming a standardized input set. This standardized input set is then loaded into the pre-trained safety assessment model. Through multi-decision tree hierarchical iterative calculations, the model outputs safety assessment results, including the damage severity level corresponding to the defect severity, the defect type determination result matching nonlinear signal characteristics, and the remaining safe operating life assessment result combining material degradation patterns and operating conditions.

[0044] Secondly, this application provides a safety assessment system for pressure-bearing components based on nonlinear ultrasound, including:

[0045] The background nonlinear threshold calibration module includes: a nonlinear signal acquisition unit that controls the robotic arm to drive the automatic probe to acquire inherent nonlinear signals and additional nonlinear signals; and a background nonlinear threshold calibration unit that establishes a mapping reference library between clamping force and corresponding additional nonlinear signals to calibrate the background nonlinear threshold.

[0046] The abnormal candidate region screening module includes: a field detection raw signal acquisition unit that divides the detection area of ​​the pressure-bearing component under test into a grid and acquires the field detection raw signal corresponding to each grid unit; an effective nonlinear parameter calculation unit that removes nonlinear interference for each field detection raw signal and calculates the effective nonlinear parameter; and an abnormal candidate region screening unit that filters abnormal candidate regions based on the effective nonlinear parameter and the background nonlinear threshold.

[0047] Location feature calculation module: includes: an equivalent emission source parameter inversion unit that inverts and calculates the equivalent emission source parameters of the abnormal nonlinear signal for all anomaly candidate regions; and a location feature calculation unit that determines the anomaly location region based on the three-dimensional spatial coordinates of the equivalent emission source parameters and outputs the location features of the anomaly location region.

[0048] Safety assessment module: includes: a safety assessment unit that uses artificial intelligence algorithms to train a safety assessment model, inputs location features into the safety assessment model, and calculates the safety assessment result of the pressure-bearing component.

[0049] Compared with the prior art, the beneficial effects of the present invention are:

[0050] 1. This invention collects additional nonlinear signals under gradient clamping force and constructs a mapping benchmark library between clamping force and additional nonlinear signals. This can accurately eliminate nonlinear interference caused by clamping coupling, achieve accurate calibration of the background nonlinear threshold, significantly improve the signal-to-noise ratio of the detection signal, and eliminate interference signals from affecting the detection results.

[0051] 2. This invention is based on a nonlinear ultrasonic propagation forward model to perform inversion calculations on anomaly candidate regions, accurately obtain the equivalent emission source parameters of the abnormal nonlinear signals, and combine with densified grid supplementary sampling detection to delineate the boundary range of the defect location, realize the precise positioning of the defect region, and meet the positioning accuracy requirements of actual engineering.

[0052] 3. This invention trains a dedicated safety assessment model using artificial intelligence algorithms. After substituting the defect location features into the model, it can quantitatively output the damage level, defect type, and remaining safe operating life, providing a complete assessment result with practical engineering guidance significance, which is in line with the safety operation and maintenance needs of pressure-bearing components of thermal power units. Attached Figure Description

[0053] Figure 1 This is a schematic diagram illustrating the steps of the safety assessment method for pressure-bearing components based on nonlinear ultrasound according to the present invention;

[0054] Figure 2 This is a system structure diagram of the pressure-bearing component safety assessment system based on nonlinear ultrasound of the present invention. Detailed Implementation

[0055] 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, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0056] Example: Figures 1-2 As shown, the present invention provides a technical solution:

[0057] like Figure 1 As shown, this application provides a safety assessment method for pressure-bearing components based on nonlinear ultrasound, including the following steps:

[0058] Step S100: Control the robotic arm to drive the automatic probe to collect inherent nonlinear signals and additional nonlinear signals; establish a mapping benchmark library between clamping force and corresponding additional nonlinear signals, and calibrate the background nonlinear threshold;

[0059] Specifically, several sets of continuously gradient-distributed clamping forces are preset, and the range of the clamping forces covers the minimum effective coupling clamping force to the maximum safe clamping force required for on-site detection.

[0060] The robotic arm is controlled to move the automatic probe into a suspended, non-contact state, triggering the nonlinear ultrasonic detector to output the first fundamental excitation signal, acquiring the first full-wave signal, and calculating the inherent nonlinear signal under zero clamping force.

