Power cable insulation state detection system based on ultralow frequency dielectric loss

Abstract

The invention discloses a power cable insulation state detection system based on ultralow frequency dielectric loss. The system comprises an excitation signal generation module, a signal acquisition and conversion module, a data processing and analysis module, an insulation state evaluation module, a database storage module and a man-machine interaction module. The signal acquisition and conversion module acquires output voltage and current signals and converts analog signals into digital signals; the data processing and analysis module processes signals based on a system identification theory, establishes an equivalent mathematical model, analyzes data, calculates dielectric parameters and provides a basis for evaluation; the insulation state evaluation module combines historical data with a standard threshold value, designs an evaluation algorithm, analyzes dielectric parameter changes, judges the cable insulation problem, sets a parameter alarm mechanism and automatically gives an alarm when the threshold value is exceeded; according to the method, the detection process is optimized based on the identification theory, ultralow frequency key frequency points are covered, the result precision is ensured through mathematical modeling and analysis, the test time is effectively shortened, and the test result accuracy is high.

Description

Technical Field

[0001] This invention belongs to the field of power cable insulation condition detection technology, specifically relating to a power cable insulation condition detection system based on ultra-low frequency dielectric loss. Background Technology

[0002] Cable insulation condition testing, as an important method for diagnosing cable insulation performance, plays a very positive role in timely detection and elimination of cable faults and improving the safe and stable operation of power systems. Currently, commonly used online testing methods for cable insulation condition testing include partial discharge method, dielectric loss factor method, DC method, and AC superposition method. However, these methods suffer from problems in practical applications, such as weak detection signals, low accuracy, and susceptibility to external interference. To solve these problems, it is necessary to develop a power cable insulation condition testing system based on ultra-low frequency dielectric loss. Summary of the Invention

[0003] The purpose of this invention is to overcome the shortcomings of the prior art and provide a power cable insulation condition detection system based on ultra-low frequency dielectric loss that has good adaptability, short testing time, and high test result accuracy, which can help improve the operation and maintenance efficiency of power systems.

[0004] The objective of this invention is achieved as follows: a power cable insulation condition detection system based on ultra-low frequency dielectric loss, comprising...

[0005] An excitation signal generation module is used to generate ultra-low frequency test excitation signals based on the characteristics of cable insulation materials;

[0006] A signal acquisition and conversion module is used to acquire the output voltage and current signals of the cable insulation system through high-precision sensors, extract ultra-low frequency band characteristics, and convert analog signals into digital signals using high-speed analog-to-digital converters.

[0007] A data processing and analysis module is used to process digital voltage and current signals based on system identification theory, establish an equivalent mathematical model of the cable insulation system, analyze data, calculate dielectric parameters, and provide a basis for evaluation.

[0008] An insulation condition assessment module is used to combine historical data with standard thresholds to design an assessment algorithm, analyze changes in dielectric parameters, identify cable insulation problems, and set up a parameter alarm mechanism to automatically alarm when the threshold is exceeded, prompting maintenance personnel to investigate.

[0009] A database storage module is used to store historical test data, cable information and evaluation results, and supports querying and retrieval, providing data support for long-term tracking;

[0010] A human-computer interaction module is used for a visual interface that supports input of cable parameters and test requirements, as well as real-time display of test progress, dielectric parameters, and evaluation reports.

[0011] Preferably, the ultra-low frequency test excitation signal is a voltage of 1-2000V and a frequency of 1mHz-1kHz, which is compatible with the IEEE P400.2 / D11 standard.

[0012] Preferably, the high-precision sensor includes a high-precision voltage transformer and a current transformer. Specifically, the voltage transformer is connected in parallel and the current transformer is connected in series in the circuit, and grounding and shielding are completed. Then, the signal is synchronously acquired by a 24-bit high-precision ADC acquisition card, and high-quality voltage and current data are output after low-pass filtering and amplification.

[0013] Preferably, the establishment of the equivalent mathematical model of the cable insulation system based on system identification theory specifically includes:

[0014] (1) Voltage and current signal preprocessing: Digital filtering technology is used to filter out high-frequency noise and power frequency interference in the acquired signal, retain the effective signal in the ultra-low frequency band, and normalize the signal to eliminate the influence of signal amplitude differences on subsequent analysis; at the same time, phase-locked loop or high-precision clock is used to trigger sampling to ensure that the sampling time of voltage and current signals is strictly synchronized.

