A method and system for detecting the loop resistance of a crossbonded metal sheath of an earth cable

Abstract

The application discloses a kind of cross interconnection ground cable metal sheath loop resistance detection method and system, it is related to ground loop resistance detection technical field.By at least two different frequency of different frequency excitation current signals to the cable metal sheath loop to be measured time-sharing injection, acquisition voltage and current response signal and are conditioned and analog-digital conversion obtain digital sequence;Signal is handled under each frequency to obtain pure waveform, and calculate complex impedance according to this, and loop resistance is solved based on the complex impedance of at least two frequencies.In denoising process, signal is decomposed into intrinsic mode component using variational mode decomposition, effective component and noise component are separated by combining effective index value, and effective component is finely denoised and reconstructed.The method can effectively suppress power frequency, harmonic and random interference, reduce waveform distortion, improve the accuracy and stability of loop resistance measurement under complex field environment.

Description

Technical Field

[0001] This invention relates to the field of grounding loop resistance detection technology, and specifically to a method and system for detecting the loop resistance of the metal sheath of a cross-connected grounding cable. Background Technology

[0002] With the widespread application of high-voltage and ultra-high-voltage power cables in urban power transmission networks, cross-interconnection grounding methods have been widely adopted due to their ability to effectively reduce induced voltage in the metal sheath and decrease circulating current losses. During long-term operation, the electrical connection status of the cable's metal sheath loop directly affects the sheath grounding effect and the reliability of the cable system. Since the sheath loop consists of the metal sheath, grounding lead, grounding device, and multiple connection joints, its loop resistance parameters can reflect, to a certain extent, the quality of joint connections, the degree of conductor corrosion, and the operating status of the grounding system. Therefore, loop resistance detection has become one of the important aspects of cable operation and maintenance and condition assessment.

[0003] In existing technologies, the live detection of the resistance of cable metal sheath loops typically employs a heterogeneous frequency signal excitation method. For example, Chinese Patent (Publication No.: CN114935694A) discloses a live detection system and method for the resistance of metal shielding in cross-connected cable lines. This system uses an excitation sensor to couple and inject an excitation voltage signal into the metal shielding, simultaneously acquiring the current and voltage parameters of the corresponding loop, constructing an equivalent impedance equation set to calculate the resistance value of the branch under test, and achieving rapid and reliable detection of the loop metal shielding resistance under cable operation.

[0004] However, in actual power field environments, due to the long cable lines, complex laying conditions, and strong surrounding electromagnetic interference, the voltage and current signals acquired during measurement are often affected by power frequency circulating currents, high-order harmonics, switching noise from power electronic equipment, and random electromagnetic interference, resulting in a large amount of noise components mixed in the measurement data. When the noise frequency is close to or superimposed on the frequency of the signal under test, effective information loss, signal waveform distortion, and phase characteristic distortion are likely to occur, thus affecting the accuracy and stability of subsequent parameter calculation results. Due to the combined effect of the above factors, existing methods for detecting the resistance of cable metal sheath loops generally suffer from insufficient anti-interference capability, significant susceptibility of measurement accuracy to environmental influences, and limited accuracy in extracting resistance parameters. Summary of the Invention

[0005] The purpose of this invention is to solve the problem mentioned in the background art of insufficient anti-interference capability of existing detection technology leading to insufficient accuracy of resistance parameter extraction, and to propose a method and system for detecting the resistance of the metal sheath loop of cross-interconnected grounding cables.

[0006] A first aspect of this invention provides a method for detecting the loop resistance of the metallic sheath of a cross-connected grounding cable, the method comprising: At least two different frequency excitation current signals are injected into the metal sheath circuit of the cable under test in a time-division manner, and the voltage response signal and current response signal of the sheath circuit at each excitation frequency are collected. After signal conditioning and analog-to-digital conversion, the original voltage digital sequence and the original current digital sequence are obtained. Denoising processing is performed on the original voltage digital sequence and the original current digital sequence at each excitation frequency to obtain the pure voltage waveform and the pure current waveform. Calculate the complex impedance at the corresponding frequency based on the pure voltage and pure current waveforms at each excitation frequency; solve the loop resistance using the complex impedances at at least two different frequencies, and output the resistance value. The noise reduction process includes: Variational mode decomposition is used to decompose the target signal into multiple intrinsic mode components; the target signal is the original voltage digital sequence or the original current digital sequence at any excitation frequency. Calculate the effective index value for each modal component, and divide the multiple intrinsic modal components into effective components and noise components based on the effective index value; The selected effective components are subjected to refined noise reduction, and the signal is reconstructed based on the noise-reduced effective components to obtain a pure voltage waveform or a pure current waveform.