[0061] The robotic arm is controlled to clamp the actuator, the probe clamping stroke is adjusted, and the nonlinear ultrasonic detector is triggered to output the second fundamental wave excitation signal. The automatic probe is controlled to emit the ultrasonic fundamental wave into the interior of the undamaged standard test block, and the second full wave signal is collected. Coherent averaging is performed to obtain the original calibration signal under the current clamping force. The fundamental wave and each higher harmonic characteristic of the original calibration signal are extracted, and the total nonlinear parameter under the current clamping force is calculated. The inherent nonlinear signal is subtracted from the total nonlinear parameter to obtain the additional nonlinear signal corresponding to the current clamping force.

[0062] Furthermore, the corresponding data of the full gradient clamping force and the additional nonlinear interference signal are retrieved. The clamping force value is used as the independent variable, and the amplitude and phase characteristics of the additional nonlinear interference signal are used as the dependent variable. An interpolation fitting algorithm is used to generate a continuous mapping function and construct a mapping benchmark library.

[0063] The robotic arm is controlled to move the automatic probe to the calibration area of ​​the undamaged base material of the pressure-bearing component to be tested. A standard clamping force value is selected, and the robotic arm is driven to adjust the probe clamping state, triggering ultrasonic excitation and signal acquisition to obtain the original base material signal of the current area. The inherent nonlinear signal is subtracted from the original base material signal, and the additional nonlinear interference signal corresponding to the standard clamping force value is retrieved from the mapping reference library and further subtracted to obtain the nonlinear signal of the base material. The fundamental and higher harmonic characteristics of the nonlinear signal of the base material are extracted, and the material background nonlinear parameters of the undamaged base material are calculated. Based on the material background nonlinear parameters superimposed with a preset error redundancy coefficient, the background nonlinear threshold is calibrated.

[0064] In one specific embodiment, a common 12Cr1MoV steel pressure-bearing component of thermal power units is selected as the test object, and a non-destructive standard test block of the same material is provided. The preset gradient clamping force range is 50N~300N. With 50N as the gradient interval, six sets of continuous clamping forces are set sequentially, namely 50N, 100N, 150N, 200N, 250N, and 300N, to fully cover the minimum effective coupling clamping force and the maximum safe clamping force for on-site testing.

[0065] The robotic arm is controlled to move the automatic probe into a suspended, non-contact state, triggering the nonlinear ultrasonic detector to output a 1MHz fundamental excitation signal. The full-wave signal is collected and calculated, and the inherent nonlinear signal parameter of the system under zero clamping force is obtained as 0.082. This value is the inherent nonlinear value brought about by the hardware and circuit of the detection system itself, and is independent of external clamping or the workpiece being tested.

[0066] A clamping force of 150N was selected for calibration. The robotic arm was controlled to press the probe, triggering a 1MHz second fundamental excitation signal. Multiple sets of full-wave signals were acquired and coherently averaged to obtain the original calibration signal. The fundamental and second and third harmonic characteristics were extracted, and the total nonlinear parameter under this clamping force was calculated to be 0.147. After subtracting the inherent nonlinear signal under zero clamping force, the additional nonlinear signal parameter corresponding to this clamping force was obtained to be 0.065.

[0067] All six sets of gradient clamping forces were sequentially acquired using additional nonlinear signals. The parameters of the additional nonlinear signals for each set were 0.021, 0.043, 0.065, 0.087, 0.109, and 0.131, respectively. Using clamping force as the independent variable and the amplitude and phase of the additional nonlinear signals as the dependent variables, a continuous mapping function was generated through an interpolation fitting algorithm, thus completing the construction of a mapping benchmark library for clamping force and additional nonlinear signals.

[0068] Using 200N as the standard clamping force, signals were collected in the undamaged area of ​​the parent material of the component under test. After deducting the inherent nonlinearity and the additional nonlinear interference from the mapping library matching, the nonlinear signal of the parent material was obtained, and the material background nonlinear parameter was calculated to be 0.093. After superimposing a 5% preset error redundancy coefficient, the final calibration yielded a background nonlinear threshold of 0.098.

