Excitation signal separation method and device and train

Abstract

This application provides a method, apparatus, and train for separating excitation signals, which can be applied to the fields of rail transit detection and signal measurement technology. The method includes: determining a parameter vector within a preset parameter range for decomposing an initial excitation signal in a first optimization stage; processing the parameter vector according to a global search optimization algorithm corresponding to the first optimization stage to obtain an intermediate parameter vector; decomposing the initial excitation signal based on the intermediate parameter vector to obtain an intermediate signal component, wherein the signal quality represented by the intermediate signal component is lower than a preset signal quality threshold; processing the intermediate parameter vector according to a local search optimization algorithm corresponding to a second optimization stage to obtain a target parameter vector; and decomposing the initial excitation signal based on the target parameter vector to obtain a target signal component, wherein the signal quality of the target signal component meets a preset signal quality threshold.

Description

Technical Field

[0001] This application relates to the field of rail transit detection and signal measurement technology, and more specifically to a method, device and train for excitation signal separation. Background Technology

[0002] Accurate analysis of micro-vibration signals is a crucial step in precision manufacturing, aerospace equipment monitoring, and precision instrument fault diagnosis. The separation accuracy directly impacts the reliability of equipment performance evaluation and fault early warning. To meet the demand for high-precision analysis of micro-vibration signals in low signal-to-noise ratio environments and achieve efficient separation of excitation signals from multi-source noise, a highly noise-resistant and high-precision laser-based micro-vibration signal separation method is urgently needed. Summary of the Invention

[0003] In view of the above problems, this application provides a method, apparatus, and train for separating excitation signals. It also provides an excitation signal separation apparatus, a train control device, a train, electronic equipment, a storage medium, and a program product.

[0004] According to a first aspect of this application, a vibration signal separation method is provided, comprising: determining a parameter vector within a preset parameter range for decomposing an initial vibration signal in a first optimization stage, wherein the initial vibration signal is obtained by acquiring vibration signals from a target component of a running train using a laser vibration measuring device, and the parameter vector is used to decompose the initial vibration signal to obtain signal components indicating the fault vibration response characteristics of the target component; processing the parameter vector according to a global search optimization algorithm corresponding to the first optimization stage to obtain an intermediate parameter vector, wherein the global search optimization algorithm matches the number of optimization operations corresponding to the first optimization stage, and decomposing the initial vibration signal based on the intermediate parameter vector to obtain an intermediate signal component, wherein the signal quality represented by the intermediate signal component is lower than a preset signal quality threshold; processing the intermediate parameter vector according to a local search optimization algorithm corresponding to a second optimization stage to obtain a target parameter vector; and decomposing the initial vibration signal according to the target parameter vector to obtain a target signal component, wherein the signal quality of the target signal component meets the preset signal quality threshold.

[0005] According to a second aspect of this application, a vibration signal separation device is provided, comprising: a determining module, configured to determine, within a preset parameter range, a parameter vector for decomposing an initial vibration signal in a first optimization stage, wherein the initial vibration signal is obtained by acquiring vibration signals from a target component of a running train using a laser vibration measuring device, and the parameter vector is used to decompose the initial vibration signal to obtain a signal component indicating the fault vibration response characteristics of the target component; a first optimization module, configured to process the parameter vector according to a global search optimization algorithm corresponding to the first optimization stage to obtain an intermediate parameter vector, wherein the global search optimization algorithm matches the number of optimization operations corresponding to the first optimization stage, and decomposes the initial vibration signal based on the intermediate parameter vector to obtain an intermediate signal component, wherein the signal quality represented by the intermediate signal component is lower than a preset signal quality threshold; a second optimization module, configured to process the intermediate parameter vector according to a local search optimization algorithm corresponding to the second optimization stage to obtain a target parameter vector; and a decomposition module, configured to decompose the initial vibration signal according to the target parameter vector to obtain a target signal component, wherein the signal quality of the target signal component meets the preset signal quality threshold.

[0006] According to a third aspect of this application, a train is provided, comprising: a train body laser vibration measurement device configured to acquire vibration signals from target components of a running train; and a processor disposed in the train body, the processor being communicatively connected to the laser vibration measurement device and configured to execute a vibration signal separation method.

[0007] A fourth aspect of this application provides an electronic device, comprising: one or more processors; and a memory for storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors perform the method described above.

[0008] A fifth aspect of this application also provides a computer-readable storage medium having executable instructions stored thereon, which, when executed by a processor, cause the processor to perform the methods described above.

[0009] A sixth aspect of this application also provides a computer program product, including a computer program that, when executed by a processor, implements the above-described method.

[0010] According to the embodiments of this application, in the first optimization stage before optimization, a global search optimization algorithm is used to expand the search range and avoid getting trapped in local optima; in the second optimization stage after optimization, a local search optimization algorithm is switched to improve the convergence accuracy, realizing the synergistic complementarity of the two algorithms, making up for the defects of a single algorithm being prone to getting trapped in local optima and insufficient convergence accuracy, effectively overcoming the technical problems of being prone to getting trapped in local optima, insufficient synergy of a single algorithm, and dependence on experience for decomposition parameter settings, and realizing adaptive and accurate matching of parameters; the initial excitation signal is decomposed according to the optimized target parameter vector, improving the separation accuracy and anti-noise interference, and significantly improving the stability, accuracy and algorithm convergence efficiency of signal decomposition. Attached Figure Description

[0011] The above-mentioned contents, other objects, features and advantages of this application will become clearer from the following description of embodiments with reference to the accompanying drawings, in which:

[0012] Figure 1 The illustration schematically shows the application scenario of the excitation signal separation method, apparatus, and train according to embodiments of this application;

[0013] Figure 2 A flowchart illustrating an excitation signal separation method according to an embodiment of this application is shown schematically.

[0014] Figure 3 A schematic diagram of a device for measuring excitation signals based on a vertical vibration generator, according to an embodiment of this application, is shown.

[0015] Figure 4 A schematic diagram of a device for measuring excitation signals based on a horizontal vibration generator, according to an embodiment of this application, is shown.

[0016] Figure 5 A schematic diagram of the initial excitation signal decomposition according to an embodiment of this application is shown;

[0017] Figure 6 A schematic diagram of the excitation signal separation device according to an embodiment of this application is shown.

[0018] Figure 7 A block diagram schematically illustrates an electronic device suitable for implementing an excitation signal separation method, apparatus, and train according to embodiments of this application. Detailed Implementation

[0019] The embodiments of this application will now be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of this application. In the following detailed description, numerous specific details are set forth to provide a thorough understanding of the embodiments of this application for ease of explanation. However, it will be apparent that one or more embodiments may be implemented without these specific details. Furthermore, descriptions of well-known structures and technologies are omitted in the following description to avoid unnecessarily obscuring the concepts of this application.

[0020] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of this application. The terms “comprising,” “including,” etc., as used herein indicate the presence of features, steps, operations, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, or components.

[0021] All terms used herein (including technical and scientific terms) have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein are to be interpreted in a manner consistent with the context of this specification, and not in an idealized or overly rigid way.

