Cage guide weak defect detection method, inspection system, storage medium and equipment

Abstract

The invention discloses a shaft guide weak defect detection method, an inspection system, a storage medium and equipment, and belongs to the field of safety monitoring of mine deep well hoisting systems. According to the method, a second-order under-damping three-stable-state stochastic resonance system model based on variable-scale frequency shift is constructed aiming at the physical problem that a high-frequency transient echo signal generated by active excitation is difficult to meet a stochastic resonance adiabatic approximation condition; performing time domain compression and frequency domain re-scaling on an original signal by introducing a frequency scale transformation factor, and transforming a high-frequency impact characteristic to a low-frequency resonance region; meanwhile, a second-order inertia term and nonlinear damping are introduced, and a dynamic system with band-pass filtering characteristics is constructed; an improved adaptive particle swarm optimization algorithm is adopted to synchronously optimize scale factors, damping coefficients and potential well parameters, and broadband noise energy is focused on defect characteristic frequency by using a variable-scale stochastic resonance effect, so that high-sensitivity detection of early defects such as microcracks and bolt looseness is realized under an extremely low signal-to-noise ratio.

Description

Technical Field

[0001] This invention relates to the field of safety monitoring technology for deep mine hoisting systems, and in particular to a method for detecting minor defects in mine shafts, an inspection system, a storage medium, and equipment. Background Technology

[0002] With the increasing intensity of deep resource extraction, mine hoists are developing towards high speed, heavy load, and deep shafts. Rigid guideways, as the "lifeline" of the hoisting system, are highly susceptible to early defects such as micro-cracks, loose bolts, and abnormal gaps. Existing detection methods have significant shortcomings: The underground environment in coal mines is harsh, and the operation of the cage is accompanied by strong aerodynamic noise, wire rope vibration, and mechanical friction (strong background noise). Traditional passive detection or simple active tapping methods produce early weak defects (such as 0.1mm microcracks or loose deep bolts) with extremely low echo signal energy, which is easily drowned out by background noise (low signal-to-noise ratio).

[0003] Existing linear filtering denoising techniques (such as wavelet transform and low-pass filtering) often filter out weak defect feature signals that overlap with the noise frequency band while filtering out environmental noise, resulting in "denoising as distortion" and causing early hidden dangers to be missed.

[0004] Therefore, there is an urgent need to provide a method and system for detecting minor defects in tank passages to solve the above problems. Summary of the Invention

[0005] The main objective of this invention is to provide a method, inspection system, storage medium, and equipment for detecting minor defects in tank passageways, thereby addressing existing technical problems.

[0006] To achieve the above objectives, the present invention provides a method for detecting minor defects in tank passageways, comprising: S1: Active transient impact is performed on the tank passage to collect the original composite signal on the surface of the tank passage, including weak defect feature signals and strong downhole background noise; S2: Receive the original composite signal and construct a second-order underdamped tristable stochastic resonance system model based on variable-scale frequency shift. The model introduces a second-order inertial term and a frequency scale transformation mechanism to match the high-frequency transient echo signal generated by active knocking. S3: Based on the characteristics of the currently acquired signal, the system parameters and damping factor of the second-order underdamped model are adaptively adjusted using an optimization algorithm. The original composite signal is then input into the second-order model after parameter optimization. The background noise energy and the defect impact signal are made to resonate in synergy using the inertial stochastic resonance mechanism, thereby outputting a signal with a significantly enhanced signal-to-noise ratio. S4: Extract defect features from the enhanced signal and combine them with the location information to determine the existence and location of the defect.

[0007] Furthermore, the dynamic equation of the variable-scale second-order tristable stochastic resonance system in step S2 is:

[0008] in, For the transformed time scale, satisfying , This is the frequency scaling factor; For system output; For nonlinear damping terms, Based on the basic damping coefficient, It is a nonlinear control factor; The structural parameters of the tristable state trap; and These are the defect feature signal and background noise after time-scale compression, respectively.

[0009] Furthermore, the optimization algorithm is an adaptive particle swarm optimization algorithm, which uses the signal-to-noise ratio of the output signal of the tristable stochastic resonance system as the fitness function, and iteratively searches for the system parameter combination that maximizes the signal-to-noise ratio.