[0069] Step S200: Divide the test area of ​​the pressure-bearing component under test into a grid, and collect the original field test signal corresponding to each grid cell; remove nonlinear interference for each original field test signal and calculate the effective nonlinear parameter; based on the effective nonlinear parameter, combine it with the background nonlinear threshold to screen abnormal candidate areas;

[0070] Specifically, based on the geometry of the pressure-bearing component to be tested and the preset detection range, the detection area is divided into equally spaced grids, and the robotic arm motion trajectory, probe pose parameters and detection execution sequence corresponding to each grid unit are determined.

[0071] Following the movement trajectory of the robotic arm, the robotic arm is driven to move the automatic probe to each grid cell in sequence. The probe pressing stroke is monitored and adjusted in real time, triggering the nonlinear ultrasonic detector to output the third fundamental wave excitation signal. The automatic probe is then controlled to emit the ultrasonic fundamental wave into the pressure-bearing component under test. The full wave signal of the corresponding grid cell is collected, and coherent averaging processing is performed to obtain the original on-site detection signal corresponding to each grid cell.

[0072] Furthermore, the effective nonlinear parameters corresponding to all grid cells are compared one by one with the background nonlinear threshold; grid cells whose effective nonlinear parameters exceed the background nonlinear threshold are selected and marked as abnormal candidate grid cells; the spatial coordinate information corresponding to all abnormal candidate grid cells is integrated and combined with the distribution boundary of the grid cells to delineate the abnormal candidate region.

[0073] In one specific embodiment, the component to be tested is selected as a 12Cr1MoV steel main steam pipe. A core detection area of ​​300mm×200mm is defined, and the area is divided into grids with equal intervals of 10mm×10mm, generating a total of 600 regular grid units. The motion trajectory of the robotic arm is planned to scan line by line, and the fixed posture of vertical probe clamping and standard clamping force of 200N is set. The detection of each point is completed in sequence from left to right and from top to bottom.

[0074] The nonlinear ultrasonic detector is triggered to output a 1MHz third fundamental wave excitation signal, which drives the robotic arm to traverse each grid cell in sequence. Five sets of full-wave signals are collected at each point and coherent averaging is performed to eliminate electromagnetic noise and random noise on site, so as to obtain the original on-site detection signal corresponding to each grid cell. The clamping force is adjusted in real time throughout the process and stably maintained within the calibration range of 200N±2N.

[0075] Interference removal was performed on the original signals of each grid, successively deducting the inherent nonlinear parameter of 0.082 and the additional nonlinear parameter of 0.087 matching the 200N clamping force. Finally, the effective nonlinear parameters of each point were calculated. The effective parameters of the undamaged parent material area were stable between 0.091 and 0.097, all of which were lower than the calibrated background nonlinear threshold of 0.098.

[0076] After threshold comparison, a total of 12 grid cells with excessive nonlinear parameters were selected. The corresponding parameter values ​​were between 0.123 and 0.176. This group of grid cells was distributed in a continuous and concentrated manner. By integrating the spatial coordinates of each cell, a continuous anomaly candidate region of 50mm×40mm was finally delineated.

[0077] Step S300: For all candidate anomaly regions, invert and calculate the equivalent emission source parameters of the anomaly nonlinear signal; based on the three-dimensional spatial coordinates of the equivalent emission source parameters, determine the anomaly location region and output the location characteristics of the anomaly location region;

[0078] Specifically, a forward model for nonlinear ultrasonic propagation is constructed by combining the isotropic properties of the metallic material of the pressure-bearing component under test and the propagation law of nonlinear ultrasound in solid media.

[0079] An objective function for inversion is constructed, which is logically defined as the deviation between the simulated received signal characteristics output by the forward model and the anomalous nonlinear signal. The goal of inversion is to minimize this deviation by adjusting the parameters of the source to be determined. Reasonable value ranges that meet the constraints are defined for the parameters of the equivalent source. The value ranges of each parameter are then divided into equally spaced grids to form a discrete set of initial parameter grid points. Each initial parameter grid point is substituted into the nonlinear ultrasonic propagation forward model to calculate the corresponding objective function deviation. Several sets of initial parameter grid points with the smallest deviations are selected. These initial parameter grid points are then substituted into the adapted global optimization algorithm to initiate the iterative solution process, thereby calculating the equivalent source parameters of the anomalous nonlinear signal.