[0015] (2) The equivalent mathematical model includes the equivalent circuit model and the state-space model. The equivalent circuit model treats the insulating medium as multiple RC branches with different time constants connected in parallel, as follows:

[0016]

[0017] In the formula, Y(jω) is the admittance, and R... i C i Let C0 be the resistance and capacitance of each branch, and C0 be the geometric capacitance.

[0018] The state-space model abstracts the cable insulation system into a state-space form, as follows:

[0019]

[0020] In the formula, x(t) is the state variable, u(t) is the input stimulus, and y(t) is the output. The model is established by identification matrices A, B, C, and D.

[0021] (3) Using the preprocessed voltage signal as input u(t) and the current signal as output y(t), construct the input-output dataset {u(t)}. k ),y(t k )}, k=1,2…,N, where N is the number of sampling points;

[0022] (4) Introduce the particle swarm optimization algorithm, encode the model parameters as particles / individuals, use the error function as the fitness, search for the global optimal parameter combination, and improve the model identification accuracy.

[0023] Preferably, the dielectric parameters include the dielectric loss factor and capacitance, and the calculation process for the dielectric parameters is as follows:

[0024] (1) Impedance Calculation: According to Ohm's law, the impedance of a cable insulation system at a certain frequency is... In the formula, the amplitude phase

[0025] (2) Calculate admittance: Admittance In the formula, the amplitude phase

[0026] (3) Calculate the dielectric loss factor: The admittance can be decomposed into conductance G(f) and susceptance B(f), i.e., Y(f) = G(f) + jB(f), then the dielectric loss factor is...

[0027] (4) Calculate capacitance: susceptance For cable insulation systems, the inductance L can be ignored, and B(f) ≈ ωC, therefore the capacitance...

[0028] Preferably, the evaluation algorithm is designed as follows:

[0029] (1) Data standardization: Historical dielectric parameters, including the loss factor tanδ and capacitance C, are stored in time series and standardized using Z-score: x ′ = (xu) / σ, where u and σ are the mean and standard deviation of historical data;

[0030] (2) Feature extraction: Deviation between current value and historical mean Δx = x curr -u hist The rate of change r = (x curr -x prev ) / x prev In the formula x curr Let be the dielectric loss factor tanδ and capacitance C, u at the current moment. hist Let x be the dielectric loss factor tanδ and capacitance C, which are the historical averages. prev Let tanδ be the dielectric loss factor and capacitance C at the previous moment;

[0031] (3) Construction of multidimensional feature vectors: [tanδ,C,Δtanδ,ΔC,r tanδ ,r C Stored in a time-series database;

[0032] (4) Standard threshold system design: set absolute thresholds based on cable characteristics and set confidence interval thresholds based on the statistical distribution of historical data;

[0033] (5) Threshold Comparison and Anomaly Detection: An alarm is triggered directly when the current parameter exceeds the absolute threshold; when the current parameter does not exceed the absolute threshold but exceeds the confidence interval threshold, the anomaly index S is calculated.

[0034]

[0035] In the formula, ω1 and ω2 are weights, and ω1 + ω2 = 1; an early warning is triggered when S > empirical value.

[0036] Due to the adoption of the above technical solution, the beneficial effects of the present invention are:

[0037] This invention employs an excitation signal generation module, a signal acquisition and conversion module, a data processing and analysis module, an insulation condition assessment module, a database storage module, and a human-computer interaction module. Based on identification theory, it optimizes the detection process, covers key frequency points of ultra-low frequency, and mathematical modeling and analysis ensure the accuracy of the results. This not only effectively shortens the test time but also ensures high accuracy of the test results.

[0038] This invention is compatible with various types of cables, and the ultra-low frequency test excitation signal generated based on the characteristics of the cable insulation material is optimal, which can effectively excite the dielectric response of the insulation medium. At the same time, it uses a high-precision sensor to collect data, which has strong anti-interference ability, fast acquisition speed and high accuracy. Together, they ensure the accuracy of the test results of the insulation status of the cable from the source.

[0039] This invention combines historical data with standard thresholds to comprehensively evaluate the insulation status of cables, further ensuring the accuracy of test results. At the same time, it automatically alarms when the threshold is exceeded, ensuring the timely implementation of maintenance work.

