A stainless steel package steel weld corrosion life prediction method and system based on multi-parameter coupling

Abstract

The application discloses a stainless steel weld corrosion life prediction method and system based on multi-parameter coupling, and the method comprises the following steps: collecting corrosion potential sequence, electrochemical impedance spectrum data and electrochemical noise data of a weld; extracting charge transfer resistance and double-layer capacitance from the impedance spectrum data, extracting noise resistance and localization factor from the noise data, extracting corrosion potential drift rate from the corrosion potential sequence, and combining into a multi-dimensional state observation vector; constructing a state space model describing the evolution of the corrosion state of the weld, using a nonlinear Wiener process to describe corrosion damage accumulation, and establishing a mapping relationship between the observation vector and the corrosion state variable; using a particle filtering algorithm to recursively update the posterior probability distribution of the corrosion state; extrapolating and predicting the time when the corrosion state reaches a preset failure threshold according to the posterior distribution, and obtaining a residual service life prediction value and a confidence interval. The multi-dimensional electrochemical parameters are fused, and the accurate prediction and uncertainty quantification of the weld corrosion life are realized.

Description

Technical Field

[0001] This invention belongs to the field of weld corrosion life prediction technology, and particularly relates to a method and system for predicting the corrosion life of stainless steel cladding welds based on multi-parameter coupling. Background Technology

[0002] Stainless steel clad composite materials are made by explosive bonding or rolling bonding of a stainless steel cladding layer and a base steel layer, and are widely used in petrochemical, marine engineering, and power equipment industries. The stainless steel clad weld is the joint area formed when the composite plates are welded together. Due to the welding thermal cycle, the microstructure and composition distribution of the weld area change, and the electrochemical potential difference between the stainless steel cladding layer and the base steel is significant, making the weld a weak point for corrosion failure. During service, the weld is exposed to corrosive media, and corrosion gradually accumulates until failure occurs. Accurately predicting the remaining service life of the weld is crucial for ensuring safe equipment operation and developing preventative maintenance plans.

[0003] Currently, weld corrosion life prediction mainly employs single-parameter extrapolation methods. For example, based on the corrosion current density measured by the linear polarization resistance method, assuming a constant corrosion rate, the time to reach the allowable corrosion depth is linearly extrapolated. Alternatively, the development of the maximum pitting depth is predicted through extreme value distribution based on the statistical characteristics of electrochemical noise. These methods suffer from the following technical problems: First, a single parameter cannot fully reflect the complex evolution of the corrosion process. Weld corrosion often includes both uniform corrosion and localized pitting corrosion. Different corrosion morphologies contribute differently to structural failure, and a single parameter cannot distinguish the evolution of these two corrosion modes, leading to significant deviations in prediction results.

[0004] Second, traditional linear extrapolation methods cannot describe the nonlinear characteristic of corrosion rate accelerating with increasing corrosion intensity. In actual corrosion processes, the corrosion rate often exhibits an accelerating trend due to the accumulation of corrosion products, acidification of the local environment, and stress concentration. Linear models cannot capture this nonlinear acceleration behavior.

[0005] Third, existing methods lack quantitative processing of measurement uncertainties. Electrochemical measurements involve various uncertainties such as instrument noise, environmental fluctuations, and sample differences. Traditional methods only provide a single predicted value and cannot provide a prediction confidence interval, making it difficult to guide practical maintenance decisions.

[0006] Fourth, existing methods have failed to effectively integrate multi-source electrochemical information. Corrosion potential, electrochemical impedance spectroscopy, and electrochemical noise reflect the corrosion state from thermodynamic, kinetic, and statistical perspectives, respectively. Coupling these information can complement each other's advantages, but existing technologies lack an effective multi-parameter fusion framework. Summary of the Invention

[0007] The purpose of this invention is to provide a method and system for predicting the corrosion life of stainless steel cladding welds based on multi-parameter coupling, so as to solve the technical problems of low single-parameter prediction accuracy, inability to describe nonlinear accelerated corrosion, and lack of uncertainty quantification in the prior art.