[0007] Optionally, the number of decomposition levels and the penalty factor of the variational mode decomposition are determined through offline optimization; specifically, a metaheuristic optimization algorithm is used to optimize the mode components with the goal of minimizing the minimum envelope entropy.

[0008] Optionally, the metaheuristic optimization algorithm is an improved gray wolf optimization algorithm; the optimization process includes: Step 1: Initialize the gray wolf population, including the location of each gray wolf; each gray wolf's location corresponds to a set of parameters to be optimized, including the decomposition level and the penalty factor; Step 2: For any gray wolf, denoted as the target gray wolf: Perform variational mode decomposition on each signal sample according to the position of the target gray wolf. For any signal sample, take the minimum value of the envelope entropy of the multiple modal components obtained by decomposition as its entropy index value, and take the average value of the entropy index value of each signal sample as the fitness value of the target gray wolf; the smaller the fitness value, the better. Step 3: Based on the fitness values ​​of each gray wolf, classify the wolf pack into different levels and determine the best individual α wolf, the second best individual β wolf, the third best individual δ wolf, and the remaining individuals ω wolf; Step 4: Use adaptive Gaussian mutation to perturb and update the positions of α wolf, β wolf, and δ wolf to obtain new α wolf, β wolf, and δ wolf. Step 5: Update the coefficient vector A using a preset nonlinear factor; the nonlinear factor decreases nonlinearly and monotonically. Step 6: For each individual gray wolf, update the position of the gray wolf according to the update rules of the standard gray wolf optimization algorithm and the current coefficient vector A, and impose boundary constraints on the position of each gray wolf, and round down the number of decomposition layers. Step 7: Repeat steps 2 to 6 until the preset termination condition is met, and output the optimal decomposition level and penalty factor.

[0009] Optionally, the step of perturbing and updating the positions of α wolf, β wolf, and δ wolf using adaptive Gaussian mutation to obtain new α wolf, β wolf, and δ wolf includes: Normalize each dimension for each gray wolf position, and calculate the standard deviation of each dimension based on the normalized data; The diversity index D of the current population is obtained by adding the standard deviations of the two dimensions. Based on the aforementioned diversity index values, calculate the variation threshold P: ; Among them, P max P min These are the maximum and minimum values ​​of the mutation threshold; Generate a random number rand between (0, 1); If the random number rand is less than the mutation threshold P, then Gaussian mutation is performed on the α wolf, β wolf, and δ wolf: ; Among them, X g It indicates the position of the α wolf, β wolf, or δ wolf; It is X g The position after mutation; Use Gaussian random numbers; If the fitness value of the mutated gray wolf position is smaller than the fitness value of the original gray wolf position, then the mutated gray wolf position is used to update α wolf, β wolf, and δ wolf.

[0010] Optionally, the formula for calculating the nonlinear factor is: ; Where a(t) is the nonlinear factor to be found; t is the current iteration number; T is the maximum iteration number; k is a positive zero adjustment coefficient; and exp() is the natural exponential function.

[0011] Optionally, the step of calculating the effective index value of each modal component and dividing the multiple intrinsic modal components into effective components and noise components based on the effective index value includes: The variance contribution rate of each modal component is calculated as its effective index value; Modal components whose effective index values ​​are greater than a preset threshold are classified as effective components. Modal components whose effective index values ​​are not greater than a preset threshold are classified as noise components.

[0012] Optionally, an improved wavelet threshold denoising algorithm is used to refine the denoising of the selected effective components, specifically including: Perform discrete wavelet transform on the target component to obtain detail coefficients and approximation coefficients; the target component is any effective component. An improved threshold function is used to process the detail coefficients: ; in, It is the j-th detail coefficient of the i-th layer; These are the processed detail coefficients; sgn() is the sign function; is the threshold of the i-th layer; exp() is the natural exponential function; b is a preset adjustment parameter; The approximation coefficients and the processed detail coefficients are subjected to inverse wavelet transform to obtain the effective components after noise reduction.

