Computer monitoring system abnormity monitoring method for large-scale electric power construction project

Abstract

The invention relates to the technical field of electric power facility anomaly monitoring, in particular to a computer monitoring system anomaly monitoring method for large-scale electric power construction engineering, which comprises the following steps of: acquiring an insulator leakage current signal, a temperature signal and an ultrasonic signal, and calculating a corresponding energy gradient; calibrating the correlation coefficient of the sensor based on the physical characteristics of an insulator material, and calculating the dynamic weight according to the energy gradient, thereby achieving the dynamic weighting of a fault sensitive signal; generating a fusion reference signal through weighted summation; constructing a dynamic threshold by combining the thermal physical parameters of the equipment, the real-time signal characteristics and the environmental parameters; calculating a signal difference quantity between the original signal and the fusion signal, and if the signal difference quantity exceeds a dynamic threshold value, triggering abnormal early warning; noise influence on a single sensor is suppressed, the signal-to-noise ratio of a fusion signal is effectively improved, the problems of abnormal missing report and false report easily caused by a fixed threshold value are avoided, and the insulator abnormal detection precision is improved.

Description

Technical Field

[0001] The present application relates to the technical field of abnormality monitoring of electric power facilities, and in particular to an abnormality monitoring method of a computer monitoring system for large-scale electric power construction projects. Background Art

[0002] Large-scale power construction projects are characterized by high voltage levels, dense equipment density, and harsh environmental conditions. Failures in critical equipment can trigger cascading failures or even catastrophic consequences. Insulators, core components in power systems that provide both insulation and mechanical support, are widely used in transmission lines, substations, and generator stator windings. Their condition directly impacts system reliability. Common insulator failures include reduced insulation resistance due to contaminant deposits, cracks and damage caused by mechanical stress, material degradation due to corona discharge, and flashover breakdown caused by lightning strikes or overvoltage. Failure to promptly detect these potential hazards can lead to line tripping, equipment damage, and even widespread grid failure.

[0003] Therefore, online insulator condition monitoring technology is crucial for large-scale power construction projects. However, the power-frequency magnetic field generated by high-voltage equipment during operation can interfere with sensor signals, introducing random or periodic noise into the detection signal and affecting the accurate assessment of insulator condition. Traditional anomaly monitoring methods typically use single sensor data and a fixed threshold for anomaly detection. This makes it difficult to fully capture the characteristics of insulator anomalies. Single sensor data is also susceptible to noise, which can easily lead to missed anomalies and false alarms. Summary of the Invention

[0004] In order to solve the above technical problems, the present application provides a computer monitoring system abnormality monitoring method for large-scale power construction projects to solve the existing problems.

[0005] The computer monitoring system anomaly monitoring method for large-scale power construction projects in this application adopts the following technical solutions: One embodiment of the present application provides a method for abnormality monitoring of a computer monitoring system for a large-scale electric power construction project, the method comprising the following steps: Collect various sensor signals of insulators at all times, including leakage current signals, temperature signals and ultrasonic signals; collect humidity signals of the environment at all times; Set a time window at each moment, and construct the signal energy gradient of each sensor signal at each moment based on the change characteristics of each sensor signal in the time window at each moment; Conduct multiple discharge tests on insulators, collect the discharge current of the insulators during the discharge tests, and construct a fault severity sequence for each discharge test. Based on the correlation between the time series of various sensor signals and the fault severity sequence, combined with the signal energy gradient, a dynamic weight coefficient for each type of sensor signal at each moment is constructed. Based on the leakage current signal, temperature signal and ultrasonic signal at each moment, as well as the corresponding dynamic weight coefficient, a fusion reference signal at each moment is constructed; Based on the material properties of the insulator and the temperature change characteristics at each moment, combined with the signal values of multiple sensors at each moment, a dynamic threshold value at each moment is constructed; Compensating the dynamic threshold based on the ambient humidity signal and the voltage level of the insulator at each moment, and constructing a compensated dynamic threshold at each moment;

[0006] A signal difference amount at each moment is constructed based on the difference between each type of sensor signal and the fusion reference signal at each moment; and insulator abnormality detection is performed based on the signal difference amount and combined with the compensated dynamic threshold.

[0007] In one embodiment, the time window at each moment is a time period of a preset length before each moment.