[0008] In a first aspect, the present invention provides a method for predicting the corrosion life of stainless steel cladding welds based on multi-parameter coupling, comprising: The corrosion potential sequence, electrochemical impedance spectroscopy data, and electrochemical noise data of stainless steel cladding welds in corrosive media are collected simultaneously. The electrochemical impedance spectroscopy data includes the real and imaginary parts of impedance at multiple frequency points, and the electrochemical noise data includes the synchronously collected potential noise sequence and current noise sequence. Charge transfer resistance and double-layer capacitance are extracted from the electrochemical impedance spectroscopy data; noise resistance and localization factor are extracted from the electrochemical noise data; and corrosion potential drift rate is extracted from the corrosion potential sequence. The charge transfer resistance, the double-layer capacitance, the noise resistance, the localization factor, and the corrosion potential drift rate are combined into a multi-dimensional state observation vector after being aligned with time. A state-space model describing the evolution of weld corrosion state is constructed. The state-space model includes a state transition equation and an observation equation. The state transition equation uses a nonlinear Wiener process to describe the accumulation of corrosion damage. The observation equation establishes a mapping relationship between the multidimensional state observation vector and the corrosion state variables. The state-space model is recursively estimated using a particle filter algorithm, and the posterior probability distribution of the corrosion state variable is updated at each sampling time. Based on the posterior probability distribution of the corrosion state variable at the current moment, the time when the corrosion state variable reaches the preset failure threshold at a future moment is extrapolated and predicted. The difference between the time and the current time is used as the predicted value of the remaining service life of the weld.

[0009] Secondly, the present invention provides a corrosion life prediction system for stainless steel cladding welds based on multi-parameter coupling, comprising: The acquisition module is configured to simultaneously acquire corrosion potential sequence, electrochemical impedance spectroscopy data and electrochemical noise data of stainless steel cladding weld in corrosive media. The electrochemical impedance spectroscopy data includes the real part and imaginary part of impedance at multiple frequency points, and the electrochemical noise data includes the synchronously acquired potential noise sequence and current noise sequence. The extraction module is configured to extract charge transfer resistance and double-layer capacitance from the electrochemical impedance spectroscopy data, noise resistance and localization factor from the electrochemical noise data, and corrosion potential drift rate from the corrosion potential sequence. The combination module is configured to combine the charge transfer resistor, the double-layer capacitance, the noise resistor, the localization factor, and the corrosion potential drift rate into a multi-dimensional state observation vector after time alignment; The module is configured to construct a state-space model describing the evolution of weld corrosion state. The state-space model includes a state transition equation and an observation equation. The state transition equation uses a nonlinear Wiener process to describe the accumulation of corrosion damage. The observation equation establishes a mapping relationship between the multidimensional state observation vector and the corrosion state variables. The update module is configured to use a particle filter algorithm to perform recursive Bayesian estimation on the state space model and update the posterior probability distribution of the corrosion state variable at each sampling time. The prediction module is configured to extrapolate and predict the time when the corrosion state variable will reach a preset failure threshold in the future based on the posterior probability distribution of the corrosion state variable at the current time, and use the difference between the time and the current time as the predicted value of the remaining service life of the weld.

[0010] Thirdly, an electronic device is provided, comprising: at least one processor, and a memory communicatively connected to the at least one processor, wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the steps of the method for predicting the corrosion life of stainless steel cladding welds based on multi-parameter coupling according to any embodiment of the present invention.

[0011] Fourthly, the present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein when the program instructions are executed by a processor, the processor performs the steps of the method for predicting the corrosion life of stainless steel cladding welds based on multi-parameter coupling according to any embodiment of the present invention.

[0012] The corrosion life prediction method and system for stainless steel cladding welds based on multi-parameter coupling proposed in this application have the following beneficial effects: First, this invention simultaneously acquires three electrochemical signals—corrosion potential, electrochemical impedance spectroscopy, and electrochemical noise—and extracts five parameters from them: charge transfer resistance, double-layer capacitance, noise resistance, localization factor, and corrosion potential drift rate. These parameters are then used to construct a multi-dimensional state observation vector, enabling multi-faceted characterization of weld corrosion. Specifically, charge transfer resistance reflects the kinetic resistance of the corrosion reaction, double-layer capacitance reflects changes in the interface microstructure, noise resistance reflects the intensity of localized corrosion, the localization factor distinguishes between pitting and uniform corrosion, and the corrosion potential drift rate reflects the stability of the passivation film. These five parameters complement each other, comprehensively depicting the complex evolution of weld corrosion.

[0013] Second, this invention uses a nonlinear Wiener process to construct the state transition equation and introduces an exponential function into the drift term, so that the corrosion rate increases exponentially with the increase of the corrosion state variable. This can accurately describe the acceleration phenomenon commonly seen in the corrosion process and overcome the problem of large prediction deviation in traditional linear models.