[0013] A second aspect of this invention provides a system for detecting the loop resistance of the metallic sheath of a cross-connected grounding cable, the system comprising: The data acquisition module is used to inject at least two different frequency excitation current signals into the metal sheath circuit of the cable under test in a time-division manner, and to acquire the voltage response signal and current response signal of the sheath circuit at each excitation frequency. After signal conditioning and analog-to-digital conversion, the original voltage digital sequence and the original current digital sequence are obtained. The signal denoising module is used to perform denoising processing on the original voltage digital sequence and the original current digital sequence at each excitation frequency to obtain a clean voltage waveform and a clean current waveform. The resistance calculation module is used to calculate the complex impedance at the corresponding frequency based on the pure voltage waveform and pure current waveform at each excitation frequency; it uses the complex impedance at at least two different frequencies to solve the loop resistance simultaneously and outputs the resistance value. The signal denoising module includes: The variational mode decomposition module is used to decompose a target signal into multiple intrinsic mode components using variational mode decomposition; the target signal is an original voltage digital sequence or an original current digital sequence at any excitation frequency; The component filtering module is used to calculate the effective index value of each modal component and divide multiple intrinsic modal components into effective components and noise components based on the effective index value. The residual noise reduction module is used to perform fine noise reduction on the selected effective components and reconstruct the signal based on the noise-reduced effective components to obtain a pure voltage waveform or a pure current waveform.

[0014] The beneficial effects of this invention are: The original signal is decomposed into intrinsic mode components by variational mode decomposition, and the effective components and noise components are accurately separated by combining effective index values. The effective components are then finely denoised and reconstructed, which significantly suppresses power frequency circulating current, harmonics and random interference. It overcomes the problems of insufficient anti-interference ability and easy waveform distortion caused by traditional methods, thus greatly improving the accuracy of resistance parameter extraction and measurement stability in complex field environments. Attached Figure Description

[0015] Figure 1 A flowchart of a method for detecting the loop resistance of a cross-connected grounding cable metal sheath provided in an embodiment of the present invention; Figure 2 This is an architecture diagram of a cross-connected grounding cable metal sheath loop resistance detection system provided in an embodiment of the present invention. Detailed Implementation

[0016] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features and effects of the present invention is provided in conjunction with the accompanying drawings and preferred embodiments.

[0017] This invention provides a method for detecting the loop resistance of the metallic sheath of a cross-connected grounding cable. See also... Figure 1 The method includes the following steps: S101, inject at least two different frequency excitation current signals into the metal sheath circuit of the cable under test in a time-division manner, and collect the voltage response signal and current response signal of the sheath circuit at each excitation frequency. After signal conditioning and analog-to-digital conversion, the original voltage digital sequence and the original current digital sequence are obtained.

[0018] S102, perform denoising processing on the original voltage digital sequence and the original current digital sequence at each excitation frequency to obtain the pure voltage waveform and the pure current waveform.

[0019] S103: Calculate the complex impedance at the corresponding frequency based on the pure voltage waveform and pure current waveform at each excitation frequency; solve the loop resistance using the complex impedances at at least two different frequencies, and output the resistance value.

[0020] In one implementation, the acquired voltage and current response signals are subjected to analog bandpass filtering during signal conditioning, so that the filter passband retains only the frequency components near the excitation frequency of the different frequencies, suppressing the 50Hz power frequency and its harmonics as well as the high-frequency noise introduced by the switching power supply, thereby reducing the impact of external interference on subsequent analysis.

[0021] In one embodiment, the noise reduction process includes: Step 1: Use variational mode decomposition to decompose the target signal into multiple intrinsic mode components; the target signal is the original voltage digital sequence or the original current digital sequence at any excitation frequency. Step 2: Calculate the effective index value for each modal component, and divide the multiple intrinsic modal components into effective components and noise components based on the effective index value; Step 3: Perform fine noise reduction on the selected effective components, and reconstruct the signal based on the noise-reduced effective components to obtain a pure voltage waveform or a pure current waveform.