[0080] Furthermore, based on the component's physical structure, a spatial grid partitioning algorithm is used to discretize the ultrasonic propagation region, assigning isotropic material parameters to the grid cells and setting medium boundary constraints. Based on the linear ultrasonic dynamic law, a numerical iterative solution algorithm is used to simulate the propagation process of the fundamental wave in an isotropic medium, obtaining the global fundamental wave sound field distribution characteristics. Combining the nonlinear interaction mechanism of the medium, a nonlinear iterative algorithm is used to calculate the initial amplitude and phase distribution of higher harmonics generated by the interaction between the fundamental wave and the anomalous emission source. Following the grid discretization format, a wave numerical transfer algorithm is used to solve the sound field characteristics of higher harmonics propagating in the medium to the probe receiving surface. Through an acoustic-electric response fitting algorithm, the sound field characteristics of the receiving surface are converted into the probe output electrical signal, and the fundamental wave and higher harmonic characteristic quantities are extracted.

[0081] Using the parameters of the abnormal nonlinear signal source as independent variables, the characteristics of the global fundamental sound field distribution, the initial amplitude and phase distribution of higher harmonics, the sound field characteristics of the probe receiving surface, and the characteristic quantities of the fundamental and higher harmonics as full-link mapping nodes, and the signal characteristics of the probe receiving end as the final dependent variable, a multivariate mapping fitting algorithm is used to establish a stable positive mapping relationship between the independent variables and the final dependent variable through each mapping node, thus completing the construction of the forward model.

[0082] Furthermore, based on the three-dimensional spatial coordinates of the equivalent emission source parameters and the grid distribution of the anomaly candidate region, the anomaly location region is determined. Centered on the anomaly location region, the detection area is divided into secondary grids according to the preset encrypted grid size. The robotic arm drives the automatic probe to perform supplementary sampling and detection along the encrypted grid path. The nonlinear interference elimination process is repeated to obtain the effective nonlinear parameters to be measured corresponding to each encrypted grid unit. Based on the distribution of the effective nonlinear parameters to be measured and combined with the radiation intensity attenuation characteristics of the equivalent emission source, the boundary of the anomaly location region is delineated, and the location characteristics of the anomaly location region are output, including three-dimensional spatial coordinate information and boundary size parameters.

[0083] In one specific embodiment, for the defined 50mm×40mm anomaly candidate region, a nonlinear ultrasonic propagation forward model is built based on the isotropic acoustic parameters of the 12Cr1MoV steel pipe. The entire ultrasonic propagation computational domain is spatially meshed using a uniform step size of 1mm, generating a total of 124,000 regular mesh elements. Fixed parameters such as longitudinal wave velocity, material nonlinearity coefficient, and sound wave attenuation coefficient are uniformly assigned to all mesh elements. At the same time, physical constraints are set for the free boundary of the outer wall of the pipe and the medium contact boundary of the inner wall, which fully conforms to the physical structure of the pressure-bearing pipe and the sound wave propagation law.

[0084] Based on the established forward model, an objective function for inversion was constructed to quantify the comprehensive deviation of the amplitude, phase, and propagation delay between the forward-modeled received signal and the field-measured abnormal nonlinear signal. A deviation convergence threshold of 0.002 was set. Simultaneously, reasonable value ranges for various parameters of the equivalent emission source were defined, strictly limiting the three-dimensional coordinates to the selected anomaly candidate region, and parameters such as radiation intensity and equivalent size to the physical range of the nonlinear response of the metallic material. The parameter value ranges were discretized at equal intervals, generating 240 initial parameter grid points. These were successively substituted into the forward model to calculate the corresponding objective function deviation values. The eight sets of parameters with the smallest deviation values ​​were selected as the initial values ​​for iterative optimization. Then, a global optimization algorithm was used for iterative solving. After 46 iterations, the deviation value decreased below the convergence threshold, ultimately yielding accurate equivalent emission source parameters: three-dimensional spatial coordinates of 1265mm axial direction, 38° circumferential direction, and 12.5mm radial wall thickness; radiation intensity of 0.169; and equivalent defect size of 3.2mm.