[0008] In one embodiment, the expression of the signal energy gradient of each type of sensor signal is: , where is the signal energy gradient of the temperature at moment t; 、 Respectively represent the temperature signals at the first and last moments in the time window at moment t; Indicates the length of the time window; Based on the leakage current signal and ultrasonic signal in each time window, the signal energy gradient of the leakage current and the signal energy gradient of the ultrasonic wave at each moment are calculated using the same calculation method as the signal energy gradient of the temperature at each moment.

[0009] In one embodiment, the process of acquiring the fault severity sequence is as follows: normalizing the time series of the insulator discharge current, and the obtained sequence is used as the fault severity sequence.

[0010] In one embodiment, the process of obtaining the dynamic weight coefficient of each type of sensor signal at each moment is as follows: Calculate the correlation between the time series of the temperature signal and each fault severity series, recorded as the first correlation; use the average of all the first correlations of the temperature signal as the correlation coefficient of the temperature signal; use the normalized value of the product of the correlation coefficient of the temperature signal and the signal energy gradient of the temperature at each moment as the dynamic weight coefficient of the temperature signal at each moment;

[0011] Based on the time series of the leakage current signal and the ultrasonic signal and the corresponding signal energy gradient at each moment, the dynamic weight coefficients of the leakage current signal and the ultrasonic signal at each moment are obtained using the same acquisition method as the dynamic weight coefficient of the temperature signal at each moment.

[0012] In one embodiment, the expression of the fusion reference signal at each moment is: , where is the fusion reference signal at time t; 、 、 are the dynamic weight coefficients of leakage current, temperature and ultrasonic signal at time t respectively; 、 、 are the leakage current, temperature and ultrasonic signal at the tth moment respectively.

[0013] In one embodiment, the expression of the dynamic threshold at each moment is: , where is the dynamic threshold at time t; K represents the thermal conductivity of the insulator; represents the specific heat capacity of the insulator; represents the temperature change rate at time t; is the leakage current signal at the tth moment; is the ultrasonic signal at the tth moment; is the normalization function; The temperature change rate at the tth moment is the ratio of the difference between the temperature at the tth moment and the temperature at the previous moment to the time interval between the two moments.

[0014] In one embodiment, the expression of the compensated dynamic threshold at each moment is: Where, represents the dynamic threshold after compensation at time t; is the dynamic threshold at time t; is the humidity compensation coefficient; represents the humidity of the environment at time t; Indicates the preset adjustment coefficient; Indicates the rated voltage of the insulator; Represents the logarithmic function with e as a real number.

[0015] In one embodiment, the expression of the signal difference at each moment is: , where is the signal difference at time t; 、 、 are the leakage current, temperature and ultrasonic signal at the tth moment respectively; is the fusion reference signal at the tth moment.

[0016] In one embodiment, the insulator abnormality detection is performed based on the signal difference and the compensated dynamic threshold, specifically: If the signal difference at each moment is greater than the corresponding compensated dynamic threshold, then the insulator is abnormal; otherwise, the insulator is not abnormal.

[0017] This application has at least the following beneficial effects: The present application collects various sensor signals from insulators at each moment, including leakage current signals, temperature signals, and ultrasonic signals. Based on the changing characteristics of each sensor signal within a time window at each moment, a signal energy gradient is constructed for each sensor signal at each moment. Combined with the correlation between the time series of each sensor signal and the insulator discharge current series, a dynamic weight coefficient is constructed for each sensor signal type at each moment, dynamically reflecting the differences in sensitivity of each sensor to fault characteristics. A fused reference signal is constructed at each moment based on the leakage current, temperature, and ultrasonic signals, as well as the corresponding dynamic weight coefficients. Using the signal energy gradients of multiple sensors as weights, the fused signal is generated by leveraging the physical complementarity of different signals, suppressing the effects of noise on a single sensor, effectively improving the signal-to-noise ratio of the fused signal, and achieving multi-dimensional state perception. A dynamic threshold is constructed based on the insulator's material properties, real-time signal characteristics, and environmental influences. A signal difference value is constructed at each moment based on the difference between each sensor signal type and the fused reference signal. The signal difference value is then compared with the dynamic threshold to determine whether the insulator is abnormal. This avoids the problem of missed or false alarms caused by fixed thresholds, reduces computational complexity, and improves the accuracy of insulator anomaly detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions and advantages of the embodiments of the present application or the prior art, the following is a brief introduction to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0019] Figure 1 A flowchart of the abnormality monitoring method for a computer monitoring system for large-scale power construction projects provided in this application;

[0020] Figure 2 Schematic diagram of the process of obtaining the dynamic weight coefficient of the temperature signal. DETAILED DESCRIPTION

[0021] In order to further illustrate the technical means and effects adopted by this application to achieve the predetermined invention objectives, the following, in conjunction with the accompanying drawings and preferred embodiments, describes in detail the specific implementation, structure, features and effects of the computer monitoring system abnormality monitoring method for large-scale power construction projects proposed in this application. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics of one or more embodiments may be combined in any suitable form.