[0022] In one implementation, the number of decomposition levels K and the penalty factor α of the variational mode decomposition are determined through offline optimization. Specifically, a metaheuristic optimization algorithm, such as particle swarm optimization, whale optimization, or gray wolf optimization, is used to optimize the mode components by minimizing their minimum envelope entropy.

[0023] In variational mode decomposition, the selection of the decomposition level K and the penalty factor α directly affects the decomposition effect. If K is too small, under-decomposition is likely to occur, resulting in the inability to effectively separate different frequency components, leading to mode aliasing and mixing of effective signals with noise components. If K is too large, over-decomposition is likely to occur, forming spurious mode components and increasing the complexity of subsequent processing. The penalty factor α is used to constrain the mode bandwidth. When α is too small, the bandwidth of each mode component is too wide, making spectral overlap more likely. When α is too large, the mode bandwidth is too narrow, potentially causing the loss of some effective feature information.

[0024] Envelope entropy characterizes the dispersion of the energy distribution within a signal envelope. When modal components contain significant noise or exhibit modal aliasing, their envelope waveforms are highly random, resulting in a large envelope entropy. Conversely, when the frequency components of the modal components are concentrated and exhibit distinct characteristics, their envelope entropy is relatively small. Therefore, by using the minimum envelope entropy of the modal components as the optimization objective and employing optimization algorithms to find the optimal decomposition level K and penalty factor α, a parameter combination with a better decomposition effect can be automatically obtained, reducing the subjectivity and uncertainty caused by manual parameter selection based on experience.

[0025] The above methods effectively suppress mode aliasing and spurious mode generation, improve the frequency band concentration and characteristic independence of each intrinsic mode component, and concentrate the effective signal energy in the corresponding mode component, thereby improving the distinction between subsequent effective and noise components. Simultaneously, it reduces the degree of noise leakage into the effective components, providing a more reliable data foundation for subsequent refined noise reduction and signal reconstruction, thereby improving the accuracy of complex impedance calculation and the anti-interference capability and accuracy of cable metal sheath loop resistance detection results.

[0026] This invention proposes an improved gray wolf optimization algorithm to optimize the number of decomposition levels K and the penalty factor α in variational mode decomposition. The optimization process includes: Step 1: Initialize the gray wolf population, including the location of each gray wolf; each gray wolf's location corresponds to a set of parameters to be optimized.

[0027] Step two: For any gray wolf, denoted as the target gray wolf: Perform variational mode decomposition on each signal sample based on the target gray wolf's position. For any signal sample, the minimum value of the envelope entropy of its multiple modal components is taken as its entropy index value, and the average value of the entropy index values ​​of each signal sample is taken as the fitness value of the target gray wolf. The smaller the fitness value, the better.

[0028] Step 3: Based on the fitness values ​​of each gray wolf, classify the wolf pack into different levels and determine the best individual α wolf, the second best individual β wolf, the third best individual δ wolf, and the remaining individuals ω wolf.

[0029] Step four: Use adaptive Gaussian mutation to perturb and update the positions of α wolf, β wolf, and δ wolf to obtain new α wolf, β wolf, and δ wolf.

[0030] Step 5: Update the coefficient vector A using a preset nonlinear factor.

[0031] Step 6: For each individual gray wolf, update the position of the gray wolf according to the update rules of the standard gray wolf optimization algorithm and the current coefficient vector A, and impose boundary constraints on the position of each gray wolf, and round down the number of decomposition layers. Step 7: Repeat steps 2 to 6 until the preset termination condition is met, and output the optimal decomposition level and penalty factor.

[0032] In one implementation, the signal samples are multiple sets of typical noisy voltage / current signal sequences selected from historical detection data. These signals cover measured data under different field interference intensities, different cable lengths, and different connection states, and are used for offline optimization of variational mode decomposition parameters.

[0033] In one implementation, the termination condition can be set to reach a preset maximum number of iterations (e.g., 100 times), or the optimal fitness value of the population remains unchanged for several consecutive generations (e.g., 10 generations). At this point, the algorithm is considered to have converged, the iteration stops, and the globally optimal decomposition level and penalty factor are output.