[0022] When using expressions such as "at least one of A, B and C", they should generally be interpreted in accordance with the meaning that is commonly understood by those skilled in the art (e.g., "a system having at least one of A, B and C" should include, but is not limited to, a system having A alone, a system having B alone, a system having C alone, a system having A and B, a system having A and C, a system having B and C, and / or a system having A, B and C, etc.).

[0023] In the technical solution of this application, the acquisition, storage, and application of user personal information comply with relevant laws and regulations, necessary confidentiality measures have been taken, and there is no violation of public order and good morals. In the technical solution of this application, user authorization or consent has been obtained before acquiring or collecting user personal information.

[0024] Currently, commonly used laser micro-vibration signal separation methods include filtering, wavelet transform, and variational mode decomposition. Filtering is simple in structure and computationally efficient, but it cannot handle complex aliasing scenarios where the target and noise frequencies are close. Wavelet transform has some time-frequency analysis capabilities, but suffers from technical problems such as mode aliasing and the dependence of decomposition accuracy on wavelet basis selection. Modal signal separation technology is preferred due to its advantages of adaptive decomposition and controllable modal bandwidth, but its decomposition effect is heavily dependent on the number of modes and the penalty factor settings. Inappropriate parameters can easily lead to insufficient decomposition, mode aliasing, or signal distortion, limiting its high-precision applications.

[0025] Among the relevant parameter optimization methods, grid search is inefficient and prone to getting trapped in local optima. Genetic algorithms and particle swarm optimization algorithms are difficult to balance global optimization and local convergence performance in low signal-to-noise ratio and nonlinear scenarios. Each single optimization algorithm has its own advantages and disadvantages. For example, the Whale algorithm has strong global optimization ability and fast convergence, while the Gray Wolf algorithm has high local convergence accuracy and good robustness. However, when used alone, the Whale algorithm is prone to getting trapped in local optima in the later stages, and the Gray Wolf algorithm has insufficient global exploration and is difficult to balance the correlation of target signal and noise suppression effect.

[0026] To at least partially address the technical problems existing in related technologies, this application provides a method for separating excitation signals. The method includes: determining a parameter vector within a preset parameter range for decomposing an initial excitation signal in a first optimization stage; processing the parameter vector according to a global search optimization algorithm corresponding to the first optimization stage to obtain an intermediate parameter vector; decomposing the initial excitation signal based on the intermediate parameter vector to obtain an intermediate signal component, wherein the signal quality represented by the intermediate signal component is lower than a preset signal quality threshold; processing the intermediate parameter vector according to a local search optimization algorithm corresponding to a second optimization stage to obtain a target parameter vector; and decomposing the initial excitation signal based on the target parameter vector to obtain a target signal component, wherein the signal quality of the target signal component meets a preset signal quality threshold.

[0027] In the technical solution of this invention, the user information (including but not limited to user personal information, user image information, user device information, such as location information) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, storage, use, processing, transmission, provision, disclosure, and application of related data all comply with relevant laws, regulations, and standards, take necessary confidentiality measures, do not violate public order and good morals, and provide corresponding operation entry points for users to choose to authorize or refuse.

[0028] Figure 1 The illustration shows an application scenario of the excitation signal separation method, apparatus, and train according to embodiments of this application.

[0029] like Figure 1 As shown, application scenario 100 according to this embodiment may include vehicle 101, network 102, and server 103. Network 102 is used as a medium to provide a communication link between vehicle 101 and server 103. Network 102 may include various connection types, such as wired, wireless communication links, or fiber optic cables, etc.

[0030] Users can operate terminal devices such as computers installed in vehicle 101 to interact with server 103 via network 102 to receive or send messages. Various communication client applications can be installed on the terminal devices, such as shopping applications, web browser applications, search applications, instant messaging tools, email clients, social media platform software, etc. (for example only).

[0031] Terminal devices can be various electronic devices with a display screen and support web browsing, including but not limited to smartphones, tablets, laptops, and desktop computers.

[0032] Server 103 can be a server that provides various services, such as a backend management server that supports websites browsed by users using terminal devices (for example only). The backend management server can analyze and process data such as received user requests, and feed back the processing results (such as web pages, information, or data obtained or generated according to user requests) to the terminal device.

[0033] It should be noted that the excitation signal separation method, device, and train provided in this application embodiment can generally be executed by vehicle 101. Accordingly, the excitation signal separation device and train control device provided in this application embodiment can generally be installed in vehicle 101.

[0034] Alternatively, the excitation signal separation method, apparatus, and train provided in this application embodiment can generally also be executed by server 103. Correspondingly, the excitation signal separation apparatus and train control apparatus provided in this application embodiment can generally also be located in server 103. The excitation signal separation method, apparatus, and train provided in this application embodiment can also be executed by a server or server cluster that is different from server 103 but capable of communicating with vehicle 101 and / or server 103. Correspondingly, the excitation signal separation apparatus and train control apparatus provided in this application embodiment can also be located in a server or server cluster that is different from server 103 but capable of communicating with vehicle 101 and / or server 103.

[0035] It should be understood that Figure 1 The number of vehicles, networks, and servers shown is merely illustrative. Depending on implementation needs, any number of terminal devices, networks, and servers can be included.

[0036] Figure 2 A flowchart illustrating an excitation signal separation method according to an embodiment of this application is shown schematically.

[0037] like Figure 2 As shown, the excitation signal separation method of this embodiment includes operations S210 to S240.

[0038] In operation S210, a parameter vector is determined within a preset parameter range for decomposing the initial excitation signal in the first optimization stage.

[0039] In operation S220, the parameter vector is processed according to the global search optimization algorithm corresponding to the first optimization stage to obtain the intermediate parameter vector.

[0040] In operation S230, the intermediate parameter vector is processed according to the local search optimization algorithm corresponding to the second optimization stage to obtain the target parameter vector.

[0041] In operation S240, the initial excitation signal is decomposed according to the target parameter vector to obtain the target signal components.

[0042] Vibration signals from target components of a running train are collected using laser vibration measurement equipment to obtain an initial excitation signal. The laser vibration measurement equipment can be a laser interferometer. The initial excitation signal is a vibration time-series signal. Adaptive threshold noise reduction processing is performed on the initial excitation signal using wavelet functions to suppress high-frequency environmental noise and eliminate baseline drift caused by equipment vibration, resulting in a preprocessed initial excitation signal.

[0043] The target components can be key rotating mechanical components, load-bearing structural components, or transmission components on the train, such as axles, bearings, gearboxes, etc.

[0044] The parameter vector is used to decompose the initial excitation signal to obtain signal components that indicate the fault vibration response characteristics of the target component.

[0045] The parameter vector can include the number of modes, penalty factor, and number of decompositions. Too many modes will lead to over-decomposition and spurious modes, while too few will result in mode merging and incomplete information. The penalty factor controls the mode bandwidth and affects the decomposition accuracy.