[0014] Third, this invention employs a particle filtering algorithm for recursive Bayesian estimation. By approximating the posterior probability distribution through a particle ensemble, it can handle nonlinear non-Gaussian system models. Simultaneously, it provides point estimates and uncertainty measures for corrosion state variables, providing confidence intervals for subsequent lifetime prediction and enhancing the reliability of the prediction results.

[0015] Fourth, this invention prevents particle degradation through a regularized resampling method, ensuring the long-term stability of particle filtering. By extrapolating and predicting the failure time of each particle, the statistical distribution of the remaining lifetime is obtained, which can output the median predicted lifetime and the prediction confidence interval, making it easier for maintenance personnel to formulate maintenance plans based on their risk preferences. Attached Figure Description

[0016] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 A flowchart illustrating a method for predicting the corrosion life of stainless steel cladding welds based on multi-parameter coupling, as provided in an embodiment of the present invention; Figure 2 A structural block diagram of a stainless steel cladding weld corrosion life prediction system based on multi-parameter coupling, provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0018] 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.

[0019] Please see Figure 1 The diagram shows a flowchart of a method for predicting the corrosion life of stainless steel cladding welds based on multi-parameter coupling, as proposed in this application.

[0020] like Figure 1 As shown, the method for predicting the corrosion life of stainless steel cladding welds based on multi-parameter coupling specifically includes the following steps: Step S101: Simultaneously collect corrosion potential sequence, electrochemical impedance spectroscopy data and electrochemical noise data of stainless steel cladding weld in corrosive medium. The electrochemical impedance spectroscopy data includes the real part and imaginary part of impedance at multiple frequency points, and the electrochemical noise data includes the synchronously collected potential noise sequence and current noise sequence.

[0021] In this step, the stainless steel-clad weld sample is immersed in a simulated seawater corrosion medium (a 3.5% sodium chloride solution by mass) at a temperature controlled at 25 degrees Celsius. An electrochemical workstation is connected, employing a three-electrode system: the working electrode is the stainless steel-clad weld sample, the reference electrode is a saturated calomel electrode, and the auxiliary electrode is a platinum electrode.

[0022] The following three types of signals were simultaneously acquired: First, corrosion potential sequence, sampled at a frequency of 1 Hz for 24 consecutive hours to obtain the sequence of corrosion potential changes over time. Second, electrochemical impedance spectroscopy data, with a frequency scan range of 100 kHz to 0.01 Hz, using a logarithmically evenly spaced frequency scan. A sinusoidal AC voltage signal with an amplitude of 10 mV was applied at each frequency point, and the amplitude and phase of the response current signal were measured to calculate the real and imaginary parts of the impedance. Third, electrochemical noise data, with potential noise and current noise simultaneously acquired under zero applied current and zero applied voltage conditions, sampled at a frequency of 10 Hz, with 1024 points collected each time, and the potential noise and current noise sequences strictly time-aligned.

[0023] Step S102: Extract charge transfer resistance and double layer capacitance from the electrochemical impedance spectroscopy data, extract noise resistance and localization factor from the electrochemical noise data, and extract corrosion potential drift rate from the corrosion potential sequence.

[0024] In this step, the electrochemical impedance spectroscopy data is converted into relaxation time distribution spectral lines. The highest frequency peak and the lowest frequency peak are identified from the relaxation time distribution spectral lines. The integral area corresponding to the lowest frequency peak is used as an estimate of the charge transfer resistance, and the ratio of the relaxation time peak value corresponding to the highest frequency peak to the charge transfer resistance is used as an estimate of the double-layer capacitance.

[0025] Furthermore, power spectral density is calculated for the potential noise sequence and the current noise sequence respectively. The square root of the ratio of the plateau values ​​of the power spectral density in the low-frequency band is taken as the noise resistance. The ratio of the standard deviation of the current noise sequence to the standard deviation of the potential noise sequence is taken as the localization factor. When the localization factor is greater than 0.1, it indicates pitting corrosion initiation. When the localization factor is less than 0.01, it indicates uniform corrosion.

[0026] The corrosion potential sequence is filtered by moving average to obtain a corrosion potential trend curve. The corrosion potential trend curve is then fitted linearly in segments. The slope of the fitted straight line in each segment is taken as the corrosion potential drift rate in that segment. A positive corrosion potential drift rate indicates that the passivation film has been repaired, while a negative value indicates that the passivation film has ruptured.