[0034] In one implementation, adaptive Gaussian mutation is used to perturb and update the positions of α wolf, β wolf, and δ wolf, resulting in new α wolf, β wolf, and δ wolf, including: Step 1: Normalize each dimension of each gray wolf location and calculate the standard deviation of each dimension based on the normalized data; add the standard deviations of the two dimensions to obtain the diversity index value D of the current population.

[0035] Step 2: Calculate the variation threshold P based on the diversity index values: Among them, P max P min These are the maximum and minimum values ​​of the mutation threshold, which can be set to 0.5 and 0.05 respectively.

[0036] Step 3: Generate a random number rand between (0, 1). If the random number rand is less than the mutation threshold P, then perform Gaussian mutation on the α wolf, β wolf, and δ wolf: ; where X g It indicates the position of the α wolf, β wolf, or δ wolf; It is X g The position after mutation; Let be a Gaussian random variable with a mean of 0 and a standard deviation of σ, where σ can take the value 1.

[0037] Step 4: Calculate the fitness value of the mutated gray wolf position. If the fitness value of the mutated gray wolf position is smaller than the fitness value of the original gray wolf position, then update α wolf, β wolf, and δ wolf using the mutated gray wolf position.

[0038] In the standard gray wolf optimization algorithm, α, β, and δ wolves act as dominant individuals, guiding the entire population towards the current optimal region. As iterations proceed, the positions of the gray wolves gradually converge, and the population diversity continuously decreases. When the population concentrates in a certain local area, premature convergence is prone to occur, meaning that the current optimal solution is not the global optimal solution, but the population has lost the ability to further explore other regions. This causes the obtained decomposition level K and penalty factor α to deviate from the optimal parameter combination.

[0039] This embodiment first normalizes the positions of each gray wolf in the population and constructs a population diversity index value D using the standard deviation of each dimension. The standard deviation reflects the dispersion of individuals in the search space; therefore, a larger D indicates a more dispersed population distribution, still possessing strong global search capabilities. A smaller D indicates a more concentrated population, with a higher risk of getting trapped in local optima. As the population gradually clusters, the system automatically increases the mutation trigger probability, causing dominant individuals to jump out of the current search area, expanding the search range, and avoiding getting trapped in local optima or premature convergence. By adaptively adjusting the Gaussian mutation probability based on population diversity, a dynamic balance between search capability and convergence performance is achieved, increasing the probability of the algorithm obtaining the global optimum, thereby improving the accuracy and stability of the variational mode decomposition parameter optimization results.

[0040] In one implementation, the update rule of the standard gray wolf optimization algorithm updates the position of each individual gray wolf (ω wolf) based on the positions of α, β, and δ wolves.

[0041] First, calculate the distance between the individual and these three types of leader wolves: Where X represents the current position vector of the individual gray wolf; , , Let these represent the position vectors of α wolf, β wolf, and δ wolf, respectively; r2 is a random vector between [0, 1].

[0042] Then update the location based on the distance: ;in, r1 is a random vector between [0, 1], and a is a convergence factor that decreases linearly with iteration, from 2 to 0, controlling the transition between exploration and development.

[0043] The final new position is: .

[0044] In one implementation, a nonlinear factor is proposed to replace the linear factor in the original standard Grey Wolf optimization algorithm. The formula for calculating this nonlinear factor is as follows: ; Where a(t) is the nonlinear factor to be found; t is the current iteration number; T is the maximum iteration number; k is a positive zero adjustment coefficient, which can be set to 15; exp() is the natural exponential function.

[0045] In the early stages of iteration, 'a' maintains a relatively large value and decreases slowly, allowing the coefficient vector A to fluctuate within a wide range. This enhances the population's ability to search the entire parameter space and prevents premature concentration in local regions. In the middle stages of iteration, 'a' decreases rapidly, enabling the algorithm to transition from the global search phase to the local optimization phase in a timely manner. In the later stages of iteration, 'a' gradually approaches zero and its change becomes more gradual, allowing individual wolves to make more precise position adjustments around the current advantageous solution, improving optimization accuracy. Therefore, compared to the linear decreasing method, this nonlinear convergence factor can more reasonably balance global search capability and local exploitation capability, reduce the risk of getting trapped in local optima and premature convergence, increase the probability of finding the global optimum, and thus obtain a better combination of variational mode decomposition parameters, improving mode separation performance.