[0010] Furthermore, the execution of the adaptive particle swarm optimization algorithm includes the following sub-steps: S3.1: Set the particle swarm size M and the maximum number of iterations T, and randomly initialize the position vector Xi and velocity vector Vi for each particle. The position vector represents a set of system control parameters. ,in For frequency scaling factor, For damping parameters, These are the potential well parameters; S3.2: For each particle, substitute the parameters it represents into the variable-scale second-order tristable stochastic resonance system, and introduce a time-scale transformation. Input the original mixed signal and calculate the output signal. The weighted kurtosis multiplied by the local signal-to-noise ratio is used as the fitness value of the particle; S3.3: Record the best historical position for each particle. and the historical best position of the entire particle swarm ; S3.4: Adjust the inertia weight w(t), individual learning factor c1(t), and social learning factor c2(t) nonlinearly according to the current iteration number t; S3.5: Update the velocity and position of each particle according to the formula; ; Where r1 and r2 are random numbers in the interval [0,1]; S3.6: Iteration and Output: Repeat steps S3.2 to S3.5 until the termination condition is met, and output the global optimal position. The corresponding parameters are used as the optimal parameter combination of the second-order three-stable state system.

[0011] Furthermore, in step S3.4, the inertia weight w(t) adopts a non-linear decreasing strategy, decreasing from the initial value to the final value as the number of iterations increases; and the learning factors satisfy: in the initial stage of iteration, c1(t) > c2(t) to enhance the global search ability; in the later stage of iteration, c1(t) < c2(t) to accelerate convergence.

[0012] Furthermore, in step S5, multi-source information fusion determination is performed. When there are characteristic peaks exceeding the threshold in the enhanced vibration signal, and the sound sensor synchronously detects abnormal audio spectrum characteristics, and / or the vision sensor captures abnormal surface texture of the shaft guide, it is comprehensively determined as a confirmed defect.

[0013] A shaft guide weak defect inspection system uses the above detection method for inspection. The inspection system includes: An inspection trolley unit for moving along the shaft guide for inspection. An active excitation unit for performing active transient knocking on the shaft guide. A positioning unit for measuring the distance of the inspection trolley unit during inspection movement.

[0014] On the other hand, the present invention also discloses a computer-readable storage medium for a shaft guide weak defect detection method, storing a computer program. When the computer program is executed by a processor, the processor is caused to execute the steps of the above method.

[0015] On yet another hand, the present invention also discloses a computer device for a shaft guide weak defect detection method, including a memory and a processor. The memory stores a computer program. When the computer program is executed by the processor, the processor is caused to execute the steps of the above method.

[0016] The beneficial effects of the present invention are reflected in: The present invention breaks through the frequency bottleneck in the application of the stochastic resonance theory and realizes the effective detection of high-frequency engineering signals. The traditional stochastic resonance theory is strictly limited by the "adiabatic approximation condition" (the input frequency must be much less than 1 Hz), resulting in its inability to be applied to the field of high-frequency vibration detection for a long time. The present invention innovatively introduces a frequency scale transformation mechanism, and finds the best "time compression ratio" through an optimization algorithm, mapping the high-frequency echo in the hundreds of hertz generated by the active knocking on the shaft guide to the low-frequency resonance region at the mathematical processing level. This improvement fundamentally solves the physical problem that high-frequency engineering signals cannot drive the stochastic resonance system, providing a new theoretical path for the high-frequency transient non-destructive detection of deep shaft guides.

[0017] The displacement-dependent nonlinear damping system constructed in this invention has the characteristic of "intelligent variable damping": During the "latency period" of weak signal input, the system maintains extremely low damping and extremely high sensitivity to micron-level crack impacts, utilizing noise energy to instantly break through the potential barrier. During the "outburst period" triggered by resonance, the damping increases rapidly with the amplitude, forming an "adaptive braking" effect. This mechanism effectively prevents false alarms caused by non-stationary aerodynamic noise downhole, enabling the output of clear and stable defect characteristic waveforms even in extreme environments with a signal-to-noise ratio as low as -30dB.