[0046] By iteratively optimizing the parameter vector using an optimization algorithm, a parameter vector for the final decomposition of the initial excitation signal is obtained.

[0047] The global search optimization algorithm is matched with the number of optimization operations corresponding to the first optimization stage. For example, if the total number of optimization operations is preset to T, the current optimization stage is determined to be the first optimization stage if the number of optimization operations is ≤ 0.5T, and the current optimization stage is determined to be the second optimization stage if the number of optimization operations is > 0.5T.

[0048] The global search optimization algorithm can be the Beluga Whale Optimization Algorithm (BWO).

[0049] The preset parameter range can be the initial search range for the White Whale optimization algorithm. The preset parameter range includes the modality number range and the penalty factor range. For example, the modality number range is (3, 15), and the penalty factor range is (100, 5000).

[0050] A parameter vector is randomly generated within the preset parameter range to decompose the initial excitation signal in the first optimization stage.

[0051] In the first optimization stage, the parameter vector is iteratively updated using a global search optimization algorithm to obtain an intermediate parameter vector.

[0052] The intermediate parameter vector represents the parameter vector updated in the first optimization phase.

[0053] Based on the intermediate parameter vector decomposition of the initial excitation signal, multiple intermediate signal components are obtained. The envelope entropy function can be used to process the multiple intermediate signal components separately to obtain the signal quality of each of the multiple intermediate signal components. The signal quality of each of the multiple intermediate signal components is lower than the preset signal quality threshold.

[0054] Signal quality characterizes the clarity of fault features separated after denoising. A smaller envelope entropy indicates higher signal quality. Signal quality values ​​all below a preset threshold indicate that the minimum envelope entropy does not meet the convergence condition of the White Whale optimization algorithm.

[0055] The local search optimization algorithm can be the Grey Wolf Optimizer (GWO).

[0056] In the second optimization stage, the intermediate parameter vector is iteratively updated using the Grey Wolf algorithm to obtain the target parameter vector.

[0057] For example, if the preset total number of optimizations is 100, the White Whale optimization algorithm is used to continuously update the parameter vector in optimization operations with ≤50 optimization operations to obtain the intermediate parameter vector; in optimization operations with >50 optimization operations, the Gray Wolf algorithm is used to continuously update the intermediate parameter vector to obtain the target parameter vector, until the number of optimization operations reaches the preset total number of optimizations or the signal quality of the target signal component meets the preset signal quality threshold, at which point the algorithm optimization stops.

[0058] The target parameter vector represents the parameter vector updated in the second optimization phase.

[0059] For each optimized and updated target parameter vector, the initial excitation signal is decomposed based on the target parameter vector to obtain multiple target signal components. It is then determined whether the optimal signal quality among the multiple target signal components meets the preset signal quality threshold. If the optimal signal quality meets the preset signal quality threshold, the algorithm optimization is stopped.

[0060] The optimal signal quality represents the minimum envelope entropy, and the corresponding target parameter vector is output after the algorithm optimization stops.

[0061] The dual-index system constructed using Spearman correlation coefficient and mutual information entropy can be used to screen signal components from multiple target signal components to evaluate the fault status of target components.

[0062] According to the embodiments of this application, in the first optimization stage before optimization, a global search optimization algorithm is used to expand the search range and avoid getting trapped in local optima; in the second optimization stage after optimization, a local search optimization algorithm is switched to improve the convergence accuracy, realizing the synergistic complementarity of the two algorithms, making up for the defects of a single algorithm being prone to getting trapped in local optima and insufficient convergence accuracy, effectively overcoming the technical problems of being prone to getting trapped in local optima, insufficient synergy of a single algorithm, and dependence on experience for decomposition parameter settings, and realizing adaptive and accurate matching of parameters; the initial excitation signal is decomposed according to the optimized target parameter vector, improving the separation accuracy and anti-interference ability, and significantly improving the stability, accuracy and algorithm convergence efficiency of signal decomposition; in addition, the parameter vector optimization process does not require complex hardware modification, and efficient signal separation can be achieved only through algorithm optimization, which is applicable to a variety of micro-vibration scenarios, and is compatible with mainstream laser interferometers and data acquisition equipment, with strong versatility and engineering practicality.

[0063] According to an embodiment of this application, the intermediate parameter vector is obtained by processing the parameter vector according to the global search optimization algorithm corresponding to the first optimization stage, including: determining the uniformly distributed random number and global exploration intensity of the i-th optimization operation in the first optimization stage, wherein the global exploration intensity is determined based on the number of optimization operations already executed and the preset total number of optimization operations; when the uniformly distributed random number is less than the global exploration intensity, the (i-1)-th intermediate parameter vector is optimized according to the exploration intensity change step size and the random parameter vector to obtain the i-th intermediate parameter vector, wherein the exploration intensity change step size is determined based on the global exploration intensity, the number of optimization operations already executed and the preset number of parameter vectors, and i is an integer greater than 1.

[0064] The uniformly distributed random number B0 is a random number between (0,1) that is randomly generated in each optimization operation. This uniformly distributed random number is used to simulate the randomness and uncertainty of the natural behavior of beluga whales in the beluga optimization algorithm.

[0065] Global exploration intensity is used to simulate the probability of whale fall; the lower the global exploration intensity, the higher the probability of whale fall.

[0066] In one embodiment, global exploration intensity As shown in formula (1):

[0067] (1).

[0068] Where i represents the number of optimization operations that have been performed, and I represents the preset total number of optimization operations.

[0069] When the uniformly distributed random number is less than the global exploration intensity, it means that the algorithm has begun to simulate the whale fall phenomenon in nature in order to maintain population diversity and escape local optima.

[0070] Explore the step size for intensity variation to simulate the step size of a whale falling.

[0071] The random parameter vector represents the position of a beluga individual r randomly selected from the current population in the beluga algorithm.

[0072] The (i-1)th intermediate parameter vector simulates the current position of the beluga whale in the (i-1)th optimization operation, and the (i)th intermediate parameter vector simulates the current position of the beluga whale in the (i-1)th optimization operation.

[0073] The preset parameter vector quantity represents the number of parameter vectors obtained by combining parameters based on a preset parameter range, used to simulate the number of beluga whale individuals in a population.

[0074] In one embodiment, the intensity change step size is explored. As shown in formula (2):

[0075] (2).

[0076] Where C2 is a step factor related to the probability of whale fall and the population size N, C2=2W f ×N, where Lb and Ub are the lower and upper bounds of the algorithm's search space, respectively, and e is the natural decay exponent.

[0077] In one embodiment, the i-th intermediate parameter vector As shown in formula (3):

[0078] (3).

[0079] Among them, r5, r6, and r7 are all random numbers between (0, 1). It is the random parameter vector in the i-th optimization operation, and the current individual position. This represents the (i-1)th intermediate parameter vector.

[0080] According to the embodiments of this application, by introducing global exploration intensity, the global search optimization can reinitialize some individuals with a certain probability during the iteration process, which effectively avoids the problem of premature convergence. When the population gets stuck in a local optimum and the envelope entropy cannot be further reduced, the whale fall mechanism can quickly introduce a new search direction to ensure that the parameter vector of the global optimal mode decomposition is finally obtained.

