Cable insulation wide temperature range dielectric loss detection system and detection method thereof

Abstract

The invention belongs to the technical field of electrical equipment insulation detection, and particularly relates to a cable insulation wide-temperature-range dielectric loss detection system which comprises a human-computer interaction interface, a main control processor, an acquisition card, a linear high-voltage amplifier, a micro-current preamplifier, a three-electrode test, a gear shifting communication control module, a power module and a signal conditioning module. The main control processor is connected with the power supply module and the acquisition card, the main control processor is connected with the temperature sensor through the signal conditioning module, the acquisition card is connected with the linear high-voltage amplifier and the micro-current pre-amplifier, and the micro-current pre-amplifier is connected with the gear shifting communication control module; the ultrathin thin film resistance type PT100 temperature sensor is combined with a wave-shaped groove isolation structure, so that the heat conduction error of adjacent temperature measurement nodes is effectively reduced. The depth of the wave-shaped groove is 1.5-3 times of the wire diameter of the thermocouple, and the groove is filled with a nanometer aerogel heat insulation material, thereby guaranteeing the independent measurement of a temperature field, and enabling the thermal interference among a plurality of measurement points to be reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of insulation detection of power equipment, and in particular to a cable insulation wide temperature range dielectric loss detection system and a detection method thereof. Background Art

[0002] In the field of power cable insulation condition assessment, ultra-low frequency dielectric loss measurement technology, due to its high sensitivity to overall cable aging, moisture, and water treeing, has become a core diagnostic method recommended by the international standard IEEE 400-2013. 10kV XLPE cables, due to their excellent electrical performance and mechanical strength, are widely used in urban distribution networks and industrial power supply systems. However, existing technologies for XLPE cables suffer from the following key drawbacks: Current testing methods assess cable condition based on measurements at room temperature (20-30°C), while actual cable operating temperatures can reach 70-90°C. Research has shown that the dielectric loss value (tan δ) and dielectric constant (ε′) of cable insulation materials exhibit nonlinear changes with increasing temperature at test frequencies between 1mHz and 1Hz, such as increased polymer polarization losses and enhanced molecular chain motion. However, existing standards and test equipment lack integrated temperature correction algorithms, resulting in significant deviations between room temperature measurements and actual high-temperature operating conditions, potentially leading to misjudgments or missed detections of aging. In addition, existing research mostly relies on empirical formulas for a single temperature point, such as the fixed temperature coefficient method, lacks dynamic modeling within a wide temperature range, and does not combine the dielectric loss spectrum characteristics for multi-parameter collaborative correction. In addition, due to the lack of effective thermal isolation design, traditional temperature sensors have significant heat conduction interference between multiple measuring points, resulting in accumulated temperature measurement errors; cold-end compensation mostly relies on manual calibration or a single reference point, making it difficult to correct ambient temperature fluctuations in real time. Noise interference and signal attenuation during long-distance signal transmission further reduce data reliability.

[0003] In summary, existing ultra-low-frequency (ULF) dielectric loss measurement systems have significant limitations when evaluating the insulation condition of 10kV cross-linked polyethylene (XLPE) cables: 1. Existing cable temperature sensors lack effective thermal isolation, leading to significant thermal conduction interference between multiple measurement points and cumulative temperature measurement errors. Cold-junction compensation often relies on manual calibration or a single reference point, making it difficult to correct for ambient temperature fluctuations in real time. Noise interference and signal attenuation during long-distance signal transmission further reduce data reliability. 2. The dielectric loss value and dielectric constant of XLPE material increase exponentially with temperature at test frequencies between 1mHz and 1Hz. However, existing standards and equipment still rely on room-temperature measurements and lack correction models for this nonlinear temperature dependence, resulting in evaluation errors under high-temperature conditions. 3. Traditional linear correction methods and static threshold criteria are unable to adapt to the thermal aging effects, water tree growth characteristics, and frequency-temperature superposition effects of XLPE. Furthermore, equipment verification and repair effectiveness evaluation are often detached from actual operating temperatures. These issues severely limit the accuracy of XLPE cable insulation condition diagnosis. An innovative method based on dynamic temperature coefficient correction is urgently needed to bridge the gap between room-temperature testing and actual high-temperature operating conditions. Summary of the Invention

[0004] The object of the present invention is to provide a cable insulation wide temperature range dielectric loss detection system and detection method thereof, so as to solve the problems raised in the above background technology.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a cable insulation wide temperature range dielectric loss detection system, comprising:

[0006] Human-computer interaction interface, main control processor, acquisition card, linear high-voltage amplifier, micro-current preamplifier, three-electrode test, gear adjustment communication control module, power module and signal conditioning module;

[0007] The main control processor is connected to the power module and the acquisition card, the main control processor is connected to the temperature sensor through the signal conditioning module, the acquisition card is connected to the linear high-voltage amplifier and the micro-current preamplifier, and the micro-current preamplifier is connected to the gear adjustment communication control module;

[0008] The main control processor includes a signal generation module, a dielectric response calculation module and a temperature correction module, wherein: the signal generation module is used to generate a digital signal corresponding to the excitation voltage; the dielectric response calculation module is used to process and analyze the response current collected by the acquisition card to obtain the dielectric loss value and dielectric constant; the temperature correction module is used to correct the calculated dielectric loss value and dielectric constant through a temperature correction algorithm;