[0018] This invention utilizes the bandpass filtering characteristics of a second-order inertial system to effectively suppress low-frequency geological interference unique to deep well environments. Addressing the extremely low-frequency baseline drift caused by wellbore deformation and geological pressure in deep mines, this invention abandons the traditional first-order overdamped model and introduces a second-order inertial term to endow the system with "mass memory." By leveraging the inherent bandpass filtering characteristics of the second-order system, it naturally blocks low-frequency geological drift and background sway below 5Hz, resonating and amplifying only the transient impact characteristics excited by impacts, significantly reducing the false alarm rate caused by environmental baseline fluctuations. Attached Figure Description

[0019] Figure 1 This is a schematic diagram of the method for detecting minor defects in the tank passage of the present invention; Figure 2 This is an ideal characteristic signal waveform diagram of the weak defects in the can passage of the present invention; Figure 3 This is a waveform diagram of the original detection signal that was submerged by strong noise under the actual working conditions of the present invention. Figure 4 This is a waveform diagram of the enhanced output signal of the adaptive tristable stochastic resonance system of the present invention; Figure 5 This is a schematic diagram of the structure of the tank passage weak defect detection system of the present invention; Figure 6 This is a schematic diagram of the active excitation unit structure of the present invention; Figure 7 This is a schematic diagram of the positioning unit structure of the present invention.

[0020] Explanation of reference numerals in the attached figures: 1. Inspection trolley unit; 109; 2. Active excitation unit; 21. Controllable vibrator; 22. Fixing buckle; 203. Tapping head; 3. Positioning unit; 31. Positioning wheel; 32. Positioning wheel connecting rod; 33. Permanent magnet. Detailed Implementation

[0021] 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 a part of the embodiments of the present invention, and not all of them. Unless otherwise specified, the embodiments and features in the embodiments of this application can be combined with each other. 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.

[0022] Please see Figure 1 This invention provides a method for detecting minor defects in well passageways. The method relies on a hardware system consisting of an inspection trolley unit 1, an active excitation unit 2, and a positioning unit 3. Addressing the high-frequency aerodynamic noise and mechanical friction interference in deep well hoisting environments, it constructs a hardware-software co-operated "variable-scale second-order nonlinear dynamics detection system" to accurately capture minor defects in well passageways. The method includes the following steps; S1: Control the active transient impact on the tank passage to collect the original composite signal on the surface of the tank passage, including weak defect feature signals and strong downhole background noise; Specifically, the original composite signal on the surface of the tank channel is collected by a vibration sensor; the active excitation unit is controlled to perform active transient impact on the tank channel to collect the original composite signal on the surface of the tank channel.

[0023] Inspection trolley unit 1 slides down the tank track at a constant speed (e.g., velocity) At this point, the active excitation unit 2 installed on the trolley begins to operate. Specifically, the controllable vibrator 21, under control commands, drives the striking head 23 to apply periodic transient impacts (the impact frequency is set to...) to the side surface of the tank passage. To ensure the stability of the impact, the retaining clip 22 securely locks the vibrator to the preset mounting plane, and during installation, the perpendicularity error between the impact head 23 and the can channel surface is calibrated. This prevents the excitation energy from being dispersed due to the shaking of the trolley.

[0024] Downhole signal feature acquisition: A high-frequency piezoelectric vibration sensor installed close to the impact point synchronously acquires the original composite signal from the surface of the tank passage. The signal contains two distinct components: Weak defect characteristic signals Induced by active excitation, the amplitude is extremely small. If there is a microcrack or loose bolt near the impact point, the impact energy will excite a set of high-frequency decaying oscillation waves (the center frequency is usually around 1000 Hz), but due to the small size of the defect, the amplitude is extremely low.

[0025] Strong background noise This includes aerodynamic noise (broadband) generated by the high-speed air cutting of the cage, rolling friction noise generated by the contact between the positioning wheel 31 and the cage guide, and extremely low-frequency baseline drift caused by the geological pressure in the wellbore. At this time, the signal-to-noise ratio is usually below -25dB.

[0026] In existing technologies, such as Figure 3 As shown, in the original signal, the defect features are completely submerged in the messy noise and cannot be identified by the naked eye or by traditional linear filters.