[0081] According to an embodiment of this application, the intermediate parameter vector is obtained by processing the parameter vector according to the global search optimization algorithm corresponding to the first optimization stage. The method further includes: determining the uniformly distributed random number and global exploration intensity of the i-th optimization operation in the first optimization stage; if the uniformly distributed random number is greater than or equal to the global exploration intensity, obtaining a balance factor based on the number of optimization operations already performed, the preset total number of optimizations, and the uniformly distributed random number; the balance factor characterizes the dynamic balance between global exploration capability and local development capability under the i-th optimization operation; and processing the (i-1)-th intermediate parameter vector according to the search algorithm corresponding to the balance factor to obtain the i-th intermediate parameter vector.

[0082] A uniformly distributed random number greater than or equal to the global exploration intensity indicates that no whale fall phenomenon has occurred.

[0083] In one embodiment, the balance factor As shown in formula (4):

[0084] (4).

[0085] in, Let be a uniformly distributed random number generated during the i-th optimization operation.

[0086] Based on the balance factor, the specific optimization stage of the Whale Optimization Algorithm where the i-th optimization operation is located is determined, thereby determining the corresponding search algorithm, updating the (i-1)-th intermediate parameter vector, and obtaining the i-th intermediate parameter vector.

[0087] According to the embodiments of this application, a balance factor is introduced to realize the transition of the population from the exploration phase to the development phase in the beluga optimization algorithm. In this way, the parameter vector is optimized by simulating the update of the population position, realizing the global optimization capability of adaptive balance, and significantly improving the convergence accuracy and stability of complex multimodal optimization problems.

[0088] According to an embodiment of this application, the (i-1)th intermediate parameter vector is processed according to a search algorithm corresponding to the balance factor to obtain the i-th intermediate parameter vector, including: when the balance factor is less than a preset stage division threshold, optimizing the (i-1)th intermediate parameter vector according to the random parameter vector, the number of parameter categories, and the preset number of parameter vectors to obtain the i-th intermediate parameter vector; when the balance factor is greater than or equal to the preset stage division threshold, optimizing the (i-1)th intermediate parameter vector according to the random parameter vector, the estimated parameter vector, and the jump intensity to obtain the i-th intermediate parameter vector, where the jump intensity characterizes the optimization speed of the intermediate parameter vector, and the estimated parameter vector characterizes the parameter vector corresponding to the preset optimal signal quality in the already executed optimization operation.

[0089] The preset stage division threshold can be 0.5.

[0090] If the balance factor is less than the preset stage division threshold, it means that the i-th optimization operation is in the exploration stage of the beluga optimization algorithm. The search algorithm in the exploration stage is designed based on the swimming behavior of beluga whales, and the search position depends on the swimming pairing behavior of beluga whales.

[0091] In one embodiment, the search algorithm for the search phase is as shown in formula (5):

[0092] (5).

[0093] The number of parameter categories is D, the number of preset parameter vectors is N, and p and r are random integers in the ranges [1,D] and [1,N], respectively. This represents the intermediate parameter vector obtained in the i-th optimization operation, used to simulate the position of the current beluga whale individual t in the j-th dimension. This represents the (i-1)th intermediate parameter vector obtained in the (i-1)th optimization operation, used to simulate the position of the current beluga whale individual t in the j-th dimension. Let r1 and r2 be random parameter vectors in the i-th optimization operation, used to simulate the position of random individual r in random dimension p, where r1 and r2 are random numbers between (0,1).

[0094] If the balance factor is greater than or equal to the preset stage division threshold, it means that the i-th optimization operation is in the development stage of the beluga optimization algorithm. The development stage simulates the beluga's predation behavior, in which individual beluga whales share their locations to hunt, and the locations of other individuals and the optimal individual location must be taken into account.

[0095] In one embodiment, jump intensity As shown in formula (6):

[0096] (6).

[0097] Where r4 are all random numbers between (0,1).

[0098] In one embodiment, the search algorithm during the development phase is shown in formula (7):

[0099] (7).

[0100] in, The predicted parameter vector represents the optimal intermediate vector that appears in the first i optimization operations. The signal quality is optimal based on the signal components decomposed from the optimal intermediate vector. The intermediate parameter vector representing the (i-1)th parameter obtained in the (i-1)th optimization operation. Let represent the intermediate parameter vector obtained in the i-th optimization operation. Let r3 and r4 represent random numbers that conform to the Levy distribution, where r3 and r4 are both random numbers between (0,1).

[0101] In one embodiment, As shown in formulas (8) and (9):

[0102] (8).

[0103] (9).

[0104] Where u and v are normally distributed random numbers, and β is a constant, usually set to 1.5.

[0105] According to an embodiment of this application, the intermediate parameter vector is processed according to the local search optimization algorithm corresponding to the second optimization stage to obtain the target parameter vector, including: decomposing the initial excitation signal based on the intermediate parameter vector to obtain multiple preset signal components; determining multiple reference signal components from the multiple preset signal components according to their respective energy entropy and preset screening rules; and optimizing the intermediate parameter vector according to the multiple reference signal components and their respective convergence coefficients to obtain the target parameter vector.

[0106] The intermediate parameter vector in the first optimization stage is the intermediate parameter vector obtained from the last optimization operation in the first optimization stage. Based on the intermediate parameter vector, the initial excitation signal is decomposed to obtain multiple preset signal components.

[0107] The energy entropy of each of the multiple preset signal components is processed using the envelope entropy function. The multiple preset signal components are then sorted from smallest to largest according to their energy entropy to obtain ordered preset signal components.

[0108] For example, if the preset filtering rule can be to filter the reference signal components at the first three positions, then three reference signal components will be determined from multiple preset signal components.

[0109] In the second optimization stage, during the i-th optimization operation, the reference signal components ranked in the top three positions by energy entropy from smallest to largest are selected based on the parameter vector obtained from the previous i optimization operations. Then, the i-1-th target parameter vector is optimized based on the three reference signal components to obtain the i-th target parameter vector.

[0110] In one embodiment, the local search optimization algorithm corresponding to the second optimization stage is shown in formula (10):

[0111] (10).

[0112] in, , , These are the reference signal components at the first, second, and third positions obtained from the previous i optimization operations, respectively. Let be the target parameter vector of the i-th element, representing the current position of the individual in the local search optimization algorithm. Let i be the (i-1)th objective parameter vector. It is a random coefficient between (0, 2), and a decreases from 2 to 0 with each iteration, thereby gradually improving the accuracy of the local search.

[0113] According to the embodiments of this application, the hierarchical guidance behavior of wolves in the gray wolf algorithm is simulated by the reference signal component, and the range of the searched target parameter vector is locally and precisely optimized; by linearly decreasing the convergence coefficient 'a', the search range is gradually narrowed, the search efficiency is significantly improved, and the parameter accuracy is greatly improved.