[0046] In one implementation, the variance contribution rate is used as an effective indicator. The effective indicator value for each modal component is calculated, and multiple intrinsic modal components are divided into effective components and noise components based on the effective indicator values, including: Step 1: Calculate the variance contribution rate of each modal component.

[0047] Step two: Classify modal components with a variance contribution rate greater than a preset threshold as effective components; classify modal components with a variance contribution rate not greater than a preset threshold as noise components. The preset threshold can be set to 1%.

[0048] The variance contribution rate quantifies the contribution of each modal component to the total energy of the original signal, thus distinguishing between the main signal components and noise components. Effective signals typically have a large variance contribution rate, while noise components have a small variance. By combining this with a preset threshold (such as 1%), effective components can be directly screened. This method is simple to calculate and highly stable.

[0049] One implementation proposes an improved wavelet thresholding denoising algorithm to perform refined denoising on the selected effective components. Specifically, it includes: Step one: Perform a discrete wavelet transform on the target component to obtain detail coefficients and approximation coefficients. The target component is any effective component.

[0050] Step two: Process the detail coefficients using an improved threshold function: ; in, It is the j-th detail coefficient of the i-th layer; These are the processed detail coefficients; sgn() is the sign function; is the threshold of the i-th layer; exp() is the natural exponential function; b is an adjustment parameter greater than 0, which can be set to 0.5.

[0051] Step 3: Perform inverse wavelet transform on the approximation coefficients and the processed detail coefficients to obtain the effective components after noise reduction.

[0052] Traditional hard thresholding functions are discontinuous at the threshold point, leading to pseudo-Gibbs oscillations in the reconstructed signal. While soft thresholding functions are continuous, they constantly compress all coefficients greater than the threshold, causing a constant bias and weakening the signal's true amplitude. The improved thresholding function proposed in this embodiment introduces an exponential term for adaptive smooth transition, making the coefficient decay more gradual. This better preserves effective feature information, avoids the fixed bias problem of soft thresholding, and eliminates the discontinuity of hard thresholding at the threshold point. Therefore, it improves the continuity and fidelity of signal reconstruction while suppressing noise, enhancing the signal-to-noise ratio of the effective components and the stability and accuracy of subsequent resistance calculations.

[0053] In one embodiment, to verify the practical effect of the proposed method for detecting the circuit resistance of the metal sheath of a cross-interconnected grounding cable, this embodiment selects a cross-interconnected grounding cable line from a 220kV substation as the test object and performs live testing while the cable is in operation. Simultaneously, the detection results of this method are compared with the processing results of the traditional heterodyne excitation method.

[0054] The test object was a single-phase metallic sheathed cable circuit. The theoretical design resistance of the circuit (measured offline under power-off conditions using a DC resistance test method, used as a reference value) was 0.215Ω. During the test, the cable load current was 480A, the measured power frequency induced voltage in the sheath was approximately 35V, and there was significant 5th and 7th harmonic interference, as well as broadband random noise from nearby frequency converter equipment. The injected heterogeneous current frequencies were 63Hz and 117Hz.

[0055] The parameters for this method are set as follows: VMD parameters: The improved Grey Wolf optimization algorithm is used, with the number of decomposition levels K=5 and the penalty factor α=1780. Effective component screening threshold: Variance contribution rate >1%. Improved wavelet threshold denoising: db4 wavelet is used, with a decomposition level of 4.

[0056] Traditional method: Digital filtering is performed using FIR notch filters (narrowband notch filters for 50Hz and its integer multiples (100Hz, 150Hz)).

[0057] The acquired voltage and current signals were processed using the two methods described above, and the complex impedances at 63Hz and 117Hz were calculated. The loop resistance was then solved by combining these methods. Ten consecutive measurements were performed, and the average resistance measurement value, absolute error, and relative error were statistically analyzed. The results are shown in Table 1.

[0058] Table 1 As shown in Table 1, compared with the traditional heterodyne excitation detection method using conventional digital filtering, the detection method proposed in this invention can reduce the relative error of loop resistance measurement from 24.7% to 1.86%. The above comparison results fully verify the effectiveness and superiority of the method of this invention.