[0085] Using the equivalent emission source's three-dimensional coordinates obtained through inversion as the core, the core anomaly point was located. The original 10mm×10mm conventional detection grid was densified to a finer 2mm×2mm grid, completing the secondary meshing of the core anomaly region and generating a total of 400 densified grid units. The robotic arm was driven to move the probe along the densified grid path to collect and detect points one by one, maintaining a standard clamping force of 200N throughout the process. The process of eliminating inherent nonlinear interference and clamping force-added nonlinear interference was repeated to measure the effective nonlinear parameters of each densified grid unit. The data showed a clear gradient distribution, with the effective nonlinear parameter value being higher the closer to the core coordinates of the equivalent emission source. The parameter at the core point reached 0.176, while the value at the peripheral points gradually decreased towards the background threshold.

[0086] Based on the attenuation law of the radiation intensity of the equivalent emission source from the inside to the outside, the effective nonlinear parameters of the densified grid are judged layer by layer, and the edge points with slight excesses are eliminated. Finally, the complete boundary of the abnormal location is accurately delineated. The actual external dimensions of the defect area are measured to be 36mm×32mm, and the burial depth range in the pipe wall thickness direction is 11.8mm-13.2mm. The three-dimensional spatial coordinates, boundary dimensions, burial depth range and other complete set of positional features of the abnormal area are output simultaneously. The overall positioning deviation is controlled within 0.5mm, which fully meets the defect positioning accuracy requirements of the pressure-bearing components of thermal power units.

[0087] Step S400: Use artificial intelligence algorithms to train a safety assessment model, input the location features into the safety assessment model, and calculate the safety assessment result of the pressure-bearing component.

[0088] Specifically, the safety assessment model employs the LightGBM lightweight gradient boosting decision tree AI algorithm for supervised training. The model is trained and validated using historical defect detection data of pressure-bearing components of thermal power units, shutdown dissection verification results, and full life-cycle operation data. Based on the location characteristics of abnormal locations, the model matches the corresponding region's effective nonlinear parameters, equivalent emission source radiation intensity, pressure-bearing component material parameters, design parameters, and historical operating condition data to complete feature normalization preprocessing, forming a standardized input set. The standardized input set is then loaded into the pre-trained safety assessment model. The model outputs safety assessment results through multi-decision tree hierarchical iterative calculations, including the damage severity level corresponding to the defect severity, the defect type determination result matching nonlinear signal characteristics, and the remaining safe operating life assessment result combining material degradation patterns and operating conditions.

[0089] In one specific embodiment, the LightGBM lightweight gradient boosting decision tree algorithm was used to build a safety assessment model. 1200 sets of historical defect data of pressure-bearing components of 12Cr1MoV steel thermal power units, 320 sets of shutdown dissection test results, and annual operating condition data of the units were collected to complete supervised training. After cross-validation, the model assessment accuracy reached 98.7%, which meets the assessment accuracy requirements of industrial sites.

[0090] By integrating the three-dimensional coordinates, boundary dimensions, burial depth, and other location features of this defect, matching the equivalent emission source radiation intensity of 0.169, the effective nonlinear parameter peak value of 0.176, and the pipeline design parameters, service duration, and operating temperature and pressure data, all features are normalized to generate a standardized model input set.

[0091] The standardized input set is imported into the trained model, and the final safety assessment result is output through multi-decision tree iterative reasoning: the degree of defect damage is determined to be Level II mild damage, the defect type is fatigue microcracks on the inner wall of the pipe, and the remaining safe operating life is assessed as 18,200 hours of continuous operation under full load conditions.

[0092] like Figure 2 As shown, this application provides a safety assessment system for pressure-bearing components based on nonlinear ultrasound, including:

[0093] The background nonlinear threshold calibration module includes: a nonlinear signal acquisition unit that controls the robotic arm to drive the automatic probe to acquire inherent nonlinear signals and additional nonlinear signals; and a background nonlinear threshold calibration unit that establishes a mapping reference library between clamping force and corresponding additional nonlinear signals to calibrate the background nonlinear threshold.