[0114] According to an embodiment of this application, decomposing an initial excitation signal based on a target parameter vector to obtain a target signal component includes: decomposing the initial excitation signal based on the target parameter vector to obtain multiple initial signal components; processing the multiple initial signal components using an entropy function to obtain the energy entropy of each of the multiple initial signal components; and determining the initial signal component corresponding to the minimum energy entropy among the multiple energy entropies as the target signal component.

[0115] After obtaining the target parameter vector in each optimization, the initial excitation signal is decomposed using the Variational Mode Decomposition (VMD) method based on the target parameter vector to obtain multiple initial signal components (Inrinsic Mode Function, IMF).

[0116] The envelope signal is obtained by performing a Hilbert transform on the initial signal components.

[0117] In one embodiment, the entropy function is as shown in formula (11):

[0118] (11).

[0119] in, H represents the energy entropy, and H represents the number of initial signal components. Let h be the h-th envelope signal.

[0120] Energy entropy measures the sparsity of a signal. The lower the energy entropy, the stronger the periodicity of the signal and the more obvious the fault characteristics.

[0121] The initial signal component corresponding to the minimum energy entropy among multiple energy entropies is determined as the target signal component. The energy entropy of the target signal component is used as the signal quality, and then compared with a preset signal quality threshold to determine whether to stop the algorithm to optimize the parameter vector.

[0122] According to the embodiments of this application, by selecting the initial signal component corresponding to the minimum energy entropy as the target signal component, the orderliness and purity of the target mode are accurately quantified, which is suitable for micro-vibration signal scenarios with low signal-to-noise ratio and nonlinear coupling. This ensures efficient separation of the target excitation signal from multi-source noise, effectively suppresses the influence of noise interference and redundant information, and significantly enhances the sparsity and identifiability of fault impact features, thereby greatly improving the algorithm's convergence efficiency and the accuracy and reliability of subsequent fault feature extraction.

[0123] According to embodiments of this application, the excitation signal separation method further includes: determining a fault signal component from multiple initial signal components; analyzing the fault signal component to determine the fault state assessment result of the target component.

[0124] The initial signal component containing the most significant fault characteristics can be selected as the fault signal component based on sparsity indicators such as energy entropy, envelope entropy, or correlation kurtosis.

[0125] The fault signal component represents the intrinsic mode information that contains the most significant fault characteristic information of the target component, separated from the initial excitation signal.

[0126] Fault characteristic frequencies and modulation information can be extracted through methods such as envelope demodulation, spectrum analysis, or deep learning.

[0127] By comparing the extracted features with preset thresholds or historical data, the fault type, severity, and development trend can be determined, thereby achieving quantitative assessment and early warning of the health status of the target component.

[0128] According to an embodiment of this application, determining a fault signal component from multiple initial signal components includes: performing correlation analysis on the initial signal components and the initial excitation signal respectively to obtain the correlation coefficients of each of the multiple initial signal components; performing nonlinear analysis on the initial signal components and the initial excitation signal respectively to obtain the mutual information entropy of each of the multiple initial signal components; and determining the initial signal component as a fault signal component if the correlation coefficient of the initial signal component is greater than a preset coefficient threshold or the mutual information entropy is greater than a preset entropy threshold.

[0129] The Spearman rank correlation coefficient can be used to process the initial signal component and the initial excitation signal to obtain the correlation coefficient. The larger the correlation coefficient, the stronger the correlation.

[0130] In one embodiment, the mutual information entropy is as shown in formula (12):

[0131] (12).

[0132] Where H(y) and H(x) are the information entropy of the initial signal component y and the initial excitation signal x, respectively, and H(y, x) is the mutual information entropy.

[0133] The greater the mutual information entropy, the stronger the nonlinear correlation.

[0134] A dual-index system is constructed based on preset coefficient thresholds and preset entropy thresholds, and the fault signal component is the target mode that satisfies the dual-index system.

[0135] If the correlation coefficient of the initial signal component is greater than a preset coefficient threshold or the mutual information entropy is greater than a preset entropy threshold, the initial signal component is identified as a fault signal component.

[0136] According to the embodiments of this application, the effective modes are screened by combining the Spearman correlation coefficient and mutual information entropy dual index system. This not only captures the correlation characteristics between the modal signal components and the initial excitation signal, but also suppresses noise interference, effectively solving the defects of low separation accuracy and weak noise resistance caused by single index screening.

[0137] According to an embodiment of this application, analyzing fault signal components to determine the fault status assessment result of the target component includes: extracting frequency features from the fault signal components to obtain fault features; and performing fault assessment on the target component based on the fault features to determine the fault status assessment result.

[0138] Fault characteristics can be frequency characteristics.

[0139] For example, Hilbert envelope demodulation was performed on the fault signal component of the bearing outer ring to extract the fault characteristic frequency of 156.2 Hz and its second harmonic of 312.4 Hz.

[0140] By comparing the measured fault characteristic frequency amplitude with the preset threshold, the fault status assessment result of the target component is determined.

[0141] For example, by comparing and determining that there is a moderate peeling fault in the outer ring, and at the same time, by combining the sideband distribution to evaluate the load fluctuation, the final fault status assessment result of "outer ring fault - moderate - recommended to repair within 7 days" is output.

[0142] Figure 3 A schematic diagram of an apparatus for measuring excitation signals based on a vertical vibration generator, according to an embodiment of this application, is shown.

[0143] Figure 4 A schematic diagram of an apparatus for measuring excitation signals based on a horizontal vibration generator, according to an embodiment of this application, is shown.

[0144] like Figure 3 , 4As shown, the vertical vibration generator produces a controllable initial excitation signal in the vertical direction, and the horizontal vibration generator produces a controllable initial excitation signal in the horizontal direction. The device mainly includes a signal generator (1), a controller (2), an exciter (3), a vibration table (4), a reflective target (5), a laser interferometer (6), a data acquisition card (7), and a data processing and display unit (8). The signal generator (1) outputs a vibration control signal with a preset frequency, which is amplified and modulated by the controller (2) to drive the vibrator (3) to work; the vibrator (3) drives the vibration table work surface (4) to generate target micro-vibration (single frequency or multi-frequency), and the reflective target (5) is pasted on the vibration table work surface (4) (non-rigid connection, only to enhance the laser reflectivity) and moves synchronously with the work surface; the laser interferometer (6) is fixed on an independent stable base, and its transmitting end emits a laser beam to non-contactly irradiate the reflective target (5), and the receiving end captures the interference light signal reflected by the target, converting the optical path change caused by the micro-vibration into an electrical signal; the data acquisition card (7) performs analog-to-digital conversion on the electrical signal output by the laser interferometer (6), generates a time-series data sequence and transmits it to the data processing and display unit (8); the data processing and display unit (8) has a built-in signal separation algorithm based on a hybrid algorithm to decompose, filter and reconstruct the initial excitation signal, and finally saves and displays the separated fault signal components.

[0145] Figure 5 A schematic diagram of the initial excitation signal decomposition according to an embodiment of this application is shown.