[0027] Step S103: The charge transfer resistor, the double-layer capacitance, the noise resistor, the localization factor, and the corrosion potential drift rate are combined into a multi-dimensional state observation vector after being aligned with time.

[0028] In this step, firstly, charge transfer resistance and double-layer capacitance are extracted from the electrochemical impedance spectroscopy (EIS) data. Specifically, the EIS data is converted into relaxation time distribution spectra. The conversion process includes: constructing a Fredholm first-kind integral equation with relaxation time as the independent variable, solving it using the Tikhonov regularization method to obtain the relaxation time distribution spectra. The highest and lowest frequency peaks are identified from these relaxation time distribution spectra. The lowest frequency peak corresponds to the longest relaxation time, reflecting the charge transfer process of the electrochemical reaction; the integral area of ​​this peak is used as an estimate of the charge transfer resistance. The highest frequency peak corresponds to the shortest relaxation time, reflecting the charging and discharging process of the double layer; the peak relaxation time of this peak is multiplied by the charge transfer resistance to obtain an estimate of the double-layer capacitance.

[0029] Secondly, noise resistance and localization factor are extracted from the electrochemical noise data. Specifically, power spectral density is calculated for both the potential noise sequence and the current noise sequence. The square root of the ratio of the plateau values ​​of the power spectral density in the low-frequency range of 0.01 Hz to 1 Hz is taken as the noise resistance. The larger the noise resistance value, the stronger the weld's resistance to uniform corrosion. The localization factor is equal to the standard deviation of the current noise sequence divided by the standard deviation of the potential noise sequence. When the localization factor is greater than 0.1, it indicates pitting corrosion initiation; when the localization factor is less than 0.01, it indicates that uniform corrosion is dominant; and when it is between the two, it indicates a mixed corrosion morphology.

[0030] Next, the corrosion potential drift rate is extracted from the corrosion potential sequence. Specifically, a five-point moving average filter is applied to the corrosion potential sequence to remove high-frequency noise, resulting in a corrosion potential trend curve. A piecewise linear fitting method is used to divide the trend curve into multiple linear segments, with the endpoints of each segment determined by an inflection point detection algorithm. The slope of the fitted straight line within each segment represents the corrosion potential drift rate for that time period. A positive drift rate indicates that the passivation film is being repaired, and the corrosion tendency is reduced; a negative drift rate indicates that the passivation film is ruptured, and the corrosion tendency is increased.

[0031] The five parameters extracted above are aligned according to the sampling time, and a five-dimensional vector is obtained at each sampling time, which is denoted as the multidimensional state observation vector.

[0032] Step S104: Construct a state-space model describing the evolution of weld corrosion state. The state-space model includes a state transition equation and an observation equation. The state transition equation uses a nonlinear Wiener process to describe the accumulation of corrosion damage. The observation equation establishes a mapping relationship between the multidimensional state observation vector and the corrosion state variables.

[0033] In this step, the corrosion state variable at the current moment is equal to the corrosion state variable at the previous moment plus the sum of the drift term and the diffusion term. The drift term contains an exponential function, the exponent of which is the product of the drift coefficient and the corrosion state variable at the previous moment. The exponential function is used to characterize the nonlinear characteristic that the corrosion rate increases rapidly with the deepening of corrosion. The diffusion term is the diffusion coefficient multiplied by the square root of the sampling time interval and then multiplied by a standard normally distributed random number.

[0034] Multidimensional state observation vector samples at different corrosion stages are obtained through accelerated corrosion tests. Simultaneously, the true corrosion depth corresponding to the multidimensional state observation vector samples is obtained as the true value of the corrosion state variable through destructive measurement. A nonlinear observation function is obtained by fitting the nonlinear observation function using a multivariate nonlinear regression method. The nonlinear observation function takes the corrosion state variable as input and the predicted value of each component in the multidimensional state observation vector as output. The observation equation represents the multidimensional state observation vector at the current moment as the value of the nonlinear observation function at the corrosion state variable at the current moment plus the observation noise vector.

[0035] Step S105: The particle filter algorithm is used to perform recursive Bayesian estimation on the state space model, and the posterior probability distribution of the corrosion state variable is updated at each sampling time.