[0059] This invention provides a system for detecting the loop resistance of the metallic sheath of a cross-connected grounding cable. See also... Figure 2 The system includes: The data acquisition module 201 is used to inject at least two different frequency excitation current signals into the metal sheath circuit of the cable under test in a time-division manner, and to acquire the voltage response signal and current response signal of the sheath circuit at each excitation frequency. After signal conditioning and analog-to-digital conversion, the original voltage digital sequence and the original current digital sequence are obtained. The signal denoising module 202 is used to perform denoising processing on the original voltage digital sequence and the original current digital sequence at each excitation frequency to obtain a clean voltage waveform and a clean current waveform. The resistance calculation module 203 is used to calculate the complex impedance at the corresponding frequency based on the pure voltage waveform and pure current waveform at each excitation frequency; and to solve the loop resistance by using the complex impedances at at least two different frequencies, and output the resistance value. The signal denoising module includes: The variational mode decomposition module is used to decompose the target signal into multiple intrinsic mode components using variational mode decomposition; the target signal is the original voltage digital sequence or the original current digital sequence at any excitation frequency; The component filtering module is used to calculate the effective index value of each modal component and divide multiple intrinsic modal components into effective components and noise components based on the effective index value. The residual noise reduction module is used to perform fine noise reduction on the selected effective components and reconstruct the signal based on the noise-reduced effective components to obtain a pure voltage waveform or a pure current waveform.

[0060] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention are within the scope of the claims of the present invention.

Claims

1. A method for detecting the loop resistance of the metallic sheath of a cross-connected grounding cable, characterized in that, The method includes: At least two different frequency excitation current signals are injected into the metal sheath circuit of the cable under test in a time-division manner, and the voltage response signal and current response signal of the sheath circuit at each excitation frequency are collected. After signal conditioning and analog-to-digital conversion, the original voltage digital sequence and the original current digital sequence are obtained. Denoising processing is performed on the original voltage digital sequence and the original current digital sequence at each excitation frequency to obtain the pure voltage waveform and the pure current waveform. Calculate the complex impedance at the corresponding frequency based on the pure voltage and pure current waveforms at each excitation frequency; solve the loop resistance using the complex impedances at at least two different frequencies, and output the resistance value. The noise reduction process includes: Variational mode decomposition is used to decompose the target signal into multiple intrinsic mode components; the target signal is the original voltage digital sequence or the original current digital sequence at any excitation frequency. Calculate the effective index value for each modal component, and divide the multiple intrinsic modal components into effective components and noise components based on the effective index value; The selected effective components are subjected to refined noise reduction, and the signal is reconstructed based on the noise-reduced effective components to obtain a pure voltage waveform or a pure current waveform.

2. The method for detecting the loop resistance of the metallic sheath of a cross-connected grounding cable according to claim 1, characterized in that, The number of decomposition levels and the penalty factor of the variational mode decomposition are determined through offline optimization; specifically, a metaheuristic optimization algorithm is used to optimize the mode components by minimizing the minimum envelope entropy.

3. The method for detecting the loop resistance of the metallic sheath of a cross-connected grounding cable according to claim 2, characterized in that, The metaheuristic optimization algorithm is an improved Grey Wolf optimization algorithm; The optimization process includes: Step 1: Initialize the gray wolf population, including the location of each gray wolf; each gray wolf's location corresponds to a set of parameters to be optimized, including the decomposition level and the penalty factor; Step 2: For any gray wolf, denoted as the target gray wolf: Perform variational mode decomposition on each signal sample according to the position of the target gray wolf. For any signal sample, take the minimum value of the envelope entropy of the multiple modal components obtained by decomposition as its entropy index value, and take the average value of the entropy index value of each signal sample as the fitness value of the target gray wolf; the smaller the fitness value, the better. Step 3: Based on the fitness values ​​of each gray wolf, classify the wolf pack into different levels and determine the best individual α wolf, the second best individual β wolf, the third best individual δ wolf, and the remaining individuals ω wolf; Step 4: Use adaptive Gaussian mutation to perturb and update the positions of α wolf, β wolf, and δ wolf to obtain new α wolf, β wolf, and δ wolf. Step 5: Update the coefficient vector A using a preset nonlinear factor; the nonlinear factor decreases nonlinearly and monotonically. Step 6: For each individual gray wolf, update the position of the gray wolf according to the update rules of the standard gray wolf optimization algorithm and the current coefficient vector A, and impose boundary constraints on the position of each gray wolf, and round down the number of decomposition layers. Step 7: Repeat steps 2 to 6 until the preset termination condition is met, and output the optimal decomposition level and penalty factor.