[0146] like Figure 5 As shown, variational mode decomposition of the initial excitation signal is divided into time-domain decomposition ( Figure 5 a) and frequency domain analysis ( Figure 5 b) Two parts.

[0147] Figure 5 A shows the time-domain decomposition results, demonstrating that the initial excitation signal was decomposed into six signal components (IMF 1 to IMF 6). The time-domain waveforms reveal that each IMF exhibits different oscillation characteristics: IMF 1 has the highest frequency and the densest waveform, while IMF 6 has the lowest frequency and the smoothest waveform. All IMFs are physically meaningful amplitude-frequency modulated (AM-FM) signals. The magnified area shows that the high-frequency IMFs (such as IMF 6) contain fine vibrational details, verifying the ability of variational mode decomposition to effectively separate signal components at different time scales. The original signal can be completely reconstructed by superimposing the IMFs.

[0148] Figure 5 b represents the frequency domain analysis results, which show the spectral distribution and center frequency of each IMF through Fourier transform. Figure 5The annotation in b shows that the center frequency of IMF 1 is approximately 5Hz, IMF 2 is approximately 15Hz, and the center frequency of subsequent IMFs increases progressively. The spectral peaks are clear and the bandwidth is narrow, demonstrating the excellent characteristics of compact frequency bands and clear center frequencies for each signal component. The magnified area shows that the spectral peaks of the high-frequency IMFs are sharp and there is no mode aliasing. There are obvious frequency intervals between the spectra, which effectively separates the frequency components of the signal and avoids the frequency aliasing problem in the decomposition process.

[0149] Based on the above-described excitation signal separation method, this application also provides an excitation signal separation device. The following will be combined with... Figure 6 The device is described in detail.

[0150] Figure 6 A schematic block diagram of an excitation signal separation device according to an embodiment of this application is shown.

[0151] like Figure 6 As shown, the excitation signal separation device 600 of this embodiment includes a determination module 610, a first optimization module 620, a second optimization module 630, and a decomposition module 640.

[0152] The determination module 610 is used to determine, within a preset parameter range, a parameter vector for decomposing the initial excitation signal in the first optimization stage. The initial excitation signal is obtained by collecting vibration signals from the target component of the running train using a laser vibration measuring device. The parameter vector is used to decompose the initial excitation signal to obtain signal components that indicate the fault vibration response characteristics of the target component.

[0153] The first optimization module 620 is used to process the parameter vector according to the global search optimization algorithm corresponding to the first optimization stage to obtain the intermediate parameter vector. The global search optimization algorithm matches the number of optimization operations corresponding to the first optimization stage. Based on the intermediate parameter vector, the initial excitation signal is decomposed to obtain the intermediate signal component. The signal quality represented by the intermediate signal component is lower than the preset signal quality threshold.

[0154] The second optimization module 630 is used to process the intermediate parameter vector according to the local search optimization algorithm corresponding to the second optimization stage to obtain the target parameter vector.

[0155] The decomposition module 640 is used to decompose the initial excitation signal according to the target parameter vector to obtain the target signal component, and the signal quality of the target signal component meets the preset signal quality threshold.

[0156] According to an embodiment of this application, the first optimization module 620 includes a first optimization submodule and a second optimization submodule.

[0157] The first optimization submodule is used to determine the uniformly distributed random number and global exploration intensity of the i-th optimization operation in the first optimization stage. The global exploration intensity is determined based on the number of optimization operations already performed and the preset total number of optimization operations.

[0158] The second optimization submodule is used to optimize the (i-1)th intermediate parameter vector based on the exploration intensity change step size and the random parameter vector when the uniformly distributed random number is less than the global exploration intensity, so as to obtain the i-th intermediate parameter vector. The exploration intensity change step size is determined based on the global exploration intensity, the number of optimization operations that have been performed, and the number of preset parameter vectors, where i is an integer greater than 1.

[0159] According to an embodiment of this application, the first optimization module 620 further includes a third optimization submodule, a fourth optimization submodule, and a fifth optimization submodule.

[0160] The third optimization submodule is used to determine the uniformly distributed random number and global exploration intensity of the i-th optimization operation in the first optimization stage.

[0161] The fourth optimization submodule is used to obtain a balance factor based on the number of optimization operations already performed, the preset total number of optimizations, and the uniformly distributed random number, when the uniformly distributed random number is greater than or equal to the global exploration intensity. The balance factor represents the degree of dynamic balance between global exploration capability and local development capability under the i-th optimization operation.

[0162] The fifth optimization submodule is used to process the (i-1)th intermediate parameter vector according to the search algorithm corresponding to the balance factor, and obtain the i-th intermediate parameter vector.

[0163] According to an embodiment of this application, the fifth optimization submodule includes a first optimization unit and a second optimization unit.

[0164] The first optimization unit is used to optimize the (i-1)th intermediate parameter vector based on the random parameter vector, the number of parameter categories, and the preset number of parameter vectors when the balance factor is less than the preset stage division threshold, so as to obtain the i-th intermediate parameter vector.

[0165] The second optimization unit is used to optimize the (i-1)th intermediate parameter vector based on the random parameter vector, the estimated parameter vector, and the jump intensity when the balance factor is greater than or equal to the preset stage division threshold, so as to obtain the i-th intermediate parameter vector. The jump intensity represents the optimization speed of the intermediate parameter vector, and the estimated parameter vector represents the parameter vector corresponding to the preset optimal signal quality in the optimization operation that has been performed.

[0166] According to an embodiment of this application, the second optimization module 630 includes a sixth optimization submodule, a seventh optimization submodule, and an eighth optimization submodule.

[0167] The sixth optimization submodule is used to decompose the initial excitation signal based on the intermediate parameter vector to obtain multiple preset signal components.

[0168] The seventh optimization submodule is used to determine multiple reference signal components from multiple preset signal components based on their respective energy entropy and preset filtering rules.

[0169] The eighth optimization submodule is used to optimize the intermediate parameter vector based on multiple reference signal components and their respective convergence coefficients to obtain the target parameter vector.

[0170] According to an embodiment of this application, the decomposition module 640 includes a first decomposition submodule, a second decomposition submodule, and a third decomposition submodule.

[0171] The first decomposition submodule is used to decompose the initial excitation signal according to the target parameter vector to obtain multiple initial signal components.

[0172] The second decomposition submodule is used to process multiple initial signal components using an entropy function to obtain the energy entropy of each initial signal component.

[0173] The third decomposition submodule is used to determine the initial signal component corresponding to the minimum energy entropy among multiple energy entropies as the target signal component.

[0174] According to an embodiment of this application, the excitation signal separation device 600 further includes an extraction module and an analysis module.

[0175] The extraction module is used to determine the fault signal component from multiple initial signal components.

[0176] The analysis module is used to analyze the fault signal components and determine the fault status assessment results of the target component.

[0177] According to an embodiment of this application, the extraction module includes a first extraction submodule, a second extraction submodule, and a third extraction submodule.