[0022] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.

[0023] The specific scheme of the abnormality monitoring method of the computer monitoring system for large-scale power construction projects provided by this application is described in detail below with reference to the accompanying drawings.

[0024] An embodiment of the present application provides a method for monitoring abnormalities in a computer monitoring system for large-scale power construction projects.

[0025] Specifically, the following abnormal monitoring method for computer monitoring systems for large-scale power construction projects is provided. Figure 1 , the method comprises the following steps: Step S1 , collecting various sensor signals of the insulator at each moment, including leakage current signal, temperature signal and ultrasonic signal; collecting humidity signal of the environment at each moment.

[0026] Insulator anomalies are often accompanied by simultaneous changes in electrical parameters (leakage current), thermal characteristics (local temperature rise), and mechanical vibration (discharge noise). A single sensor, such as one that only measures leakage current, cannot fully capture the fault characteristics and is susceptible to environmental interference. This application uses multi-dimensional physical signals to collaboratively monitor insulator status, addressing the problem of single-sensor information being incomplete and susceptible to interference.

[0027] A set of sensors is installed on both the conductor side (high potential) and the ground side (low potential) of the insulator string to monitor the overall potential distribution of the insulator string. Each set of sensors includes a leakage current sensor, a temperature sensor, an ultrasonic sensor, and a humidity sensor.

[0028] This application uses a leakage current sensor, a temperature sensor, and an ultrasonic sensor to collect leakage current signals, temperature signals, and ultrasonic signals of the insulator, respectively. Specifically: (1) Leakage current refers to the current generated by the tiny conductive path on the surface of an insulator or the internal insulating medium under the operating voltage. When the shed is damaged, the coating is aged (such as embrittled silicone rubber on composite insulators), or the dirt is unevenly distributed on the surface of the insulator, local electric field concentration (such as at the edge of the shed or the tip of the crack) will induce corona discharge or surface micro-discharge. These discharge phenomena will ionize the air to generate ion flow, forming a pulsed leakage current. Although its average amplitude is small, it contains the early characteristics of insulation degradation.

[0029] To collect leakage current signals, this application uses a through-hole Rogowski coil (bandwidth 0-100kHz) for this purpose, with a data collection interval of 1ms. In other embodiments of this application, implementers can adjust the data collection interval based on their specific circumstances. It should be noted that implementers may also use other leakage current sensors for leakage current signal collection, and this application does not impose any specific limitations.

[0030] The average value of the leakage current signals collected by the leakage current sensors on the conductor side and the ground side at each moment is used as the leakage current signal of the insulator at each moment.

[0031] A second-order Butterworth low-pass filter (cut-off frequency 20kHz) is used to filter the leakage current signal to suppress the high-frequency noise generated by the ultra-high voltage corona; and a signal amplifier is used to amplify the filtered leakage current signal. In this application, the gain effect of the amplifier is 100 times, which improves the accuracy of small current detection.

[0032] (2) There is a strong correlation between the change of temperature signal and insulator abnormality. When there is slight contamination on the surface of the insulator or internal cracks, the leakage current may not increase significantly, but the Joule heat generated by the discharge will cause the local temperature to rise. The leakage current reflects the discharge intensity, and the temperature reflects the discharge duration effect.

[0033] For temperature signal acquisition, this application uses a single-point infrared temperature measurement module (Melexis MLX90614), with a data acquisition interval of 1 second. In other embodiments of this application, implementers can adjust the data acquisition interval based on their specific circumstances. It should be noted that implementers can also use other temperature sensors for temperature signal acquisition based on their specific circumstances, and this application does not impose any specific restrictions.

[0034] The average value of the temperature signals collected by the temperature sensors on the conductor side and the grounding side at each moment is used as the temperature signal of the insulator at each moment.

[0035] (3) Ultrasonic signals are a key characteristic of insulator discharge anomalies. Local discharges caused by surface contamination, cracks, or internal defects in insulators generate broadband acoustic signals. The ultrasonic components between 20kHz and 100kHz, especially those around 40kHz, are important carriers of discharge energy. Unlike ambient noise, such as wind, rain, and bird activity, which concentrate their energy below 20kHz, the ultrasonic signal amplitude is positively correlated with the discharge intensity and can reflect the discharge initiation and development process.