[0094] The abnormal candidate region screening module includes: a field detection raw signal acquisition unit that divides the detection area of ​​the pressure-bearing component under test into a grid and acquires the field detection raw signal corresponding to each grid unit; an effective nonlinear parameter calculation unit that removes nonlinear interference for each field detection raw signal and calculates the effective nonlinear parameter; and an abnormal candidate region screening unit that filters abnormal candidate regions based on the effective nonlinear parameter and the background nonlinear threshold.

[0095] Location feature calculation module: includes: an equivalent emission source parameter inversion unit that inverts and calculates the equivalent emission source parameters of the abnormal nonlinear signal for all anomaly candidate regions; and a location feature calculation unit that determines the anomaly location region based on the three-dimensional spatial coordinates of the equivalent emission source parameters and outputs the location features of the anomaly location region.

[0096] Safety assessment module: includes: a safety assessment unit that uses artificial intelligence algorithms to train a safety assessment model, inputs location features into the safety assessment model, and calculates the safety assessment result of the pressure-bearing component.

[0097] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

Claims

1. A safety assessment method for pressure-bearing components based on nonlinear ultrasound, characterized in that, Includes the following steps: The robotic arm is controlled to drive an automatic probe to collect inherent nonlinear signals and additional nonlinear signals; a mapping benchmark library between clamping force and corresponding additional nonlinear signals is established, and the background nonlinear threshold is calibrated. The testing area of ​​the pressure-bearing component to be tested is divided into grids, and the original field test signals corresponding to each grid unit are collected. For each original signal detected on-site, nonlinear interference is removed, and the effective nonlinear parameter is calculated. Based on the effective nonlinear parameter, anomaly candidate regions are screened in combination with the background nonlinear threshold. For all candidate anomaly regions, the equivalent emission source parameters of the anomalous nonlinear signal are inverted and calculated; based on the three-dimensional spatial coordinates of the equivalent emission source parameters, the anomaly location region is determined, and the location characteristics of the anomaly location region are output. The safety assessment model is trained using artificial intelligence algorithms. The location features are input into the safety assessment model to calculate the safety assessment result of the pressure-bearing component.

2. The method for safety assessment of pressure-bearing components based on nonlinear ultrasound according to claim 1, characterized in that, The controlled robotic arm drives an automatic probe to collect inherent nonlinear signals and additional nonlinear signals, including: Several sets of continuously gradient-distributed clamping forces are preset, and the range of the clamping forces covers the minimum effective coupling clamping force to the maximum safe clamping force required for on-site detection. The robotic arm is controlled to move the automatic probe into a suspended, non-contact state, triggering the nonlinear ultrasonic detector to output the first fundamental excitation signal, acquiring the first full-wave signal, and calculating the inherent nonlinear signal under zero clamping force. The robotic arm is controlled to clamp the actuator, the probe clamping stroke is adjusted, and the nonlinear ultrasonic detector is triggered to output the second fundamental wave excitation signal. The automatic probe is controlled to emit the ultrasonic fundamental wave into the interior of the undamaged standard test block, and the second full wave signal is collected. Coherent averaging is performed to obtain the original calibration signal under the current clamping force. The fundamental wave and each higher harmonic characteristic of the original calibration signal are extracted, and the total nonlinear parameter under the current clamping force is calculated. The inherent nonlinear signal is subtracted from the total nonlinear parameter to obtain the additional nonlinear signal corresponding to the current clamping force.

3. The method for safety assessment of pressure-bearing components based on nonlinear ultrasound according to claim 1, characterized in that, The establishment of a mapping reference library between clamping force and corresponding additional nonlinear signal, and the calibration of the background nonlinear threshold, include: The corresponding data of full-gradient clamping force and additional nonlinear interference signal are retrieved. The clamping force value is used as the independent variable, and the amplitude and phase characteristics of the additional nonlinear interference signal are used as the dependent variable. An interpolation fitting algorithm is used to generate a continuous mapping function and construct a mapping benchmark library. The robotic arm is controlled to move the automatic probe to the calibration area of ​​the undamaged base material of the pressure-bearing component to be tested. A standard clamping force value is selected, and the robotic arm is driven to adjust the probe clamping state, triggering ultrasonic excitation and signal acquisition to obtain the original base material signal of the current area. The inherent nonlinear signal is subtracted from the original base material signal, and the additional nonlinear interference signal corresponding to the standard clamping force value is retrieved from the mapping reference library and further subtracted to obtain the nonlinear signal of the base material. The fundamental and higher harmonic characteristics of the nonlinear signal of the base material are extracted, and the material background nonlinear parameters of the undamaged base material are calculated. Based on the material background nonlinear parameters superimposed with a preset error redundancy coefficient, the background nonlinear threshold is calibrated.