[0178] The first extraction submodule is used to perform correlation analysis on the initial signal components and the initial excitation signal respectively, and obtain the correlation coefficients of each of the multiple initial signal components.

[0179] The second extraction submodule is used to perform nonlinear analysis on the initial signal components and the initial excitation signal respectively, and obtain the mutual information entropy of each of the multiple initial signal components.

[0180] The third extraction submodule is used to identify the initial signal component as a fault signal component when the correlation coefficient of the initial signal component is greater than a preset coefficient threshold or the mutual information entropy is greater than a preset entropy threshold.

[0181] According to an embodiment of this application, the analysis module includes a first analysis submodule and a second analysis submodule.

[0182] The first analysis submodule is used to extract frequency features from the fault signal components to obtain fault features.

[0183] The second analysis submodule is used to perform fault assessment on the target component based on the fault characteristics and determine the fault status assessment result.

[0184] According to embodiments of this application, any plurality of modules among the determining module 610, the first optimization module 620, the second optimization module 630, and the decomposition module 640 can be combined into one module, or any one of these modules can be split into multiple modules. Alternatively, at least part of the functionality of one or more of these modules can be combined with at least part of the functionality of other modules and implemented in one module. According to embodiments of this application, at least one of the determining module 610, the first optimization module 620, the second optimization module 630, and the decomposition module 640 can be at least partially implemented as hardware circuitry, such as a field-programmable gate array (FPGA), a programmable logic array (PLA), a system-on-a-chip, a system-on-a-substrate, a system-on-package, an application-specific integrated circuit (ASIC), or implemented in hardware or firmware by any other reasonable means of integrating or packaging the circuitry, or implemented in software, hardware, or firmware, or in any appropriate combination of any of these three implementation methods. Alternatively, at least one of the determining module 610, the first optimization module 620, the second optimization module 630, and the decomposition module 640 may be at least partially implemented as a computer program module, which can perform corresponding functions when the computer program module is run.

[0185] Figure 7 A block diagram schematically illustrates an electronic device suitable for implementing an excitation signal separation method, apparatus, and train according to embodiments of this application.

[0186] like Figure 7 As shown, an electronic device 700 according to an embodiment of this application includes a processor 701, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 702 or a program loaded from a storage portion 708 into a random access memory (RAM) 703. The processor 701 may include, for example, a general-purpose microprocessor (e.g., a CPU), an instruction set processor and / or an associated chipset and / or a special-purpose microprocessor (e.g., an application-specific integrated circuit (ASIC)), etc. The processor 701 may also include onboard memory for caching purposes. The processor 701 may include a single processing unit or multiple processing units for performing different actions of the method flow according to an embodiment of this application.

[0187] RAM 703 stores various programs and data required for the operation of electronic device 700. Processor 701, ROM 702, and RAM 703 are interconnected via bus 704. Processor 701 executes various operations of the method flow according to embodiments of this application by executing programs in ROM 702 and / or RAM 703. It should be noted that programs may also be stored in one or more memories other than ROM 702 and RAM 703. Processor 701 may also execute various operations of the method flow according to embodiments of this application by executing programs stored in one or more memories.

[0188] According to embodiments of this application, the electronic device 700 may further include an input / output (I / O) interface 705, which is also connected to a bus 704. The electronic device 700 may also include one or more of the following components connected to the I / O interface 705: an input section 706 including a keyboard, mouse, etc.; an output section 707 including a cathode ray tube (CRT), liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 708 including a hard disk, etc.; and a communication section 709 including a network interface card such as a LAN card, modem, etc. The communication section 709 performs communication processing via a network such as the Internet. A drive 710 is also connected to the I / O interface 705 as needed. A removable medium 711, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on the drive 710 as needed so that computer programs read from it can be installed into the storage section 708 as needed.

[0189] This application also provides a computer-readable storage medium, which may be included in the device / apparatus / system described in the above embodiments; or it may exist independently and not assembled into the device / apparatus / system. The computer-readable storage medium carries one or more programs, which, when executed, implement the excitation signal separation method according to the embodiments of this application.

[0190] According to embodiments of this application, the computer-readable storage medium can be a non-volatile computer-readable storage medium, such as including but not limited to: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this application, the computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. For example, according to embodiments of this application, the computer-readable storage medium may include ROM 702 and / or RAM 703 and / or one or more memories other than ROM 702 and RAM 703 described above.

[0191] Embodiments of this application also include a computer program product comprising a computer program containing program code for performing the methods shown in the flowchart. When the computer program product is run on a computer system, the program code enables the computer system to implement the excitation signal separation method provided in the embodiments of this application.

[0192] When the computer program is executed by the processor 701, it performs the functions defined in the system / apparatus of this application embodiment. According to the embodiments of this application, the systems, apparatuses, modules, units, etc., described above can be implemented by computer program modules.

[0193] In one embodiment, the computer program may rely on a tangible storage medium such as an optical storage device or a magnetic storage device. In another embodiment, the computer program may also be transmitted and distributed in the form of signals over a network medium, and may be downloaded and installed via the communication section 709, and / or installed from a removable medium 711. The program code contained in the computer program can be transmitted using any suitable network medium, including but not limited to: wireless, wired, etc., or any suitable combination thereof.

[0194] In such an embodiment, the computer program can be downloaded and installed from a network via the communication section 709, and / or installed from the removable medium 711. When the computer program is executed by the processor 701, it performs the functions defined in the system of this application embodiment. According to the embodiments of this application, the systems, devices, apparatuses, modules, units, etc., described above can be implemented by computer program modules.

[0195] According to embodiments of this application, program code for executing the computer programs provided in the embodiments of this application can be written in any combination of one or more programming languages. Specifically, these computational programs can be implemented using high-level procedural and / or object-oriented programming languages, and / or assembly / machine languages. Programming languages ​​include, but are not limited to, languages ​​such as Java, C++, Python, "C", or similar programming languages. The program code can be executed entirely on the user's computing device, partially on the user's device, partially on a remote computing device, or entirely on a remote computing device or server. In cases involving remote computing devices, the remote computing device can be connected to the user's computing device via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computing device (e.g., via the Internet using an Internet service provider).

[0196] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram or flowchart, and combinations of blocks in a block diagram or flowchart, may be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0197] Those skilled in the art will understand that the features described in the various embodiments and / or claims of this application can be combined or combined in various ways, even if such combinations or combinations are not explicitly described in this application. In particular, the features described in the various embodiments and / or claims of this application can be combined or combined in various ways without departing from the spirit and teachings of this application. All such combinations and / or combinations fall within the scope of this application.

[0198] The embodiments of this application have been described above. However, these embodiments are merely illustrative and not intended to limit the scope of this application. Although various embodiments have been described above, this does not mean that the measures in the various embodiments cannot be used advantageously in combination. The scope of this application is defined by the appended claims and their equivalents. Various substitutions and modifications can be made by those skilled in the art without departing from the scope of this application, and all such substitutions and modifications should fall within the scope of this application.