[0036] For ultrasonic signal acquisition, this application uses a narrowband piezoelectric ceramic sensor (center frequency 40kHz), with a 50ms data acquisition interval. In other embodiments of this application, implementers can adjust the ultrasonic signal acquisition interval based on their specific circumstances. It should be noted that implementers can use other ultrasonic sensors for ultrasonic signal acquisition based on their specific circumstances, and this application does not impose any specific restrictions.

[0037] The average value of the ultrasonic signals collected by the ultrasonic sensors on the conductor side and the ground side at each moment is taken as the ultrasonic signal of the insulator at each moment.

[0038] Ultrasonic signals are key to detecting insulator discharge anomalies. They are filtered using a bandpass filter. In this application, the filter's passband is 35-45 kHz, eliminating the effects of wind and rain noise and power frequency harmonics. An automatic gain control (AGC) circuit is used to automatically control the ultrasonic signal's gain, dynamically adapting to changes in the discharge signal's amplitude.

[0039] In addition, a humidity sensor collects humidity signals from the insulator's surroundings, with a data collection interval of 1 second. Although insulator anomalies don't cause humidity changes, ambient humidity can affect insulator resistance, causing errors.

[0040] Since the acquisition frequencies of various sensors are different, the GPS module is used to output a 1PPS pulse signal for time synchronization. The data points are marked by timestamps, and the linear interpolation method is used to synchronize the low-frequency signal to the high-frequency time series to ensure the time alignment of multiple signals.

[0041] In strong electromagnetic interference environments, leakage current sensors are prone to high-frequency oscillation or baseline drift due to electromagnetic induction coupling. The back-end amplifier circuit of ultrasonic sensors may be affected by electric field coupling, reducing the signal-to-noise ratio. The readout circuit and analog-to-digital conversion process of temperature sensors may suffer from magnetic field interference, reducing temperature measurement accuracy. However, due to differences in interference coupling paths, sensitive frequency bands, and physical principles among different sensors, strong electromagnetic interference is unlikely to have a serious impact on all sensors simultaneously. Multi-sensor fusion technology can compensate for interfered channels with undistorted signals, significantly improving monitoring reliability in complex electromagnetic environments and achieving multi-dimensional signal complementarity and anti-interference capabilities.

[0042] To facilitate subsequent processing, this application normalizes the leakage current signal, temperature signal, ultrasonic signal and humidity signal at all times using the Z-score normalization algorithm to eliminate the dimension effect.

[0043] Step S2: setting a time window at each moment, and constructing a signal energy gradient of each sensor signal at each moment based on the change characteristics of each sensor signal in the time window at each moment.

[0044] In insulator anomaly monitoring, the signal's mutation characteristics are key to fault early warning. Partial discharge can cause a transient increase in leakage current and a sudden rise in temperature in the discharge area. These changes manifest as a rapid increase in signal power. Traditional methods directly use the signal's absolute value or static statistics, such as the effective and average leakage current values, for anomaly monitoring. However, these methods struggle to capture short-term power changes and are susceptible to interference from slowly changing environmental noise.

[0045] Taking the temperature signal as an example, the difference between the temperature signal at the start and end of the time window is used to calculate the signal energy gradient of the temperature at each moment, reflecting the real-time change rate of the temperature signal power. The expression is: , where is the signal energy gradient of the temperature at moment t; 、 Respectively represent the temperature signals at the first and last moments in the time window at moment t in the temperature signal time series; Indicates the length of the time window. Preferably, in the embodiment of the present application, The value of is set to 10ms, covering the typical discharge pulse duration. As other embodiments of this application, the implementer can set it according to the actual situation. The time window of each moment in this application is a time period of a preset time length before each moment. In the embodiment of this application, the preset time length is .

[0046] Based on the leakage current signal and ultrasonic signal in each time window, the signal energy gradient of the leakage current and the signal energy gradient of the ultrasonic wave at each moment are calculated using the same calculation method as the signal energy gradient of the temperature at each moment.

[0047] This application uses differential approximation to replace integral differentiation to reduce computing power requirements. The energy gradient is essentially the time rate of change of signal power, reflecting the increase in signal energy per unit time. When a fault occurs, the energy gradient of the signal will have a significant positive peak, while it will maintain low fluctuations during normal operation. The larger the energy gradient, the more likely the signal is from a fault feature rather than noise, and a higher weight will be assigned in subsequent steps to enhance the fusion signal's ability to characterize the fault. To avoid the signal being affected by load fluctuations and climatic conditions and exhibiting non-stationary characteristics, the energy gradient can dynamically track the effective changes in the signal and avoid the hysteresis of a fixed threshold.