[0027] S2: Receive the original composite signal and construct a variable-scale second-order tristable stochastic resonance system model based on nonlinear damping control; Traditional detection algorithms cannot directly process the high-frequency impact echoes acquired in S1. This step constructs the following second-order nonlinear dynamic equation as the core part of signal processing:

[0028] in, For the transformed time scale, satisfying R is the frequency scaling factor; x is the system output. For nonlinear damping terms: Basic damping, This is a nonlinear factor. It provides low damping in the small-signal region to facilitate oscillation initiation and high damping in the large-signal region to maintain steady state. The second-order inertial term imparts bandpass filtering characteristics to the system, effectively filtering out extremely low-frequency geological interference from the well. The physical mechanism under operating conditions is explained below: 1. Frequency scaling mechanism (for the high-frequency characteristics of the striking head): Due to the frequency of the knock echo generated by the active excitation unit 2 The adiabatic approximation region far exceeds the requirements of stochastic resonance. This model introduces a transformation scale. Mathematically, this is equivalent to "slowing down" the high-speed signal acquired by the sensor by a factor of R (where R is the parameter to be optimized, typically around 1000), making it absorbable by the stochastic resonance system.

[0029] 2. Second-order inertial term (Regarding geological interference in deep wells): A "system quality" term is introduced. Physically, this gives the algorithm bandpass filtering characteristics. It can effectively block extremely low frequencies caused by wellbore deformation. The trend term interference is only sensitive to the transient impact characteristics excited by the controllable exciter 21.

[0030] 3. Nonlinear damping term (Regarding non-steady wind noise): Latent state (sensitivity): When the trolley passes through a defect-free section, the signal is weak ( ), minimum damping ( The system is in a highly sensitive standby state, ready to capture the impact of tiny cracks at any time.

[0031] Steady-state response (wind noise resistance): When encountering strong airflow disturbances that cause a sharp increase in amplitude, the nonlinear damping term increases rapidly, acting as an "adaptive brake" to prevent the output signal from diverging and ensure that the system does not give false alarms in harsh environments.

[0032] S3: Based on the background noise intensity of the currently acquired original mixed signal, the system parameters of the variable-scale second-order tristable stochastic resonance model are adaptively adjusted using an optimization algorithm, and the original mixed signal is input into the parameter-optimized model. The weak defect characteristics are enhanced by using the nonlinear damped assisted inertial stochastic resonance effect.

[0033] Specifically, S31, particle initialization: Set the population size to N (e.g., N=30), and the maximum number of iterations is... (For example The six key control parameters of the system are treated as a 6-dimensional particle. The position of each particle i is randomly initialized. and speed The position vector is defined as follows: The parameter value range is set according to the high-frequency characteristics of the impact echo in the tank passage: Frequency scale transformation factor Set the value range to Used to map high-frequency signals to the low-frequency region; basic damping coefficient Value range [0.1, 1.0]; Nonlinear control factor : Value range [0,5]; Potential well structure parameters a, b, c: Value range [0,5].

[0034] S32, each particle The representative parameters are substituted into the second-order nonlinear differential equation described in step S2. The numerical solution is then obtained using the dimension-reduced discretized fourth-order Runge-Kutta method (Rescaled SRK4). 1. Dimensionality reduction: Let This transforms the second-order equations into a system of first-order differential equations.

[0035] 2. Variable step size setting: Set the calculation step size based on the current particle's scale factor R. ,in This is the original sampling frequency. This step ensures that the waveform can be accurately reproduced at different scales.

[0036] 3. Iterative calculation: Obtaining the system output signal .

[0037] Calculate fitness value The method employs a "weighted kurtosis and local signal-to-noise ratio product" to simultaneously consider both impact characteristics (kurtosis) and periodic energy (signal-to-noise ratio).

[0038] Where K is the kurtosis value of the output signal. This represents the local signal-to-noise ratio at the defect characteristic frequency. Weighting coefficients.

[0039] Objective: To find The optimal combination of parameters.

[0040] S33, Individual and Global Extremum Updates: Individual extreme values Recording particles The highest fitness level ever achieved.

[0041] Global Extrema Records the highest fitness level achieved by the entire population throughout its history.

[0042] S34. Adaptive Parameter Adjustment: To address the premature convergence problem of traditional PSO, an adaptive parameter adjustment method is introduced based on the number of iterations. Nonlinearly changing inertial weights and learning factors : Adaptive inertial weights: A non-linear decreasing strategy is adopted, with larger weights in the early stage to facilitate global search and smaller weights in the later stage to facilitate fine convergence.