[0009] The acquisition card includes a digital-to-analog conversion module and an analog-to-digital conversion module, wherein: the digital-to-analog conversion module is used to convert the digital signal of the excitation voltage into a continuous analog output; the analog-to-digital conversion module is used to perform synchronous digital sampling on the input analog voltage and current;

[0010] The linear high-voltage amplifier is used to amplify the output voltage of the acquisition card;

[0011] The micro-current preamplifier includes a micro-current amplification module and a low-pass filter module, wherein: the micro-current amplification module is used to amplify the response current; the low-pass filter module is used for filtering;

[0012] The gear adjustment communication control module dynamically adjusts the gears of the micro-current amplification and low-pass filtering based on the test frequency and the feedback current amplitude to achieve automatic calibration;

[0013] The power module provides stable power, supports AC / DC input and battery backup;

[0014] The three-electrode test includes a high-voltage electrode, a low-voltage electrode, and a ground electrode, wherein the high-voltage electrode is connected to a linear high-voltage amplifier, and the low-voltage electrode is connected to a micro-current preamplifier;

[0015] The human-computer interaction interface is used to display the dielectric loss spectrum, temperature correction curve, user data input and device status, and supports parameter setting and data export;

[0016] The temperature sensor includes a housing, a data acquisition system, a wavy groove, a straight groove, a cold end compensation module and a hot end measurement module;

[0017] The wavy groove is located between the cold-end compensation module and the hot-end measurement module. The cold-end compensation module and the hot-end measurement module are connected to the data acquisition system. Adjacent hot-end measurement modules are isolated by straight grooves.

[0018] The data acquisition system has a built-in cold junction compensation algorithm to achieve automatic temperature correction;

[0019] The wavy grooves and the straight grooves are used to isolate thermal interference between adjacent temperature measurement nodes in the thermocouple array;

[0020] The cold end compensation module is used to monitor the reference end temperature in real time and compensate;

[0021] The hot end measurement module is used to measure the temperature of the object being measured in real time.

[0022] Preferably, the acquisition card is interconnected with the main control processor through a high-speed digital interface, and a duplex communication protocol is used to realize bidirectional data transmission. The digital excitation signal generated by the main control processor is directly written into the internal cache through the DMA controller of the acquisition card. At the same time, the acquisition card transmits the converted response data back to the main control processor memory through the interrupt mechanism. The analog output channel of the acquisition card is connected to the differential input end of the linear high-voltage amplifier through a shielded twisted pair cable. The interface adopts a BNC connector and is connected in series with a 1kΩ impedance matching resistor. The output end is configured with an overvoltage protection circuit to limit the impact of instantaneous surge voltage on the DAC module.

[0023] Preferably, the hot end measurement module comprises a plurality of distributed thermocouples, the measuring ends of which are fixed to the surface of the object to be measured, and the signal output end is connected to the data acquisition system through a flexible wire. The depth of the wavy groove and the straight groove is 1.5 to 3 times the diameter of the thermocouple wire, and the groove is filled with thermal insulation material.

[0024] The data acquisition system is equipped with a digital signal processing unit, which forms a closed-loop feedback with the cold-end compensation module and the hot-end measurement module.

[0025] A cable insulation wide temperature range dielectric loss detection method, based on a cable insulation wide temperature range dielectric loss detection system, comprises the following steps:

[0026] (1) Measure the dielectric loss and dielectric constant of XLPE cables at different temperatures;

[0027] (2) Establish a segmented mathematical model: In the range of 20 to 70°C, use the exponential model formula In the range of 90-130℃, the quadratic exponential model formula is used. Where Δy indicates that there may be individual differences in XLPE cables;

[0028] (3) Optimize the model parameters a1, b1, a2, b2, and c by nonlinear least squares method;

[0029] (4) Storing the model parameters in the non-volatile memory unit flash memory in the main control processor;

[0030] (5) The temperature correction module combines the temperature value T measured by the temperature sensor 修正 The measured data is corrected in real time and the dielectric loss value at the target temperature is output.

[0031] Preferably, the segmented mathematical model uses the lsqcurvefit function in MATLAB to perform nonlinear least squares optimization, including: automatically searching for the optimal parameter combination in the mathematical model through a mathematical iterative algorithm so that the sum of squares of the residuals between the model prediction value and the experimental measurement value is minimized, and the parameters a1, b1, a2, b2, and c are obtained at this time;

[0032] Substitute the dielectric loss value of the cable at room temperature measured experimentally into the formula to calculate To simplify the extrapolation calculation, take the logarithm of both sides of the above formula to change it to the following:

[0033] lny=lna1+b1T

[0034] lny=lna2+b2T+cT 2

[0035] Therefore, the temperature value from the temperature sensor and the dielectric loss value of the dielectric response calculation module can be extrapolated to the dielectric loss value at any temperature in the temperature range of 20°C-130°C using the following formula.