4. The method for safety assessment of pressure-bearing components based on nonlinear ultrasound according to claim 1, characterized in that, The testing area of ​​the pressure-bearing component to be tested is divided into grids, and the original field test signals corresponding to each grid cell are collected, including: Based on the geometry of the pressure-bearing component to be tested and the preset detection range, the detection area is divided into equally spaced grids, and the robotic arm motion trajectory, probe pose parameters and detection execution sequence corresponding to each grid unit are determined. Following the movement trajectory of the robotic arm, the robotic arm is driven to move the automatic probe to each grid cell in sequence. The probe pressing stroke is monitored and adjusted in real time, triggering the nonlinear ultrasonic detector to output the third fundamental wave excitation signal. The automatic probe is then controlled to emit the ultrasonic fundamental wave into the pressure-bearing component under test. The full wave signal of the corresponding grid cell is collected, and coherent averaging processing is performed to obtain the original on-site detection signal corresponding to each grid cell.

5. The method for safety assessment of pressure-bearing components based on nonlinear ultrasound according to claim 1, characterized in that, The method of filtering abnormal candidate regions based on effective nonlinear parameters and background nonlinear thresholds includes: The effective nonlinear parameters corresponding to all grid cells are compared one by one with the background nonlinear threshold; grid cells whose effective nonlinear parameters exceed the background nonlinear threshold are selected and marked as anomalous candidate grid cells; the spatial coordinate information corresponding to all anomalous candidate grid cells is integrated and combined with the distribution boundary of the grid cells to delineate the anomalous candidate region.

6. The method for safety assessment of pressure-bearing components based on nonlinear ultrasound according to claim 1, characterized in that, The process of inverting and calculating the equivalent emission source parameters of the anomalous nonlinear signal for all candidate regions of anomalies includes: By combining the isotropic properties of the metallic material of the pressure-bearing component under test and the propagation law of nonlinear ultrasound in solid media, a forward model of nonlinear ultrasound propagation is constructed. An objective function for inversion is constructed, which is logically defined as the deviation between the simulated received signal characteristics output by the forward model and the anomalous nonlinear signal. The goal of inversion is to minimize this deviation by adjusting the parameters of the source to be determined. Reasonable value ranges that meet the constraints are defined for the parameters of the equivalent source. The value ranges of each parameter are then divided into equally spaced grids to form a discrete set of initial parameter grid points. Each initial parameter grid point is substituted into the nonlinear ultrasonic propagation forward model to calculate the corresponding objective function deviation. Several sets of initial parameter grid points with the smallest deviations are selected. These initial parameter grid points are then substituted into the adapted global optimization algorithm to initiate the iterative solution process, thereby calculating the equivalent source parameters of the anomalous nonlinear signal.

7. The method for safety assessment of pressure-bearing components based on nonlinear ultrasound according to claim 6, characterized in that, The nonlinear ultrasound propagation forward model is constructed by combining the isotropic properties of the metallic material of the pressure-bearing component under test and the propagation law of nonlinear ultrasound in a solid medium, including: Based on the component's physical structure, a spatial grid partitioning algorithm is used to discretize the ultrasonic propagation region, assigning isotropic material parameters to the grid cells and setting medium boundary constraints. Based on the linear ultrasonic dynamics, a numerical iterative solution algorithm is used to simulate the propagation process of the fundamental wave in an isotropic medium, obtaining the global fundamental wave sound field distribution characteristics. Combining the nonlinear interaction mechanism of the medium, a nonlinear iterative algorithm is used to calculate the initial amplitude and phase distribution of higher harmonics generated by the interaction between the fundamental wave and the anomalous emission source. Following the grid discretization format, a wave numerical transfer algorithm is used to solve for the sound field characteristics of higher harmonics propagating in the medium to the probe receiving surface. Through an acoustic-electric response fitting algorithm, the sound field characteristics of the receiving surface are converted into the probe output electrical signal, extracting the fundamental wave and higher harmonic characteristic quantities. Using the parameters of the abnormal nonlinear signal source as independent variables, the characteristics of the global fundamental sound field distribution, the initial amplitude and phase distribution of higher harmonics, the sound field characteristics of the probe receiving surface, and the characteristic quantities of the fundamental and higher harmonics as full-link mapping nodes, and the signal characteristics of the probe receiving end as the final dependent variable, a multivariate mapping fitting algorithm is used to establish a stable positive mapping relationship between the independent variables and the final dependent variable through each mapping node, thus completing the construction of the forward model.