[0036] In this step, the initialization step extracts a specified number of particles from the prior distribution to form an initial particle set, with each particle carrying the same initial weight. In the prediction step, each particle is propagated forward one step according to the state transition equation, evolving the corrosion state variable of each particle to the next time step according to a nonlinear Wiener process, resulting in a predicted particle set. In the update step, the multidimensional state observation vector at the current time step is obtained, and the likelihood weight of each predicted particle is calculated according to the observation equation. The likelihood weight is equal to the value of the observation noise probability density function at the difference between the current observation and the predicted observation. The likelihood weight of each predicted particle is multiplied by the weight at the previous time step, and the weights of all particles are normalized. In the resampling step, the predicted particles are sampled and replaced according to the normalized weights; particles with large weights are copied, and particles with small weights are eliminated, resulting in a posterior particle set. The weighted average of the posterior particle set is used as the point estimate of the corrosion state variable, and the weighted covariance matrix of the posterior particle set is used as the estimation uncertainty measure.

[0037] It should be noted that the resampling step adopts a regularized resampling method, which adds a Gaussian perturbation to the kernel density estimation bandwidth during the particle replication process. The kernel density bandwidth of the regularized resampling is calculated according to the functional relationship between the negative exponent of the number of particles and the dimension of the state space. The kernel density bandwidth is used to prevent particle degradation while ensuring particle diversity.

[0038] Step S106: Based on the posterior probability distribution of the corrosion state variable at the current moment, extrapolate and predict the time when the corrosion state variable will reach the preset failure threshold at a future moment, and use the difference between the time and the current time as the predicted value of the remaining service life of the weld.

[0039] In this step, starting from the posterior particle set of the corrosion state variable at the current moment, the corrosion state evolution trajectory is simulated independently for each particle according to the state transition equation time by time. For each particle, the prediction time step corresponding to the first time the corrosion state variable of the particle reaches the preset failure threshold is recorded. The arithmetic mean of the failure time steps of all particles is taken and multiplied by the sampling time interval to obtain the remaining service life prediction value. At the same time, the 50th percentile of the failure time steps of all particles is taken as the median predicted service life, and the 16th percentile and 84th percentile are taken as the lower limit and upper limit of the 68% prediction confidence interval. Please see Figure 2 The diagram shows a structural block diagram of a stainless steel cladding weld corrosion life prediction system based on multi-parameter coupling, as proposed in this application.

[0040] like Figure 2As shown, the stainless steel cladding weld corrosion life prediction system 200 includes an acquisition module 210, an extraction module 220, a combination module 230, a construction module 240, an update module 250, and a prediction module 260.

[0041] The acquisition module 210 is configured to simultaneously acquire corrosion potential sequences, electrochemical impedance spectroscopy (EIS) data, and electrochemical noise data of stainless steel cladding welds in a corrosive medium. The EIS data includes real and imaginary impedance parts at multiple frequency points, and the EIS data includes synchronously acquired potential noise sequences and current noise sequences. The extraction module 220 is configured to extract charge transfer resistance and double-layer capacitance from the EIS data, noise resistance and localization factor from the EIS data, and corrosion potential drift rate from the corrosion potential sequence. The combination module 230 is configured to combine the charge transfer resistance, double-layer capacitance, noise resistance, localization factor, and corrosion potential drift rate into a multidimensional state view after time alignment. The system includes a measurement vector; a construction module 240, configured to construct a state-space model describing the evolution of weld corrosion state, the state-space model including a state transition equation and an observation equation, the state transition equation using a nonlinear Wiener process to describe corrosion damage accumulation, and the observation equation establishing a mapping relationship between the multidimensional state observation vector and the corrosion state variables; an update module 250, configured to use a particle filter algorithm to perform recursive Bayesian estimation on the state-space model, updating the posterior probability distribution of the corrosion state variables at each sampling time; and a prediction module 260, configured to extrapolate and predict the time when the corrosion state variables reach a preset failure threshold in the future based on the posterior probability distribution of the corrosion state variables at the current time, and using the difference between the time and the current time as the predicted value of the remaining service life of the weld.

[0042] It should be understood that Figure 2 The modules and references described in the document Figure 1 The steps described in the text correspond to those in the method described above. Therefore, the operations, features, and corresponding technical effects described above also apply to the method described in the text. Figure 2 The various modules in the document will not be described in detail here.