[0040] In summary, this invention has the advantages of good adaptability, short testing time, and high accuracy of test results. Attached Figure Description

[0041] Figure 1 This is a schematic diagram of the system components of the present invention. Detailed Implementation

[0042] The technical solution of the present invention will be further described in detail below through embodiments and in conjunction with the accompanying drawings.

[0043] like Figure 1As shown, this invention provides a power cable insulation condition detection system based on ultra-low frequency dielectric loss, including an excitation signal generation module, a signal acquisition and conversion module, a data processing and analysis module, an insulation condition assessment module, a database storage module, and a human-computer interaction module. Based on the theory of ultra-low frequency dielectric loss detection and system identification, it realizes fully automated detection from signal input to insulation condition assessment.

[0044] The excitation signal generation module is used to generate ultra-low frequency test excitation signals based on the characteristics of cable insulation materials. The output voltage range is 1-2000V, the frequency range is 1mHz-1kHz, and it supports signal types such as sine waves and is compatible with the IEEE P400.2 / D11 standard.

[0045] The signal acquisition and conversion module is used to acquire the output voltage and current signals of the cable insulation system through high-precision sensors, extract ultra-low frequency band characteristics, and use high-speed analog-to-digital converters to convert analog signals into digital signals, ensuring acquisition accuracy and speed.

[0046] The high-precision sensor uses high-precision voltage transformers and current transformers. Specifically, the voltage transformers are connected in parallel and the current transformers are connected in series in the circuit. The circuit is grounded and shielded. Then, the signal is synchronously acquired by a 24-bit high-precision ADC acquisition card. After low-pass filtering, amplification and other processing, high-quality voltage and current data are output, laying the foundation for subsequent analysis.

[0047] The data processing and analysis module is used to process digital voltage and current signals based on system identification theory, establish an equivalent mathematical model of the cable insulation system, analyze the data, and calculate dielectric parameters such as dielectric loss factor and capacitance to provide a basis for evaluation.

[0048] First, an equivalent mathematical model of the cable insulation system is established based on system identification theory, specifically as follows:

[0049] (1) Voltage and current signal preprocessing: Digital filtering technology is used to filter out high-frequency noise and power frequency interference in the acquired signal, retain the effective signal in the ultra-low frequency band, and normalize the signal to eliminate the influence of signal amplitude difference on subsequent analysis; at the same time, phase-locked loop or high-precision clock is used to trigger sampling to ensure that the sampling time of voltage and current signals is strictly synchronized to avoid phase calculation deviation.

[0050] (2) The equivalent mathematical model includes the equivalent circuit model and the state-space model. The equivalent circuit model treats the insulating medium as multiple RC branches with different time constants connected in parallel, as follows:

[0051]

[0052] In the formula, Y(jω) is the admittance, used to represent the conductivity of the medium under an alternating electric field (angular frequency ω), and R... i C i Let C0 be the resistance and capacitance of each branch, and C0 be the geometric capacitance.

[0053] The state-space model abstracts the cable insulation system into a state-space form, as follows:

[0054]

[0055] In the formula, x(t) is the state variable, u(t) is the input stimulus, and y(t) is the output. The model is established by identification matrices A, B, C, and D.

[0056] (3) Using the preprocessed voltage signal as input u(t) and the current signal as output y(t), construct the input-output dataset {u(t)}. k ),y(t k )}, k=1,2…,N, where N is the number of sampling points;

[0057] (4) Introduce the particle swarm optimization algorithm, encode the model parameters as particles / individuals, use the error function as the fitness, search for the global optimal parameter combination, and improve the model identification accuracy.

[0058] Then, the dielectric parameters are calculated, as follows:

[0059] (1) Impedance Calculation: According to Ohm's law, the impedance of a cable insulation system at a certain frequency f is... In the formula, the amplitude phase

[0060] (2) Calculate admittance: Admittance In the formula, the amplitude phase

[0061] (3) Calculate the dielectric loss factor: The admittance can be decomposed into conductance G(f) and susceptance B(f), i.e., Y(f) = G(f) + jB(f), then the dielectric loss factor is...

[0062] (4) Calculate capacitance: susceptance For cable insulation systems, the inductance L can be ignored, and B(f) ≈ ωC, therefore the capacitance...

[0063] The insulation condition assessment module is used to combine historical data with standard thresholds to design assessment algorithms, analyze changes in dielectric parameters, determine problems such as cable insulation aging and moisture absorption, and set up a parameter alarm mechanism to automatically alarm when the threshold is exceeded, prompting maintenance and troubleshooting.