[0036] For the temperature range of 20℃-70℃:

[0037] ln tanδ t =b1·(t-25)+ln tanδ 25 +ln tanδ Δ

[0038] For the temperature range 90℃-130℃:

[0039] ln tanδ t =b1·(t-25)+c·(t d -25 2 )+ln tanδ 25 +ln tanδ Δ

[0040] Preferably, the temperature value T measured by the temperature sensor 修正 Obtained by the built-in adaptive Kalman filter algorithm of the hot end measurement module and the cold end compensation module: Define the state vector X k =[T 测量 , α] T , observation vector Z k =T 冷端 , update the Kalman gain and state estimation through the recursive formula: X k =X k-1 +K k (Z k -HX k-1 ), where: K k is the Kalman gain, which determines the weight of the observation value on the state correction, H=[1,ΔT 环境 ], output the corrected temperature value T 修正 =X k (1), the compensation coefficient α is calibrated by the following formula:

[0041]

[0042] in, The data acquisition system continuously compares and corrects the temperature T 修正 With reference temperature T 冷端 , if the residual exceeds the limit for 5 consecutive times, that is, |T 修正 -T 修正 |>0.5℃, the adaptive algorithm is triggered to update α.

[0043] Preferably, the temperature correction module calculates the specific temperature value desired by the user transmitted from the human-computer interaction interface and the dielectric loss value calculated by the main control processor to obtain the dielectric loss value at the specific temperature, sets the temperature weight coefficient in the temperature transition zone, and adjusts the weight coefficient using an adaptive method based on fuzzy logic. The weight is dynamically adjusted through the membership function and the rule base to achieve a nonlinear transition that is more in line with the material characteristics.

[0044] Compared with the prior art, the present invention has the following beneficial effects:

[0045] 1. The present invention adopts an ultra-thin film resistive PT100 temperature sensor, combined with a wavy groove isolation structure, to effectively reduce the thermal conduction error of adjacent temperature measurement nodes. The depth of the wavy groove is 1.5 to 3 times the diameter of the thermocouple wire. The groove is filled with nano-aerogel insulation material to ensure independent measurement of the temperature field and reduce thermal interference between multiple measurement points. The distributed thermocouple array uses flexible wire connections to adapt to the irregular surface of the cable, ensuring a fit error of less than 0.2mm, avoiding the temperature measurement deviation caused by poor contact in traditional fixed structures. At the same time, the temperature sensor adopts closed-loop cold end compensation technology, built-in Pt100 RTD reference end sensor, and cooperates with the adaptive Kalman filter algorithm to automatically correct the cold end drift caused by ambient temperature fluctuations. Using differential signal transmission, the signal-to-noise ratio is greatly improved, avoiding the signal attenuation problem caused by long-distance transmission.

[0046] 2. The segmented model proposed in the present invention uses an exponential model in the low-temperature zone to accurately fit the thermal activation characteristics of the polarization loss, and introduces an exponential term in the high-temperature zone to adapt to the rapid rise of the conductivity loss. The establishment of the segmented model can effectively improve the error. Compared with the use of a single high-order model, the segmented model will not overfit. Each zone independently optimizes parameters to avoid low-temperature data interfering with the high-temperature zone fitting. In practical applications, it has good generalization performance and requires fewer undetermined parameters than a single high-order model, which reduces the amount of calculation. In embedded systems, the segmented model has lower memory usage and is suitable for real-time processing.

[0047] 3. The present invention adopts adaptive wide temperature range modeling based on fuzzy logic in the temperature transition zone. The dielectric loss value of XLPE cable shows non-monotonic and nonlinear changes in the range of 70-90°C (polymer crystallization melting zone). The traditional fixed weight or linear interpolation method causes a sharp increase in correction error. The contribution weight (ω∈[0,1]) is adjusted in real time according to the temperature deviation (ΔT) and the dielectric loss change rate (d(tanδ) / dT). The dynamic characteristics of the dielectric loss change rate are quantified by the Gaussian membership function and combined with the temperature deviation for comprehensive decision-making to avoid the rigidity of manually setting fixed coefficients. The proposed 9 fuzzy rules cover typical working conditions. Even if the difference in material batches causes the tanδ-T curve to shift, the system can still automatically adapt. At the same time, it is proposed that the precise value of ω calculated by the center of gravity method (such as ω = 0.5 at 80℃) can effectively solve the problem of forced switching of the segmented model in the traditional method at the critical temperature, which leads to a jump in the dielectric loss value. The correction result can be continuously differentiable in the transition range, eliminating the distortion introduced by the artificial threshold and avoiding the false triggering of insulation warnings, such as the false high tanδ caused by the jump being misjudged as "severe aging".

[0048] 4. The device of the present invention has a compact structure and is easy to install. It does not require manual intervention from signal generation to data processing, thus achieving full process automation. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 Schematic diagram of the ultra-low frequency dielectric loss detection system based on the temperature correction model in the present invention;

[0050] Figure 2 Schematic diagram of the working process of the ultra-low frequency dielectric loss detection system of the present invention;

[0051] Figure 3 This is a schematic diagram of the optimized temperature sensor in the present invention;

[0052] Figure 4 This is a flow chart of the adaptive Kalman filter algorithm in the present invention;

[0053] Figure 5 This is a flow chart of the adaptive algorithm based on fuzzy logic in the present invention;

[0054] Figure 6 The dielectric constant diagram extrapolated and experimentally measured in the present invention;

[0055] Figure 7 This is a diagram of the dielectric loss values extrapolated and experimentally measured in the present invention.