[0043] In other embodiments, the present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein when the program instructions are executed by a processor, the processor performs the method for predicting the corrosion life of stainless steel cladding welds based on multi-parameter coupling in any of the above method embodiments. In one embodiment, the computer-readable storage medium of the present invention stores computer-executable instructions, which are configured as follows: The corrosion potential sequence, electrochemical impedance spectroscopy data, and electrochemical noise data of stainless steel cladding welds in corrosive media are collected simultaneously. The electrochemical impedance spectroscopy data includes the real and imaginary parts of impedance at multiple frequency points, and the electrochemical noise data includes the synchronously collected potential noise sequence and current noise sequence. Charge transfer resistance and double-layer capacitance are extracted from the electrochemical impedance spectroscopy data; noise resistance and localization factor are extracted from the electrochemical noise data; and corrosion potential drift rate is extracted from the corrosion potential sequence. The charge transfer resistance, the double-layer capacitance, the noise resistance, the localization factor, and the corrosion potential drift rate are combined into a multi-dimensional state observation vector after being aligned with time. A state-space model describing the evolution of weld corrosion state is constructed. The state-space model includes a state transition equation and an observation equation. The state transition equation uses a nonlinear Wiener process to describe the accumulation of corrosion damage. The observation equation establishes a mapping relationship between the multidimensional state observation vector and the corrosion state variables. The state-space model is recursively estimated using a particle filter algorithm, and the posterior probability distribution of the corrosion state variable is updated at each sampling time. Based on the posterior probability distribution of the corrosion state variable at the current moment, the time when the corrosion state variable reaches the preset failure threshold at a future moment is extrapolated and predicted. The difference between the time and the current time is used as the predicted value of the remaining service life of the weld.

[0044] Computer-readable storage media may include a stored program area and a stored data area, wherein the stored program area may store an operating system and an application program required for at least one function; the stored data area may store data created based on the use of the stainless steel cladding weld corrosion life prediction system based on multi-parameter coupling. Furthermore, the computer-readable storage medium may include high-speed random access memory, and may also include memory, such as at least one disk storage device, flash memory device, or other non-volatile solid-state storage device. In some embodiments, the computer-readable storage medium may optionally include memory remotely configured relative to a processor, which can be connected to the stainless steel cladding weld corrosion life prediction system based on multi-parameter coupling via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0045] Figure 3 This is a schematic diagram of the structure of the electronic device provided in the embodiment of the present invention, such as... Figure 3As shown, the device includes a processor 310 and a memory 320. The electronic device may also include an input device 330 and an output device 340. The processor 310, memory 320, input device 330, and output device 340 can be connected via a bus or other means. Figure 3 Taking a bus connection as an example, the memory 320 is the computer-readable storage medium described above. The processor 310 executes various server functions and data processing by running non-volatile software programs, instructions, and modules stored in the memory 320, thereby realizing the stainless steel cladding weld corrosion life prediction method based on multi-parameter coupling described in the above embodiment. The input device 330 can receive input digital or character information and generate key signal inputs related to user settings and function control of the stainless steel cladding weld corrosion life prediction system based on multi-parameter coupling. The output device 340 may include a display screen or other display device.

[0046] The aforementioned electronic device can execute the method provided in the embodiments of the present invention, and has the corresponding functional modules and beneficial effects for executing the method. Technical details not described in detail in this embodiment can be found in the method provided in the embodiments of the present invention.

[0047] In one implementation, the above-described electronic device is applied to a stainless steel cladding weld corrosion life prediction system based on multi-parameter coupling, for a client, and includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to: The corrosion potential sequence, electrochemical impedance spectroscopy data, and electrochemical noise data of stainless steel cladding welds in corrosive media are collected simultaneously. The electrochemical impedance spectroscopy data includes the real and imaginary parts of impedance at multiple frequency points, and the electrochemical noise data includes the synchronously collected potential noise sequence and current noise sequence. Charge transfer resistance and double-layer capacitance are extracted from the electrochemical impedance spectroscopy data; noise resistance and localization factor are extracted from the electrochemical noise data; and corrosion potential drift rate is extracted from the corrosion potential sequence. The charge transfer resistance, the double-layer capacitance, the noise resistance, the localization factor, and the corrosion potential drift rate are combined into a multi-dimensional state observation vector after being aligned with time. A state-space model describing the evolution of weld corrosion state is constructed. The state-space model includes a state transition equation and an observation equation. The state transition equation uses a nonlinear Wiener process to describe the accumulation of corrosion damage. The observation equation establishes a mapping relationship between the multidimensional state observation vector and the corrosion state variables. The state-space model is recursively estimated using a particle filter algorithm, and the posterior probability distribution of the corrosion state variable is updated at each sampling time. Based on the posterior probability distribution of the corrosion state variable at the current moment, the time when the corrosion state variable reaches the preset failure threshold at a future moment is extrapolated and predicted. The difference between the time and the current time is used as the predicted value of the remaining service life of the weld.