[0043]

[0044] in, , .

[0045] Learning factor adaptation: .

[0046] Early stage big Smaller particles rely more on their own experience, enhancing ergodicity; later stages Small Larger particles are more "commanded by the collective" and accelerate towards the global optimum.

[0047] S35, Particle State Update: Update the velocity and position of each particle according to the following formula:

[0048] in, for A random number between [a certain number of points].

[0049] S36. Termination Decision: Determine whether the maximum number of iterations has been reached. or global optimal fitness If the improvement fails after multiple consecutive attempts (convergence), then output the globally optimal particle. Corresponding parameters The optimal parameter combination for the current variable damping second-order tristable system is determined; otherwise, return to step S32 to continue iteration.

[0050] S4: Extract defect features from the enhanced signal and combine them with the location information to determine the existence and location of the defect.

[0051] Specifically, this involves optimizing the original signal input parameters for a variable-scale second-order system. The processing effect is as follows: Under the synergistic effect of nonlinear damping and inertial resonance, background noise (especially broadband wind noise) is effectively suppressed by the high-damping mechanism, while weak defect impact signals are significantly amplified by the low-damping mechanism. For example... Figure 4 As shown in the waveform processed by this invention, the originally submerged defect signal appears as a clear, high-amplitude isolated pulse after processing. Decision logic: Set a detection threshold. When the output pulse amplitude exceeds At that time, combined with the encoder data of the positioning unit, the precise depth location of the defect in the wellbore is output.

[0052] Please see Figure 5-7 The present invention also provides a tank passage minor defect inspection system, including an inspection trolley unit 1, for inspection movement along the tank passage; The active excitation unit 2 is used to perform active transient impact on the tank passage, including a controllable vibrator 21, a fixing buckle 22, and a striking head 23. Specifically, the vibrator fixing buckle bolt is fastened to the preset mounting plane at the front of the inspection trolley shell, the controllable vibrator 21 is embedded in the groove of the fixing buckle, and the striking head 23 is adjusted to be perpendicular to the surface of the tank passage (verticality error ≤ 0.5°) to prevent the vibrator from shifting due to vibration during the inspection process.

[0053] The positioning unit 3, used to measure the distance traveled by the inspection trolley unit, includes a positioning wheel 31, a positioning wheel connecting rod 32, and a permanent magnet 33. Specifically, the positioning wheel 31 is connected to the hinge hole at one end of the positioning wheel connecting rod 32 with a bracket via a pin, ensuring that the positioning wheel 31 can rotate flexibly around the pin. The hinge hole at the other end of the positioning wheel connecting rod 32 is hinged to the hinge hole at the rear of the inspection trolley housing via a pin. The permanent magnet 33 is embedded in the pre-reserved groove in the hub of the positioning wheel 31, ensuring that the permanent magnet 33 is coaxial with the positioning wheel 31 and that there are no protrusions on the wheel surface. An installation hole is pre-set in the middle of the positioning wheel connecting rod 32, and the tilt sensor is fixed by bolts.

[0054] On the other hand, the present invention also discloses a method for detecting weak defects in tank passageways and a mobile source emission prediction system based on time-series feature migration, comprising the following units: In another aspect, the present invention also discloses a computer-readable storage medium for a method of detecting minor defects in a tank passage, which stores a computer program. When the computer program is executed by a processor, the processor performs the steps of the method described above.

[0055] In another aspect, the present invention also discloses a computer device for detecting minor defects in tank passages, including a memory and a processor. The memory stores a computer program, and when the computer program is executed by the processor, the processor performs the steps of the method described above.

[0056] In another embodiment provided in this application, a computer program product containing instructions for a method for detecting weak defects in a tank passage is also provided. When the program is run on a computer, it causes the computer to execute any of the mobile source emission prediction methods based on time-series feature migration described in the above embodiments.

[0057] It is understood that the systems, devices, and storage media provided in the embodiments of the present invention correspond to the methods provided in the embodiments of the present invention, and the explanations, examples, and beneficial effects of the relevant content can be referred to the corresponding parts of the above methods.

[0058] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially as a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid state disk (SSD)).