[0048] Step S3: Perform multiple discharge tests on the insulator, collect the discharge current of the insulator during the discharge test, and construct a fault severity sequence for each discharge test; based on the correlation between the time series of various sensor signals and the fault severity sequence, combined with the signal energy gradient, construct a dynamic weight coefficient for each type of sensor signal at each moment.

[0049] In multi-sensor fusion monitoring, different sensors naturally have different sensitivities to the same fault. Traditional solutions use fixed weights, which cannot be dynamically adjusted based on real-time operating conditions, potentially causing critical signals to be overwhelmed by noise.

[0050] First, the correlation coefficients between various sensor signals and insulator faults are analyzed. Specifically, the following steps are performed: a discharge test is performed on the insulator. In this application, the insulator is subjected to a power frequency withstand voltage test. The discharge current of the insulator is collected at each moment during the discharge test. The time series of the discharge current is normalized, and the resulting sequence is used as the fault severity sequence. Multiple discharge tests are performed on the insulator to obtain the fault severity sequence for each discharge test. The Pearson correlation coefficient between the time series of the temperature signal and each fault severity sequence is calculated, denoted as the first correlation. The average of all first correlations of the temperature signal is used as the correlation coefficient of the temperature signal. Furthermore, based on the time series of the leakage current signal and the ultrasonic signal, the correlation coefficients of the leakage current signal and the ultrasonic signal are obtained using the same acquisition method as the correlation coefficient of the temperature signal. This correlation coefficient represents the degree of correlation between the sensor signal and the physical mechanism of the insulator fault. Preferably, in this embodiment of the application, the number of discharge tests is set to 3. As other embodiments of the application, the implementer can set the number of discharge tests according to actual conditions.

[0051] Among them, the Pearson correlation coefficient is a well-known technology, and the specific process will not be repeated here. It should be noted that for the correlation between the time series of the temperature signal and the fault severity series, this application only provides a correlation calculation method. There are many existing correlation calculation methods. Implementers can also use other correlation algorithms to calculate the correlation between the time series of the temperature signal and the fault severity series. This application does not make specific restrictions.

[0052] Furthermore, the product of the signal energy gradient of each type of sensor signal at each moment and the correlation coefficient is calculated and recorded as a first product; and the normalized value of the first product is used as the dynamic weight coefficient of each type of sensor signal at each moment.

[0053] Preferably, in an embodiment of the present application, the normalization method of the first product is: calculate the sum of the first products of the leakage current, temperature and ultrasonic signal at any moment, recorded as the first sum; and use the ratio of the first product of each type of sensor at any moment to the first sum as the dynamic weight coefficient of each type of sensor at any moment.

[0054] The inherent physical characteristics of the sensor, namely its sensitivity to specific faults, are taken into account, while also taking into account the real-time changing characteristics of the signal, namely reflecting the current fault intensity through the energy gradient. Taking the leakage current sensor as an example, the dynamic weight coefficient is calculated as follows: Through adaptive adjustment of sensor weights, when the energy gradient of a sensor signal suddenly increases, its weight increases simultaneously, ensuring that the fused signal prioritizes the fault characteristics of that sensor signal. Sensors with low energy gradients and weak physical correlations are assigned low weights to reduce the impact of noise. This addresses the problem of traditional fixed weights being unable to adapt to changing operating conditions and having weak anti-interference capabilities. By combining the inherent characteristics of sensors with real-time fault characteristics, intelligent weighting of multi-sensor data is achieved, enabling the fused signal to dynamically focus on the most critical monitoring indicators, significantly improving fault identification capabilities in complex environments.

[0055] Step S4: constructing a fusion reference signal at each moment based on the leakage current signal, temperature signal and ultrasonic signal at each moment, and the corresponding dynamic weight coefficient.

[0056] The fusion reference signal at each moment is generated according to different sensor signals and dynamic weights. The expression is: Where, is the fusion reference signal at time t; 、 、 are the dynamic weight coefficients of leakage current, temperature and ultrasonic signal at time t respectively; 、 、 are the leakage current, temperature, and ultrasonic signal at time t, respectively. A fusion reference signal is generated through weighted summation to suppress the noise interference of a single sensor.