[0064] The evaluation algorithm design process is as follows:

[0065] (1) Data standardization: Historical dielectric parameters, including the loss factor tanδ and capacitance C, are stored in time series and standardized using Z-score: x ′ = (xu) / σ, where u and σ are the mean and standard deviation of historical data;

[0066] (2) Feature extraction: Deviation between current value and historical mean Δx = x curr -u hist The rate of change r = (x curr -x prev ) / x prev In the formula, x curr Let be the dielectric loss factor tanδ and capacitance C, u at the current moment. hist Let x be the dielectric loss factor tanδ and capacitance C, which are the historical averages. prev Let tanδ be the dielectric loss factor and capacitance C at the previous moment;

[0067] (3) Construction of multidimensional feature vectors: [tanδ,C,Δtanδ,ΔC,r tanδ ,r C Stored in a time-series database;

[0068] (4) Standard threshold system design: Absolute thresholds are set based on cable characteristics (based on cable age and usage time; for example, at 0.1Hz, tanδ: new cable ≤0.5%, operating cable ≤2%; capacitance change rate: year-on-year ≤5%, month-on-month ≤2%). Simultaneously, confidence interval thresholds are set based on historical data statistical distribution. Warning threshold = u hist +kσ hist In the formula, k = 2 (corresponding to a 95% confidence level), σ hist The standard deviation of historical data is used to identify anomalies that exceed the normal fluctuation range;

[0069] (5) Threshold Comparison and Anomaly Detection: An alarm is triggered directly when the current parameter exceeds the absolute threshold; when the current parameter does not exceed the absolute threshold but exceeds the confidence interval threshold, the anomaly index S is calculated.

[0070]

[0071] In the formula, ω1 and ω2 are weights, and ω1 + ω2 = 1; an early warning is triggered when S > the experience value (derived by experts based on their work experience, such as 2).

[0072] For example, during the rainy season, it was found that the tanδ of a 10kV cable increased from 0.8% to 3.1% at 0.1Hz, exceeding the absolute threshold of 2%, and the capacitance increased by 13%. It was determined that the insulation was damp, and after excavation, it was confirmed that the lead sheath was damaged and water had entered.

[0073] The database storage module stores historical test data, cable information, and evaluation results, supports querying and retrieval, and provides data support for long-term tracking.

[0074] The human-computer interaction module is used to provide a visual interface that supports input of cable parameters and test requirements, as well as real-time display of test progress, dielectric parameters, and evaluation reports.

[0075] In summary, this invention treats the cable insulation system as a two-port network, selects a suitable mathematical model based on the characteristics of the cable insulation material, and then designs the optimal excitation signal and frequency range based on the identification system design. The excitation signal is applied to one end of the cable insulation system, and the output voltage and current signals are obtained from the other end. Finally, based on the obtained voltage and current signals, an equivalent mathematical model is established to describe the cable insulation system as a two-port network, replacing the traditional frequency sweep mode. This not only shortens the test time but also improves the accuracy of the test results.

[0076] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of the claims of the present invention.

Claims

1. A power cable insulation condition detection system based on ultra-low frequency dielectric loss, characterized in that: It includes an excitation signal generation module for generating ultra-low frequency test excitation signals based on the characteristics of cable insulation materials; A signal acquisition and conversion module is used to acquire the output voltage and current signals of the cable insulation system through high-precision sensors, extract ultra-low frequency band characteristics, and convert analog signals into digital signals using high-speed analog-to-digital converters. A data processing and analysis module is used to process digital voltage and current signals based on system identification theory, establish an equivalent mathematical model of the cable insulation system, analyze data, calculate dielectric parameters, and provide a basis for evaluation. An insulation condition assessment module is used to combine historical data with standard thresholds to design an assessment algorithm, analyze changes in dielectric parameters, identify cable insulation problems, and set up a parameter alarm mechanism to automatically alarm when the threshold is exceeded, prompting maintenance personnel to investigate. A database storage module is used to store historical test data, cable information and evaluation results, and supports querying and retrieval, providing data support for long-term tracking; A human-computer interaction module is used for a visual interface that supports inputting cable parameters and test requirements, as well as real-time display of test progress, dielectric parameters, and evaluation reports.