[0056] In the figure: 1. Housing; 2. Data acquisition system; 3. Wavy groove; 4. Straight groove; 5. Cold end; 6. Hot end; 7. Copper and constantan twisted pair; 8. Human-computer interaction interface; 9. Ultra-low frequency dielectric loss detection device; 10. Three-electrode system; 11. High-voltage electrode; 12. Low-voltage electrode; 13. Ground electrode; 14. Acrylic housing; 15. Temperature sensor. DETAILED DESCRIPTION

[0057] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0058] In the description of the present invention, it should be understood that the terms "upper", "lower", "front", "back", "left", "right", "top", "bottom", "inside", "outside", etc., indicating directions or positional relationships, are based on the directions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific direction, be constructed and operated in a specific direction, and therefore should not be understood as limiting the present invention.

[0059] Example 1:

[0060] like Figure 1-7 As shown, the present invention provides a cable insulation wide temperature range dielectric loss detection system, which includes: a human-computer interaction interface 8, an ultra-low frequency dielectric loss detection device 9, a three-electrode system 10, and a temperature sensor 11.

[0061] The ultra-low frequency dielectric loss detection device 9 includes: a main control processor, an acquisition card, a linear high-voltage amplifier, a micro-current preamplifier, a gear adjustment communication control module, a power supply module and a signal conditioning module;

[0062] The human-computer interaction interface 8 is connected to the main control processor in the ultra-low frequency dielectric loss detection device 9 through a USB interface, and is used to display the dielectric loss spectrum, temperature correction curve, device status and user-input target temperature value, and supports parameter setting and data export.

[0063] The main control processor mainly includes a signal generation module, a dielectric response calculation module, and a temperature correction module. It uses an STM32H743VIT6 microcontroller produced by STMicroelectronics, an ARM Cortex-M7 core, a double-precision floating-point unit (FPU) and a hardware FFT accelerator, and a main frequency of 480MHz. Specifically, a 1MB Flash memory is integrated on the chip to store the temperature correction model parameters (α = 1.32×10 -3 ,β=0.56,γ=2.15×10 -4 ,δ=0.78,ε=1.05), the parameters are loaded from the external EEPROM through the SPI interface and connected to the human-computer interaction interface 8 through the USB3.0 interface with a transmission rate of 5Gbps; an isolated CAN bus (ADM3053 chip) is used to communicate with the acquisition card.

[0064] The three-electrode assembly 10 comprises a high-voltage electrode 11, a low-voltage electrode 12, a ground electrode 13, and an acrylic housing 14. Both the high-voltage and low-voltage electrodes 11 and 13 are made of copper. The high-voltage electrode is connected via a wire to the output interface of the ultra-low-frequency dielectric loss detector 9, namely, the linear high-voltage amplifier within the ultra-low-frequency dielectric loss detector 9. The excitation voltage is amplified and applied to the cable section. The low-voltage electrode is connected via a wire to the input interface of the ultra-low-frequency dielectric loss detector 9, namely, the microcurrent preamplifier within the ultra-low-frequency dielectric loss detector 9. This transmits the response current to the main control processor within the ultra-low-frequency dielectric loss detector 9 to calculate the dielectric loss value and dielectric constant.

[0065] The temperature sensor 15 comprises a housing 1, a data acquisition system 2, a wavy groove 3, a linear groove 4, a cold end 5, and a hot end 6. The wavy groove 3 is located between the cold end 5 and hot end 6 temperature measurement nodes of the thermocouple to reduce thermal conduction errors at the temperature measurement nodes. The linear groove 4 is located between adjacent hot end 6 temperature measurement nodes in the thermocouple to avoid thermal interference between adjacent nodes and ensure independent measurement of each node. The cold end 5 is located within the housing 1 and monitors the reference end temperature in real time for software compensation. The hot end 6 measurement module is attached to the lower surface of the cable section and is used to measure the temperature of the object being measured in real time. The cold end 5 and hot end temperature measurement nodes are connected via a low-thermoelectric potential differential signal conductor, which is made of copper and constantan twisted pair 7 and coated with a polyimide insulation layer to ensure signal transmission stability at high temperatures. The cold-end temperature measurement node (5) is connected to the input port of data acquisition system (2) via a flexible wire. Data acquisition system (2) is equipped with a digital signal processing unit, which forms a closed-loop feedback loop with the cold-end compensation module and the hot-end measurement module, adjusting the compensation coefficient in real time to keep the temperature measurement error ≤±0.5°C. The thermocouple array consists of six Pt100 thin film sensors arranged in a 2x4 matrix with a spacing of 5mm. The wavy grooves (3) in the thermal insulation structure are 1.2mm deep (twice the wire diameter) and filled with nano-SiO aerogel (density 0.1g / cm 3 , thermal conductivity 0.08W / (m·K)). Signal processing: 24-bit ADC (ADS1256) sampling, Kalman filter algorithm iteration period 10ms, steady-state error ±0.4°C (measured value).