4. The method for detecting the loop resistance of the metallic sheath of a cross-connected grounding cable according to claim 3, characterized in that, The step of perturbing and updating the positions of α wolf, β wolf, and δ wolf using adaptive Gaussian mutation to obtain new α wolf, β wolf, and δ wolf includes: Normalize each dimension for each gray wolf position, and calculate the standard deviation of each dimension based on the normalized data; The diversity index D of the current population is obtained by adding the standard deviations of the two dimensions. Based on the aforementioned diversity index values, calculate the variation threshold P: ； Among them, P max P min These are the maximum and minimum values ​​of the mutation threshold; Generate a random number rand between (0, 1); If the random number rand is less than the mutation threshold P, then Gaussian mutation is performed on the α wolf, β wolf, and δ wolf: ； Among them, X g It indicates the position of the α wolf, β wolf, or δ wolf; It is X g The position after mutation; Use Gaussian random numbers; If the fitness value of the mutated gray wolf position is smaller than the fitness value of the original gray wolf position, then the mutated gray wolf position is used to update α wolf, β wolf, and δ wolf.

5. The method for detecting the loop resistance of the metallic sheath of a cross-connected grounding cable according to claim 3, characterized in that, The formula for calculating the nonlinear factor is as follows: ； Where a(t) is the nonlinear factor to be found; t is the current iteration number; T is the maximum iteration number; k is a positive zero adjustment coefficient; and exp() is the natural exponential function.

6. The method for detecting the loop resistance of the metallic sheath of a cross-connected grounding cable according to claim 1, characterized in that, The calculation of the effective index value for each modal component, and the division of multiple intrinsic modal components into effective components and noise components based on the effective index value, includes: The variance contribution rate of each modal component is calculated as its effective index value; Modal components whose effective index values ​​are greater than a preset threshold are classified as effective components. Modal components whose effective index values ​​are not greater than a preset threshold are classified as noise components.

7. The method for detecting the loop resistance of the metallic sheath of a cross-connected grounding cable according to claim 1, characterized in that, An improved wavelet threshold denoising algorithm is used to perform refined denoising on the selected effective components, specifically including: Perform discrete wavelet transform on the target component to obtain detail coefficients and approximation coefficients; the target component is any effective component. An improved threshold function is used to process the detail coefficients: ； in, It is the j-th detail coefficient of the i-th layer; These are the processed detail coefficients; sgn() is the sign function; is the threshold of the i-th layer; exp() is the natural exponential function; b is a preset adjustment parameter; The approximation coefficients and the processed detail coefficients are subjected to inverse wavelet transform to obtain the effective components after noise reduction.

8. A system for detecting the loop resistance of the metallic sheath of a cross-connected grounding cable, characterized in that, The system includes: The data acquisition module is used to inject at least two different frequency excitation current signals into the metal sheath circuit of the cable under test in a time-division manner, and to acquire the voltage response signal and current response signal of the sheath circuit at each excitation frequency. After signal conditioning and analog-to-digital conversion, the original voltage digital sequence and the original current digital sequence are obtained. The signal denoising module is used to perform denoising processing on the original voltage digital sequence and the original current digital sequence at each excitation frequency to obtain a clean voltage waveform and a clean current waveform. The resistance calculation module is used to calculate the complex impedance at the corresponding frequency based on the pure voltage waveform and pure current waveform at each excitation frequency; it uses the complex impedance at at least two different frequencies to solve the loop resistance simultaneously and outputs the resistance value. The signal denoising module includes: The variational mode decomposition module is used to decompose a target signal into multiple intrinsic mode components using variational mode decomposition; the target signal is an original voltage digital sequence or an original current digital sequence at any excitation frequency; The component filtering module is used to calculate the effective index value of each modal component and divide multiple intrinsic modal components into effective components and noise components based on the effective index value. The residual noise reduction module is used to perform fine noise reduction on the selected effective components and reconstruct the signal based on the noise-reduced effective components to obtain a pure voltage waveform or a pure current waveform.

Images