2. The power cable insulation condition detection system based on ultra-low frequency dielectric loss according to claim 1, characterized in that: The ultra-low frequency test excitation signal is a voltage of 1-2000V and a frequency of 1mHz-1kHz, which is compatible with the IEEE P400.2 / D11 standard.

3. The power cable insulation condition detection system based on ultra-low frequency dielectric loss according to claim 1, characterized in that: The high-precision sensor includes a high-precision voltage transformer and a current transformer. Specifically, the voltage transformer is connected in parallel and the current transformer is connected in series in the circuit. The circuit is grounded and shielded. Then, the signal is synchronously acquired by a 24-bit high-precision ADC acquisition card. After low-pass filtering and amplification, high-quality voltage and current data are output.

4. The power cable insulation condition detection system based on ultra-low frequency dielectric loss according to claim 1, characterized in that, The equivalent mathematical model of the cable insulation system established based on system identification theory is as follows: (1) Voltage and current signal preprocessing: Digital filtering technology is used to filter out high-frequency noise and power frequency interference in the acquired signal, retain the effective signal in the ultra-low frequency band, and normalize the signal to eliminate the influence of signal amplitude differences on subsequent analysis; at the same time, phase-locked loop or high-precision clock is used to trigger sampling to ensure that the sampling time of voltage and current signals is strictly synchronized. (2) The equivalent mathematical model includes the equivalent circuit model and the state-space model. The equivalent circuit model treats the insulating medium as multiple RC branches with different time constants connected in parallel, as follows: In the formula, Y(jω) is the admittance, and R... i C i Let C0 be the resistance and capacitance of each branch, and C0 be the geometric capacitance. The state-space model abstracts the cable insulation system into a state-space form, as follows: In the formula, x(t) is the state variable, u(t) is the input stimulus, and y(t) is the output. The model is established by identification matrices A, B, C, and D. (3) Using the preprocessed voltage signal as input u(t) and the current signal as output y(t), construct the input-output dataset {u(t)}. k ),y(t k )}, k=1,2…,N, where N is the number of sampling points; (4) Introduce the particle swarm optimization algorithm, encode the model parameters as particles / individuals, use the error function as the fitness, search for the global optimal parameter combination, and improve the model identification accuracy.

5. The power cable insulation condition detection system based on ultra-low frequency dielectric loss according to claim 1, characterized in that, The dielectric parameters include the loss factor and capacitance. The calculation process for the dielectric parameters is as follows: (1) Impedance Calculation: According to Ohm's law, the impedance of a cable insulation system at a certain frequency is... In the formula, the amplitude phase (2) Calculate admittance: Admittance In the formula, the amplitude phase (3) Calculate the dielectric loss factor: The admittance can be decomposed into conductance G(f) and susceptance B(f), i.e., Y(f) = G(f) + jB(f), then the dielectric loss factor is... (4) Calculate capacitance: susceptance For cable insulation systems, the inductance L can be ignored, and B(f) ≈ ωC, therefore the capacitance...

6. The power cable insulation condition detection system based on ultra-low frequency dielectric loss according to claim 1, characterized in that, The evaluation algorithm design process is as follows: (1) Data standardization: Historical dielectric parameters, including the loss factor tanδ and capacitance C, are stored in time series and standardized using Z-score: x ′ = (xu) / σ, where u and σ are the mean and standard deviation of historical data; (2) Feature extraction: Deviation between current value and historical mean Δx = x curr -u hist The rate of change r = (x curr -x prev ) / x prev In the formula x curr Let be the dielectric loss factor tanδ and capacitance C, u at the current moment. hist Let x be the dielectric loss factor tanδ and capacitance C, which are the historical averages. prev Let tanδ be the dielectric loss factor and capacitance C at the previous moment; (3) Construction of multidimensional feature vectors: [tanδ,C,Δtanδ,ΔC,r tanδ ,r C Stored in a time-series database; (4) Standard threshold system design: set absolute thresholds based on cable characteristics and set confidence interval thresholds based on the statistical distribution of historical data; (5) Threshold comparison and anomaly judgment: When the current parameter exceeds the absolute threshold, an alarm is triggered directly; The anomaly index S is calculated when the current parameter does not exceed the absolute threshold but exceeds the confidence interval threshold: In the formula, ω1 and ω2 are weights, and ω1 + ω2 = 1; an early warning is triggered when S > empirical value.

Images