[0066] The distributed thermocouple array in the measurement module of the hot end 6 is closely attached to the cable surface, sensing the temperature change in real time and generating a thermocouple potential signal V 热端 The compensation module of the cold end 5 monitors the ambient temperature through the Pt100 RTD sensor and collects the ambient temperature T in real time. 冷端 The two are physically isolated by the wavy groove 3 to block heat conduction. The depth of the wavy groove 3 is 1.5 to 3 times the diameter of the thermocouple wire. The interior is filled with nano-aerogel insulation material with a thermal conductivity of <0.1W / (mK). Through the thermal resistance model Block the heat conduction path; the thermocouple signal is differentially transmitted to the data acquisition system 2 via the copper-constantan twisted pair. The signal transmission satisfies: V 有效 =V 热端 -V 冷端 +ε 噪声 (ε 噪声≤1μV), electromagnetic interference is suppressed through an independent shielding layer and optoelectronic coupling. After the 24-bit ADC module digitizes the analog signal, the Kalman filter algorithm dynamically corrects the temperature drift error. The compensation coefficient is optimized through the least squares method and closed-loop compared with the cold end 5 reference value. If the residual exceeds the limit, adaptive learning is triggered to update the parameters. At the same time, the laser displacement sensor verifies the thermocouple fit in real time to ensure that the contact impedance is ≤0.1Ω·cm 2 The corrected temperature data is synchronized to the dielectric loss calculation module to drive the wide-temperature range model. Compared to traditional sensors, this design reduces thermal interference by 90% and controls long-term drift error to within ±0.5°C through triple innovations: trench thermal isolation, differential transmission, and a closed-loop algorithm. This significantly improves temperature measurement reliability under complex high-voltage cable operating conditions.

[0067] In the present invention, since the heating of laboratory equipment may cause the ambient temperature to fluctuate slowly, the fixed compensation coefficient cannot dynamically track the cold end drift, resulting in the accumulation of measurement errors. In addition, the temperature sensor is located near the cable slice, and the single-ended signal transmission is easily interfered by the high-voltage electric field coupling of the dielectric loss detection system, resulting in a low signal-to-noise ratio, which seriously affects the compensation accuracy. Therefore, in order to improve the compensation accuracy, the present invention proposes to use an adaptive Kalman filter algorithm for cold end compensation. First, a state space model of the cold end temperature drift is established: the state vector X is defined. k =[T 测量 , α] T , observation vector Z k =T 冷端 Update the Kalman gain and state estimation through the recursive formula: X k =X k-1 +K k (Z k -HX k-1 ), where: H = [1, ΔT 环境 ], output corrected temperature value T 修正 =X k (1) The compensation coefficient α is calibrated by the following formula:

[0068]

[0069] in, Data acquisition system 2 continuously compares and corrects temperature T 修正 With reference temperature T 冷端 , if the residual exceeds the limit for 5 consecutive times, that is, |T 修正 -T 修正 |>0.5℃, the adaptive algorithm is triggered to update α. The corrected temperature value T 修正 The input is sent to the dielectric response calculation module as the input parameter of the segmented-fuzzy fusion model to achieve nonlinear correction of the dielectric loss value over a wide temperature range.

[0070] The temperature correction module measures the dielectric loss value of the XLPE cable at different temperatures for multiple times to obtain the dielectric loss value and dielectric constant of the cross-linked polyethylene (XLPE) cable at different test temperatures. Related research has shown that there is an exponential relationship between the dielectric constant and the dielectric loss value and the temperature. Therefore, a mathematical model of the dielectric loss value and the dielectric constant and the temperature is established. For temperatures in the range [20,70] (unit: °C), For temperatures in the range [90,130] (unit: °C), Where Δy represents the individual differences that may exist in XLPE cables. A mathematical iterative algorithm automatically searches for the optimal parameter combination within the mathematical model to minimize the sum of squared residuals between the model predictions and the experimental measurements. This results in the obtained parameters a1, b1, a2, b2, and c. These parameters are stored in the non-volatile flash memory of the main control processor. During system initialization, these parameters are loaded into random access memory for real-time calculations.

[0071] A temperature weight coefficient is set in the temperature transition zone [70,90] and adjusted using an adaptive method based on fuzzy logic. The weight is dynamically adjusted through membership functions and a rule base, achieving a nonlinear transition that better matches the material properties. The input variables are the temperature deviation (ΔT) and the rate of change of the dielectric loss (Δtanδ / ΔT). The temperature deviation refers to the difference between the current temperature T and the midpoint of the transition zone (80°C), expressed in degrees Celsius. The rate of change of the dielectric loss refers to the slope of the dielectric loss between adjacent temperature points, reflecting the dynamic characteristics of the material's polarization loss. The output variable is the weight coefficient ω(T), which ranges from 0 to 1, where 0 represents complete reliance on the low-temperature model and 1 represents complete reliance on the high-temperature model. The fuzzy logic system design dynamically adjusts the weight through membership functions and fuzzy rules, achieving intelligent correction that better matches the material properties.