Claims

1. A method for separating excitation signals, characterized in that, The method includes: Within a preset parameter range, a parameter vector is determined for decomposing the initial excitation signal in the first optimization stage. The initial excitation signal is obtained by collecting vibration signals of the target component of the running train using a laser vibration measuring device. The parameter vector is used to decompose the initial excitation signal to obtain signal components that indicate the fault vibration response characteristics of the target component. The parameter vector is processed by the global search optimization algorithm corresponding to the first optimization stage to obtain an intermediate parameter vector. The global search optimization algorithm is matched with the number of optimization operations corresponding to the first optimization stage. The initial excitation signal is decomposed based on the intermediate parameter vector to obtain an intermediate signal component. The signal quality represented by the intermediate signal component is lower than a preset signal quality threshold. The intermediate parameter vector is processed according to the local search optimization algorithm corresponding to the second optimization stage to obtain the target parameter vector; The initial excitation signal is decomposed according to the target parameter vector to obtain the target signal component, and the signal quality of the target signal component satisfies the preset signal quality threshold.

2. The method according to claim 1, characterized in that, The step of processing the parameter vector according to the global search optimization algorithm corresponding to the first optimization stage to obtain the intermediate parameter vector includes: Determine the uniformly distributed random number and global exploration intensity of the i-th optimization operation in the first optimization stage, wherein the global exploration intensity is determined based on the number of optimization operations already performed and the preset total number of optimization operations; When the uniformly distributed random number is less than the global exploration intensity, the (i-1)th intermediate parameter vector is optimized according to the exploration intensity change step size and the random parameter vector to obtain the i-th intermediate parameter vector. The exploration intensity change step size is determined based on the global exploration intensity, the number of optimization operations already performed, and the number of preset parameter vectors, where i is an integer greater than 1.

3. The method according to claim 1, characterized in that, The step of processing the parameter vector according to the global search optimization algorithm corresponding to the first optimization stage to obtain the intermediate parameter vector further includes: Determine the uniformly distributed random number and the global exploration intensity for the i-th optimization operation in the first optimization phase; When the uniformly distributed random number is greater than or equal to the global exploration intensity, a balance factor is obtained based on the number of optimization operations already performed, the preset total number of optimizations, and the uniformly distributed random number. The balance factor represents the degree of dynamic balance between global exploration capability and local development capability under the i-th optimization operation. The i-1th intermediate parameter vector is processed according to the search algorithm corresponding to the balance factor to obtain the i-th intermediate parameter vector.

4. The method according to claim 3, characterized in that, The step of processing the (i-1)th intermediate parameter vector according to the search algorithm corresponding to the balance factor to obtain the i-th intermediate parameter vector includes: When the balance factor is less than the preset stage division threshold, the (i-1)th intermediate parameter vector is optimized according to the random parameter vector, the number of parameter categories and the preset number of parameter vectors to obtain the i-th intermediate parameter vector; When the balance factor is greater than or equal to the preset stage division threshold, the (i-1)th intermediate parameter vector is optimized according to the random parameter vector, the estimated parameter vector and the jump intensity to obtain the i-th intermediate parameter vector. The jump intensity represents the optimization speed of the intermediate parameter vector, and the estimated parameter vector represents the parameter vector corresponding to the preset optimal signal quality in the optimization operation that has been performed.

5. The method according to claim 1, characterized in that, The intermediate parameter vector is processed according to the local search optimization algorithm corresponding to the second optimization stage to obtain the target parameter vector, including: Based on the intermediate parameter vector, the initial excitation signal is decomposed to obtain multiple preset signal components; Based on the energy entropy of each of the multiple preset signal components and the preset filtering rules, multiple reference signal components are determined from the multiple preset signal components; The intermediate parameter vector is optimized based on the plurality of reference signal components and their respective convergence coefficients to obtain the target parameter vector.

6. The method according to claim 1, characterized in that, The step of decomposing the initial excitation signal according to the target parameter vector to obtain the target signal components includes: The initial excitation signal is decomposed according to the target parameter vector to obtain multiple initial signal components; The energy entropy of each of the initial signal components is obtained by processing the multiple initial signal components using an entropy function. The initial signal component corresponding to the minimum energy entropy among multiple energy entropies is determined as the target signal component.

7. The method according to claim 1, characterized in that, The method further includes: Determine the fault signal component from multiple initial signal components; The fault signal components are analyzed to determine the fault status assessment result of the target component.

8. The method according to claim 7, characterized in that, Determining the fault signal component from multiple initial signal components includes: Correlation analysis was performed on the initial signal components and the initial excitation signal respectively to obtain the correlation coefficients of each of the initial signal components; Nonlinear analysis is performed on the initial signal components and the initial excitation signal respectively to obtain the mutual information entropy of each of the initial signal components; If the correlation coefficient of the initial signal component is greater than a preset coefficient threshold or the mutual information entropy is greater than a preset entropy threshold, the initial signal component is determined to be the fault signal component.

9. The method according to claim 7, characterized in that, The step of analyzing the fault signal components to determine the fault status assessment result of the target component includes: Frequency features are extracted from the fault signal components to obtain fault features; Based on the fault characteristics, the target component is assessed for fault status, and the assessment result is determined.

10. An excitation signal separation device, characterized in that, include: The determination module is used to determine, within a preset parameter range, a parameter vector for decomposing the initial excitation signal in the first optimization stage. The initial excitation signal is obtained by collecting vibration signals of the target component of a running train using a laser vibration measuring device. The parameter vector is used to decompose the initial excitation signal to obtain signal components that indicate the fault vibration response characteristics of the target component. The first optimization module is used to process the parameter vector according to the global search optimization algorithm corresponding to the first optimization stage to obtain an intermediate parameter vector. The global search optimization algorithm is matched with the number of optimization operations corresponding to the first optimization stage. The initial excitation signal is decomposed based on the intermediate parameter vector to obtain an intermediate signal component. The signal quality represented by the intermediate signal component is lower than a preset signal quality threshold. The second optimization module is used to process the intermediate parameter vector according to the local search optimization algorithm corresponding to the second optimization stage to obtain the target parameter vector. The decomposition module is used to decompose the initial excitation signal according to the target parameter vector to obtain the target signal component, wherein the signal quality of the target signal component satisfies the preset signal quality threshold.

11. A train, comprising: Train car body; A laser vibration measurement device is configured to collect vibration signals from target components of the train in operation. A processor is disposed in the train body, the processor is communicatively connected to the laser vibration measuring device, and the processor is configured to execute the excitation signal separation method according to any one of claims 1 to 9.

12. An electronic device, comprising: One or more processors; Memory, used to store one or more programs. Wherein, when the one or more programs are executed by the one or more processors, the one or more processors implement the method of any one of claims 1 to 9.

13. A computer-readable storage medium having a computer program or instructions stored thereon, characterized in that, When the computer program or instructions are executed by a processor, they implement the steps of the method according to any one of claims 1 to 9.

14. A computer program product comprising a computer program that, when executed by a processor, implements the method according to any one of claims 1 to 9.

Images