8. The method for safety assessment of pressure-bearing components based on nonlinear ultrasound according to claim 1, characterized in that, The method of determining the abnormal location region based on the three-dimensional spatial coordinates of the equivalent emission source parameters and outputting the location characteristics of the abnormal location region includes: Based on the three-dimensional spatial coordinates of the equivalent emission source parameters and the grid distribution of the anomaly candidate region, the anomaly location region is determined. Centered on the anomaly location region, the detection area is divided into secondary grids according to the preset encrypted grid size. The robotic arm drives the automatic probe to perform supplementary sampling and detection along the encrypted grid path. The nonlinear interference elimination process is repeated to obtain the effective nonlinear parameters to be measured corresponding to each encrypted grid unit. Based on the distribution of the effective nonlinear parameters to be measured and combined with the radiation intensity attenuation characteristics of the equivalent emission source, the boundary of the anomaly location region is delineated, and the location characteristics of the anomaly location region are output, including three-dimensional spatial coordinate information and boundary size parameters.

9. The method for safety assessment of pressure-bearing components based on nonlinear ultrasound according to claim 1, characterized in that, The process of training a safety assessment model using artificial intelligence algorithms, inputting location features into the safety assessment model, and calculating the safety assessment result for the pressure-bearing component includes: The safety assessment model employs the LightGBM lightweight gradient boosting decision tree AI algorithm for supervised training. The model is trained and validated using historical defect detection data of pressure-bearing components in thermal power units, shutdown dissection verification results, and full lifecycle operation data. Based on the location characteristics of abnormal areas, the model matches the corresponding region's effective nonlinear parameters, equivalent emission source radiation intensity, pressure-bearing component material parameters, design parameters, and historical operating condition data to complete feature normalization preprocessing, forming a standardized input set. This standardized input set is then loaded into the pre-trained safety assessment model. Through multi-decision tree hierarchical iterative calculations, the model outputs safety assessment results, including the damage severity level corresponding to the defect severity, the defect type determination result matching nonlinear signal characteristics, and the remaining safe operating life assessment result combining material degradation patterns and operating conditions.

10. A pressure-bearing component safety assessment system based on nonlinear ultrasound, using the pressure-bearing component safety assessment method based on nonlinear ultrasound as described in any one of claims 1-9, characterized in that, include: The background nonlinear threshold calibration module includes: a nonlinear signal acquisition unit that controls the robotic arm to drive the automatic probe to acquire inherent nonlinear signals and additional nonlinear signals; and a background nonlinear threshold calibration unit that establishes a mapping reference library between clamping force and corresponding additional nonlinear signals to calibrate the background nonlinear threshold. The abnormal candidate region screening module includes: a field detection raw signal acquisition unit that divides the detection area of ​​the pressure-bearing component under test into a grid and acquires the field detection raw signal corresponding to each grid unit; an effective nonlinear parameter calculation unit that removes nonlinear interference for each field detection raw signal and calculates the effective nonlinear parameter; and an abnormal candidate region screening unit that filters abnormal candidate regions based on the effective nonlinear parameter and the background nonlinear threshold. Location feature calculation module: includes: an equivalent emission source parameter inversion unit that inverts and calculates the equivalent emission source parameters of the abnormal nonlinear signal for all anomaly candidate regions; and a location feature calculation unit that determines the anomaly location region based on the three-dimensional spatial coordinates of the equivalent emission source parameters and outputs the location features of the anomaly location region. Safety assessment module: includes: a safety assessment unit that uses artificial intelligence algorithms to train a safety assessment model, inputs location features into the safety assessment model, and calculates the safety assessment result of the pressure-bearing component.

Images