[0072] The temperature deviation (ΔT) uses the linguistic variables {"low temperature side", "middle", "high temperature side"}, and the membership function type uses the triangular function, as shown in the figure. The specific calculation formula is as follows:

[0073] Low temperature side: When ΔT∈[-20℃,0℃]:

[0074]

[0075] Middle: When ΔT∈[-10℃,10℃]:

[0076]

[0077] High temperature side: When ΔT∈[0℃,20℃]:

[0078]

[0079] The linguistic variables of the dielectric loss change rate (Δtanδ / ΔT) are designed as: {“smooth”, “medium”, “violent”}, the membership function type uses the Gaussian function, and the specific calculation formula of the membership degree is:

[0080] According to relevant literature records and relevant standards, the relevant parameters of the Gaussian membership function are determined as follows:

[0081]

[0082] The weight coefficient ω(T) is designed as a language variable: {“low weight”, “medium weight”, “high weight”}. In summary, 9 rules are designed for dynamic weight adjustment of fuzzy rules, covering typical working conditions:

[0083]

[0084]

[0085] Defuzzification using the centroid method To calculate the exact ω(T), where μ i is the activation degree of the i-th rule, ω i is the weight center value of the corresponding output. Then the temperature correction formula is tanδ 修正 =(1-ω(T))·tanδ 低温 (T)+ω(T)·tanδ 高温 (T), if ΔT>10℃ (i.e. T>90℃), directly use the high temperature model; if ΔT<-10℃ (i.e. T<70℃), use the low temperature model.

[0086] Example 2:

[0087] Based on Example 1, the XLPE cable to be tested is tested, and the specific implementation method is as follows.

[0088] The samples used in this article were obtained from 10kV commercial XLPE-insulated cables produced by Shanghai Qifan Cable Co., Ltd. Cable slice specimens were obtained by slicing the insulation layer of cable segments using a J / Q Sliver cable slicer, which cuts the cable axially. Slicing parameters were set as follows: a cutting speed of 0.2 mm / s, a feed rate of 0.1 mm / cut, and deionized water circulation to prevent thermal damage. The cable slices were 80 mm wide and 0.2 mm thick. For testing purposes, the cable slices were ultimately cut into square film specimens measuring 80 mm long, 80 mm wide, and 0.2 mm thick. After sample preparation, the resulting slices were ultrasonically cleaned in deionized water using a KQ-500DE ultrasonic cleaner at 40 kHz and 80 W for 30 minutes to remove surface impurities and cutting residue. The specimens were then dried in a 90°C drying oven for 24 hours to remove crosslinking byproducts from the cable insulation and the effects of mechanical stress during slicing. Install the XLPE cable slices in a three-electrode system (high voltage electrode, low voltage electrode, ground electrode, i.e. high voltage electrode 11, low voltage electrode 12, ground electrode 13). The specific steps are as follows:

[0089] (1) The distance between the high-voltage electrode and the low-voltage electrode is set to 50 mm. The ground electrode is in close contact with the back of the sample through a spring clamping mechanism. The contact pressure is controlled at 5N±0.5N to ensure that the contact impedance is ≤0.1Ω·cm 2 .

[0090] (2) The hot end of the temperature sensor is tightly attached to the sample surface through thermal grease (model GD-414), with a fit error of <0.1 mm, and the contact status is monitored in real time by a laser displacement sensor (model LK-G5000).

[0091] (3) After the system is powered on, the initial parameters are set through the human-computer interaction interface: excitation voltage 1kV (linear high-voltage amplifier output ripple <0.3%), test frequency range 1mHz~1kHz (15 frequency points are selected in logarithmic intervals), and target temperature 105℃.

[0092] Perform system calibration before testing:

[0093] (1) Electrical signal calibration: Use standard capacitors (100pF±0.1%, model: SR104) and standard resistors (1GΩ±0.5%, model: VHP101) to replace the sample to verify the accuracy of the dielectric loss calculation module and ensure that the dielectric loss measurement error is less than ±0.1%.

[0094] (2) Temperature calibration: Place the temperature sensor in a constant temperature oil bath (FLUKE 9144), select 10 temperature points within the range of 20°C to 130°C, and compare them with a standard platinum resistance thermometer (PTB-330) through a cold junction compensation module to ensure that the corrected temperature error is ≤±0.3°C.

[0095] To verify the accuracy of the dielectric loss (tanδ) and dielectric constant (ε') extrapolated from 20°C room temperature data to 105°C and evaluate the effectiveness of the segmented-fuzzy fusion model for wide-temperature correction, a programmable ESPEC SH-260 high- and low-temperature test chamber with a temperature range of -70°C to 150°C and an accuracy of ±0.2°C, equipped with a nitrogen atmosphere, was used. The three-electrode system and XLPE specimen were placed within the chamber. The high- and low-voltage leads were connected to an external detection device through PTFE-sealed terminals (with a withstand voltage of 20 kV) located on the chamber's sidewalls to prevent specimen deformation caused by thermal shock. The chamber was heated at a rate of 2°C / min. The target temperature was set at 105°C and maintained constant for 30 minutes. Testing was initiated when the temperature fluctuation was less than ±0.1°C. To suppress XLPE oxidation side reactions at high temperatures, the nitrogen flow rate was set at 10 L / min.