[0048] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of various embodiments or some parts of embodiments.

[0049] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for predicting the corrosion life of stainless steel cladding welds based on multi-parameter coupling, characterized in that, include: The corrosion potential sequence, electrochemical impedance spectroscopy data, and electrochemical noise data of stainless steel cladding welds in corrosive media are collected simultaneously. The electrochemical impedance spectroscopy data includes the real and imaginary parts of impedance at multiple frequency points, and the electrochemical noise data includes the synchronously collected potential noise sequence and current noise sequence. Charge transfer resistance and double-layer capacitance are extracted from the electrochemical impedance spectroscopy data; noise resistance and localization factor are extracted from the electrochemical noise data; and corrosion potential drift rate is extracted from the corrosion potential sequence. The charge transfer resistance, the double-layer capacitance, the noise resistance, the localization factor, and the corrosion potential drift rate are combined into a multi-dimensional state observation vector after being aligned with time. A state-space model describing the evolution of weld corrosion state is constructed. The state-space model includes a state transition equation and an observation equation. The state transition equation uses a nonlinear Wiener process to describe the accumulation of corrosion damage. The observation equation establishes a mapping relationship between the multidimensional state observation vector and the corrosion state variables. The state-space model is recursively estimated using a particle filter algorithm, and the posterior probability distribution of the corrosion state variable is updated at each sampling time. Based on the posterior probability distribution of the corrosion state variable at the current moment, the time when the corrosion state variable reaches the preset failure threshold at a future moment is extrapolated and predicted. The difference between the time and the current time is used as the predicted value of the remaining service life of the weld.

2. The method for predicting the corrosion life of stainless steel cladding welds based on multi-parameter coupling according to claim 1, characterized in that, The specific steps for extracting the charge transfer resistance and the electric double layer capacitance from the electrochemical impedance spectroscopy data include: converting the electrochemical impedance spectroscopy data into relaxation time distribution spectral lines, identifying the highest frequency peak and the lowest frequency peak from the relaxation time distribution spectral lines, using the integral area corresponding to the lowest frequency peak as an estimated value of the charge transfer resistance, and using the ratio of the relaxation time peak value corresponding to the highest frequency peak to the charge transfer resistance as an estimated value of the electric double layer capacitance.

3. The method for predicting the corrosion life of stainless steel cladding welds based on multi-parameter coupling according to claim 1, characterized in that, The specific steps for extracting the noise resistance and the localization factor from the electrochemical noise data include: calculating the power spectral density of the potential noise sequence and the current noise sequence respectively; taking the square root of the ratio of the plateau values ​​of the power spectral density in the low-frequency band as the noise resistance; and taking the ratio of the standard deviation of the current noise sequence to the standard deviation of the potential noise sequence as the localization factor. A localization factor greater than 0.1 indicates pitting corrosion initiation, and a localization factor less than 0.01 indicates uniform corrosion.

4. The method for predicting the corrosion life of stainless steel cladding welds based on multi-parameter coupling according to claim 1, characterized in that, The specific steps for extracting the corrosion potential drift rate from the corrosion potential sequence include: performing a moving average filter on the corrosion potential sequence to obtain a corrosion potential trend curve; performing linear segmented fitting on the corrosion potential trend curve; taking the slope of the fitted straight line in each segment as the corrosion potential drift rate in that segment; a positive corrosion potential drift rate indicates that the passivation film has been repaired, and a negative value indicates that the passivation film has ruptured.

5. The method for predicting the corrosion life of stainless steel cladding welds based on multi-parameter coupling according to claim 1, characterized in that, The state transition equation, described by a nonlinear Wiener process, is as follows: the corrosion state variable at the current moment is equal to the corrosion state variable at the previous moment plus the sum of the drift term and the diffusion term. The drift term contains an exponential function, the exponent of which is the product of the drift coefficient and the corrosion state variable at the previous moment. The exponential function is used to characterize the nonlinear characteristic that the corrosion rate increases rapidly with the deepening of corrosion. The diffusion term is the diffusion coefficient multiplied by the square root of the sampling time interval and then multiplied by a standard normally distributed random number.