[0059] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0060] The various embodiments in this specification are described in a related manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the system embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0061] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. 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. Such 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 detecting minor defects in tank passageways, characterized in that... , including: S1: Actively and transiently strike the guide rail, and collect the original composite signal on the surface of the guide rail, including weak defect feature signals and strong background noise underground; S2: Receive the original composite signal, and construct a second-order underdamped bistable stochastic resonance system model based on variable-scale frequency shift. The model introduces a second-order inertia term and a frequency scale transformation mechanism to match the high-frequency transient echo signal generated by active striking; S3: Based on the characteristics of the currently collected signal, adaptively adjust the system parameters and damping factor of the second-order underdamped model using an optimization algorithm, and input the original composite signal into the second-order model with optimized parameters. Utilize the inertial stochastic resonance mechanism to make the background noise energy and the defect impact signal generate cooperative resonance, thereby outputting a signal with a significantly enhanced signal-to-noise ratio; S4: Extract defect features from the enhanced signal, and combine the positioning information to determine the existence and location of the defect.

2. The method for detecting minor defects in a tank passageway as described in claim 1, characterized in that: The dynamic equation of the variable-scale second-order bistable stochastic resonance system in step S2 is: in, For the transformed time scale, satisfying , This is the frequency scaling factor; For system output; For nonlinear damping terms, Based on the basic damping coefficient, It is a nonlinear control factor; The structural parameters of the tristable state trap; and These are the defect feature signal and background noise after time-scale compression, respectively.

3. A method for detecting minor defects in a tank passageway as described in claim 1 or 2, characterized in that: The optimization algorithm is an adaptive particle swarm optimization algorithm, which uses the signal-to-noise ratio of the output signal of the bistable stochastic resonance system as the fitness function, and iteratively searches for the system parameter combination that maximizes the signal-to-noise ratio.

4. The method for detecting minor defects in a tank passageway as described in claim 3, characterized in that: The execution of the adaptive particle swarm optimization algorithm includes the following sub-steps: S3.1: Set the particle swarm size M and the maximum number of iterations T, and randomly initialize the position vector Xi and velocity vector Vi for each particle. The position vector represents a set of system control parameters. ,in For frequency scaling factor, For damping parameters, These are the potential well parameters; S3.2: For each particle, substitute the parameters it represents into the variable-scale second-order tristable stochastic resonance system, and introduce a time-scale transformation. Input the original mixed signal and calculate the output signal. The weighted kurtosis multiplied by the local signal-to-noise ratio is used as the fitness value of the particle; S3.3: Record the best historical position for each particle. and the historical best position of the entire particle swarm ; S3.4: Non-linearly adjust the inertia weight w(t), the individual learning factor c1(t), and the social learning factor c2(t) according to the current iteration number t; S3.5: Update the velocity and position of each particle according to the formula; ； where r1 and r2 are random numbers in the interval [0, 1]; S3.6: Iteration and Output: Repeat steps S3.2 to S3.5 until the termination condition is met, and output the global optimal position. The corresponding parameters are the optimal parameter combination for the second-order tristable system.

5. The method for detecting minor defects in a tank passageway as described in claim 4, characterized in that: In step S3.4, the inertia weight w(t) adopts a non-linear decreasing strategy, decreasing from the initial value to the final value as the iteration number increases; and the learning factors satisfy: in the initial stage of iteration, c1(t) > c2(t) to enhance the global search ability; in the later stage of iteration, c1(t) < c2(t) to accelerate convergence.

6. The method for detecting minor defects in a tank passageway as described in claim 1, characterized in that: In step S5, multi-source information fusion determination is performed. When there are characteristic peaks exceeding the threshold in the enhanced vibration signal, and abnormal audio frequency spectrum features are synchronously detected by the sound sensor, and / or abnormal surface texture of the guide rail is captured by the vision sensor, it is comprehensively determined as a confirmed defect.

7. A system for inspecting minor defects in tank passageways, characterized in that: Perform inspection using the inspection method described in any one of claims 1-6. The inspection system includes; An inspection trolley unit for performing inspection movement along the guide rail; An active excitation unit for actively and transiently striking the guide rail; A positioning unit for measuring the distance of the inspection movement of the inspection trolley unit.

8. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, the processor executes the steps of the method described in any one of claims 1 to 6.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the computer program is executed by the processor, the processor executes the steps of the method described in any one of claims 1 to 6.

Images