[0096] The extrapolated and experimentally measured dielectric loss values and dielectric constants are as follows: Figure 6 、 7 shown.

[0097] At 80°C, the fuzzy logic algorithm calculates the weight coefficient ω = 0.53, indicating that the contributions of the low-temperature model and the high-temperature model are close to balanced. The corrected dielectric loss value is 2.4% lower than the uncorrected value, verifying the smooth transition ability of the fuzzy rule in the nonlinear region.

[0098] The errors between the dielectric loss value and dielectric constant at the target temperature detected by the device and the dielectric loss value and dielectric constant at 105°C in the experiment are shown in the following table:

[0099] Evaluation Metrics MSE RMSE MAE <![CDATA[R 2 ]]> MaxErr ε' 0.017033 0.13051 0.11259 0.98943 0.25834 tanδ 0.021304 0.14596 0.12042 0.97348 0.28195

[0100] According to the experimental evaluation indicators, the established mathematical model shows excellent performance in predicting the dielectric constant (ε') and dielectric loss value (tanδ) of XLPE cables: the goodness of fit of the model to the dielectric constant (R 2 =0.989) and the explanatory power of dielectric loss value (R 2=0.973) are both close to ideal values, indicating a high degree of agreement between the predicted trends and experimental data. Experimental verification demonstrates that this model achieves accurate predictions of the material's dielectric properties through its efficient algorithm. The mean square error (MSE) between the model's outputted dielectric loss and the experimentally measured values is 0.021, and the MSE between the predicted and measured dielectric constant values is 0.017, demonstrating a high degree of agreement between the predicted results and the experimental data. Compared with traditional empirical models or similar methods, this model significantly reduces prediction bias, maintaining the error within ±3%. This demonstrates the high accuracy and reliability of its extrapolated dielectric parameters at the target temperature, making it widely applicable in fields such as material design and electronic device development, providing an efficient theoretical tool for engineering practice. The model's stability and engineering applicability are verified across a wide frequency range of 1 MHz–1 kHz. Overall, this model demonstrates high-precision prediction capabilities, providing reliable theoretical support for the analysis of XLPE cable dielectric properties. Further expansion to multiple temperature conditions is possible to enhance its generalizability.

[0101] The above shows and describes the basic principles and main features of the present invention and the advantages of the present invention. For those skilled in the art, it is obvious that the present invention is not limited to the details of the above exemplary embodiments, and the present invention can be implemented in other specific forms without departing from the spirit or basic characteristics of the present invention; therefore, no matter from which point of view, the embodiments should be regarded as exemplary and non-restrictive. The scope of the present invention is limited by the appended claims rather than the above description. Therefore, it is intended that all changes that fall within the meaning and scope of the equivalent elements of the claims are included in the present invention, and any figure signs in the claims should not be regarded as limiting the claims involved.

[0102] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A cable insulation wide temperature range dielectric loss detection system, comprising: Human-computer interaction interface, main control processor, acquisition card, linear high-voltage amplifier, micro-current preamplifier, three-electrode test, gear adjustment communication control module, power supply module and signal conditioning module; its characteristics are: The main control processor is connected to the power module and the acquisition card, the main control processor is connected to the temperature sensor through the signal conditioning module, the acquisition card is connected to the linear high-voltage amplifier and the micro-current preamplifier, and the micro-current preamplifier is connected to the gear adjustment communication control module; The main control processor includes a signal generation module, a dielectric response calculation module and a temperature correction module, wherein: the signal generation module is used to generate a digital signal corresponding to the excitation voltage; the dielectric response calculation module is used to process and analyze the response current collected by the acquisition card to obtain the dielectric loss value and dielectric constant; the temperature correction module is used to correct the calculated dielectric loss value and dielectric constant through a temperature correction algorithm; The acquisition card includes a digital-to-analog conversion module and an analog-to-digital conversion module, wherein: the digital-to-analog conversion module is used to convert the digital signal of the excitation voltage into a continuous analog output; the analog-to-digital conversion module is used to perform synchronous digital sampling on the input analog voltage and current; The linear high-voltage amplifier is used to amplify the output voltage of the acquisition card; The micro-current preamplifier includes a micro-current amplification module and a low-pass filter module, wherein: the micro-current amplification module is used to amplify the response current; the low-pass filter module is used for filtering; The gear adjustment communication control module dynamically adjusts the gears of the micro-current amplification and low-pass filtering based on the test frequency and the feedback current amplitude to achieve automatic calibration; The power module provides stable power, supports AC / DC input and battery backup; The three-electrode test includes a high-voltage electrode, a low-voltage electrode, and a ground electrode, wherein the high-voltage electrode is connected to a linear high-voltage amplifier, and the low-voltage electrode is connected to a micro-current preamplifier; The human-computer interaction interface is used to display the dielectric loss spectrum, temperature correction curve, user data input and device status, and supports parameter setting and data export; The temperature sensor includes a housing, a data acquisition system, a wavy groove, a straight groove, a cold end compensation module and a hot end measurement module; The wavy groove is located between the cold-end compensation module and the hot-end measurement module. The cold-end compensation module and the hot-end measurement module are connected to the data acquisition system. Adjacent hot-end measurement modules are isolated by straight grooves. The data acquisition system has a built-in cold junction compensation algorithm to achieve automatic temperature correction; The wavy grooves and the straight grooves are used to isolate thermal interference between adjacent temperature measurement nodes in the thermocouple array; The cold end compensation module is used to monitor the reference end temperature in real time and compensate; The hot end measurement module is used to measure the temperature of the object being measured in real time.