6. The method for predicting the corrosion life of stainless steel cladding welds based on multi-parameter coupling according to claim 1, characterized in that, The specific construction method of the observation equation is as follows: multidimensional state observation vector samples at different corrosion stages are obtained through accelerated corrosion tests, and the true corrosion depth corresponding to the multidimensional state observation vector samples is obtained as the true value of the corrosion state variable through destructive measurement. A nonlinear observation function is obtained by fitting the nonlinear observation function using a multivariate nonlinear regression method. The nonlinear observation function takes the corrosion state variable as input and the predicted value of each component in the multidimensional state observation vector as output. The observation equation represents the multidimensional state observation vector at the current moment as the value of the nonlinear observation function at the corrosion state variable at the current moment plus the observation noise vector.

7. The method for predicting the corrosion life of stainless steel cladding welds based on multi-parameter coupling according to claim 1, characterized in that, The specific steps of recursive Bayesian estimation using the particle filter algorithm include: In the initialization step, a specified number of particles are drawn from the prior distribution to form an initial particle set, with each particle carrying the same initial weight; in the prediction step, each particle is propagated forward one step according to the state transition equation, and the corrosion state variable of each particle evolves to the next time step according to a nonlinear Wiener process, resulting in a predicted particle set; in the update step, the multidimensional state observation vector at the current time step is obtained, and the likelihood weight of each predicted particle is calculated according to the observation equation. The likelihood weight is equal to the value of the observation noise probability density function at the difference between the current observation value and the predicted observation value. The likelihood weight of each predicted particle is multiplied by the weight at the previous time step, and the weights of all particles are normalized; in the resampling step, the predicted particles are sampled and replaced according to the normalized weights, with particles with large weights being copied and particles with small weights being eliminated, resulting in a posterior particle set; the weighted average of the posterior particle set is used as the point estimate of the corrosion state variable, and the weighted covariance matrix of the posterior particle set is used as the estimation uncertainty measure.

8. The method for predicting the corrosion life of stainless steel cladding welds based on multi-parameter coupling according to claim 7, characterized in that, The resampling step employs a regularized resampling method, incorporating a Gaussian perturbation of the kernel density estimation bandwidth during particle replication. The kernel density bandwidth of the regularized resampling is calculated as a function of the negative exponent of the number of particles and the dimension of the state space.

9. The method for predicting the corrosion life of stainless steel cladding welds based on multi-parameter coupling according to claim 1, characterized in that, The specific steps for extrapolating and predicting the remaining service life of the weld include: starting from the posterior particle set of the corrosion state variable at the current moment, independently simulating the corrosion state evolution trajectory of each particle step by step according to the state transition equation; recording the prediction time step corresponding to when the corrosion state variable of each particle first reaches the preset failure threshold; taking the arithmetic mean of the failure time steps of all particles and multiplying it by the sampling time interval to obtain the predicted value of the remaining service life; taking the 50th percentile of the failure time steps of all particles as the median predicted service life; and taking the 16th percentile and 84th percentile as the lower and upper limits of the 68% prediction confidence interval.

10. A corrosion life prediction system for stainless steel cladding welds based on multi-parameter coupling, characterized in that, include: The acquisition module is configured to simultaneously acquire corrosion potential sequence, electrochemical impedance spectroscopy data and electrochemical noise data of stainless steel cladding weld in corrosive media. The electrochemical impedance spectroscopy data includes the real part and imaginary part of impedance at multiple frequency points, and the electrochemical noise data includes the synchronously acquired potential noise sequence and current noise sequence. The extraction module is configured to extract charge transfer resistance and double-layer capacitance from the electrochemical impedance spectroscopy data, noise resistance and localization factor from the electrochemical noise data, and corrosion potential drift rate from the corrosion potential sequence. The combination module is configured to combine the charge transfer resistor, the double-layer capacitance, the noise resistor, the localization factor, and the corrosion potential drift rate into a multi-dimensional state observation vector after time alignment; The module is configured to construct a state-space model describing the evolution of weld corrosion state. The state-space model includes a state transition equation and an observation equation. The state transition equation uses a nonlinear Wiener process to describe the accumulation of corrosion damage. The observation equation establishes a mapping relationship between the multidimensional state observation vector and the corrosion state variables. The update module is configured to use a particle filter algorithm to perform recursive Bayesian estimation on the state space model and update the posterior probability distribution of the corrosion state variable at each sampling time. The prediction module is configured to extrapolate and predict the time when the corrosion state variable will reach a preset failure threshold in the future based on the posterior probability distribution of the corrosion state variable at the current time, and use the difference between the time and the current time as the predicted value of the remaining service life of the weld.

Images