2. The cable insulation wide temperature range dielectric loss detection system according to claim 1, characterized in that: The acquisition card is interconnected with the main control processor via a high-speed digital interface and uses a duplex communication protocol to achieve bidirectional data transmission. The digital excitation signal generated by the main control processor is directly written into the internal cache via the acquisition card's DMA controller. At the same time, the acquisition card transmits the converted response data back to the main control processor's memory through an interrupt mechanism. The analog output channel of the acquisition card is connected to the differential input of the linear high-voltage amplifier via a shielded twisted pair cable. The interface uses a BNC connector and is connected in series with a 1kΩ impedance matching resistor. The output end is equipped with an overvoltage protection circuit to limit the impact of transient surge voltage on the DAC module.

3. The cable insulation wide temperature range dielectric loss detection system according to claim 1, characterized in that: The hot end measurement module includes multiple distributed thermocouples, whose measuring ends are fixed to the surface of the object to be measured, and the signal output end is connected to the data acquisition system through a flexible wire. The depth of the wavy groove and the straight groove is 1.5 to 3 times the diameter of the thermocouple wire, and the groove is filled with thermal insulation material; The data acquisition system is equipped with a digital signal processing unit, which forms a closed-loop feedback with the cold-end compensation module and the hot-end measurement module.

4. A cable insulation wide temperature range dielectric loss detection method, based on the cable insulation wide temperature range dielectric loss detection system according to any one of claims 1 to 6, characterized in that: The following steps are involved: (1) Measure the dielectric loss and dielectric constant of XLPE cables at different temperatures; (2) Establish a segmented mathematical model: In the range of 20 to 70°C, use the exponential model formula In the range of 90-130℃, the quadratic exponential model formula is used. Where Δy indicates that there may be individual differences in XLPE cables; (3) Optimize the model parameters a1, b1, a2, b2, and c by nonlinear least squares method; (4) Storing the model parameters in the non-volatile memory unit flash memory in the main control processor; (5) The temperature correction module combines the temperature value T measured by the temperature sensor 修正 The measured data is corrected in real time and the dielectric loss value at the target temperature is output.

5. A cable insulation wide temperature range dielectric loss detection method according to claim 4, characterized in that: The segmented mathematical model uses the lsqcurvefit function in MATLAB to perform nonlinear least squares optimization, including: automatically searching for the optimal parameter combination in the mathematical model through a mathematical iterative algorithm to minimize the sum of squares of the residuals between the model prediction value and the experimental measurement value, and then obtaining the parameters a1, b1, a2, b2, and c; Substitute the dielectric loss value of the cable at room temperature measured experimentally into the formula to calculate To simplify the extrapolation calculation, take the logarithm of both sides of the above formula to change it to the following: lny=lna1+b1T daughter=lna2+b2T+cT 2 Therefore, the temperature value from the temperature sensor and the dielectric loss value of the dielectric response calculation module can be extrapolated to the dielectric loss value at any temperature in the temperature range of 20°C-130°C using the following formula. For the temperature range of 20℃-70℃: ln tanδ t =b1·(t-25)+ln tanδ 25 +ln tanδ Δ For the temperature range 90℃-130℃: ln tanδ t =b1·(t-25)+c·(t d -25 2 )+ln tanδ 25 +ln tanδ Δ。 6. A cable insulation wide temperature range dielectric loss detection method according to claim 4, characterized in that: The temperature value T measured by the temperature sensor 修正 Obtained by the built-in adaptive Kalman filter algorithm of the hot end measurement module and the cold end compensation module: Define the state vector X k =[T 测量 , α] T , observation vector Z k =T cold end, update the Kalman gain and state estimation through the recursive formula: X k =X k-1 +K k (Z k -HX k-1 ), where: K k is the Kalman gain, which determines the weight of the observation value on the state correction, H=[1,ΔT 环境 ], output the corrected temperature value T 修正 =X k (1), the compensation coefficient α is calibrated by the following formula: in, The data acquisition system continuously compares and corrects the temperature T 修正 With reference temperature T 冷端 , if the residual exceeds the limit for 5 consecutive times, that is, |T 修正 -T 修正 |>0.5℃, the adaptive algorithm is triggered to update α.

7. The method for detecting dielectric loss of cable insulation over a wide temperature range according to claim 5, wherein: The temperature correction module calculates the specific temperature value desired by the user transmitted from the human-computer interaction interface and the dielectric loss value calculated by the main control processor to obtain the dielectric loss value at the specific temperature, sets the temperature weight coefficient in the temperature transition zone, and adjusts the weight coefficient using an adaptive method based on fuzzy logic. The weight is dynamically adjusted through the membership function and rule base to achieve a nonlinear transition that is more in line with the material characteristics.

Images