[0057] Step S5 : constructing a dynamic threshold value at each moment based on the material properties of the insulator and the temperature variation characteristics at each moment, combined with the signal values of multiple sensors at each moment.

[0058] Existing technologies often use a single sensor to collect data and set fixed thresholds, failing to account for individual device variations. Insulators made of different materials have varying thermal capacities and insulation properties, making fixed thresholds inaccurate. Increased humidity can cause insulator surface resistivity to decrease, making fixed thresholds prone to false alarms in humid weather.

[0059] This application sets dynamic thresholds based on physical device parameters and real-time environmental data, allowing the thresholds to dynamically adjust with individual device characteristics, load changes, and environmental conditions, solving the problem that traditional fixed thresholds are difficult to adapt to environmental changes. The expression for the dynamic threshold is: Where, is the dynamic threshold at time t; K represents the thermal conductivity of the insulator; represents the specific heat capacity of the insulator; represents the temperature change rate at time t; is the leakage current signal at the tth moment; is the ultrasonic signal at the tth moment; is the normalization function.

[0060] It should be noted that the normalization function in the above formula can be , is an exponential function with a natural constant as the base; the implementer may also use other methods to Normalization is performed, and this application does not impose specific restrictions. K represents the thermal conductivity of the insulator material, reflecting the heat flow per unit time and per unit area when the temperature gradient is 1K. It is directly related to the material type and can be directly obtained by querying the material manual. For example, the thermal conductivity coefficient of ceramic insulators is , the thermal conductivity of the composite insulator is It quantifies the heat dissipation efficiency of the equipment. The smaller the thermal conductivity (such as composite insulators), the higher the internal power loss corresponding to the same temperature rise, the higher the fault sensitivity, and the smaller the dynamic threshold should be. The larger the thermal conductivity, the larger the dynamic threshold.

[0061] Characterizes the heat storage capacity of the material, reflecting the amount of heat required to raise the temperature of the unit mass of the material by 1K, and is related to the material and structure of the insulator. The specific heat capacity of the insulator can be obtained by querying the material manual, such as ceramic insulators. , composite insulators Combining temperature rise with calculations of heat storage changes, materials with large specific heat capacity (such as composite insulators) experience slow temperature changes and require greater energy to trigger the threshold, thus avoiding misjudgments. The greater the specific heat capacity, the greater the dynamic threshold.

[0062] This method represents the change in insulator surface temperature per unit time. This application calculates the temperature difference between time t and the previous time, and uses the ratio of this difference to the time interval between the two times as the temperature change rate at time t. This method captures rapid temperature rises caused by partial discharge or poor contact, avoiding misjudgment of threshold values.

[0063] Step S6: compensating the dynamic threshold based on the ambient humidity signal and the voltage level of the insulator at each moment, and constructing a compensated dynamic threshold at each moment.

[0064] In high humidity environments, moisture in the air forms a film on the insulator surface. In particular, dirt particles absorb moisture, significantly reducing the surface resistivity and increasing leakage current. Traditional fixed thresholds cannot adapt to these normal signal fluctuations caused by humidity changes. This can lead to misinterpretations of insulation degradation in humid weather, such as misinterpreting increased leakage current caused by moisture as a contamination discharge. In dry environments, early faults can also be missed (for example, if the thresholds don't dynamically adjust as humidity decreases, miniscule leakage currents may go undetected).

[0065] Based on dielectric physics theory, the effect of humidity on the conductive properties of the insulator surface can be quantified by an exponential model. This application constructs a humidity compensation coefficient. , the expression is: , where Represents a preset adjustment coefficient, which is used to quantify the proportional relationship between the voltage level and the humidity sensitivity, and has a value range of 0.03 to 0.07. Preferably, in the embodiment of the present application, The value of is set to 0.05; Indicates the rated voltage of the insulator; Represents a logarithmic function with e as the real number. Since higher voltage levels increase the electric field strength on the insulator surface and the impact of humidity changes on the surface conductivity is more significant, a logarithmic function is used to increase the compensation coefficient as the rated voltage increases. This dynamically adjusts the threshold's response to humidity, ensuring accurate environmental compensation for devices at all voltage levels.

[0066] Furthermore, the dynamic threshold after compensation at each moment is constructed, and the expression is: , where represents the dynamic threshold after compensation at time t, Indicates the humidity of the environment at time t. The larger the value, the larger the sensor signal may fluctuate normally over a larger range due to environmental factors under the current working condition. The signal is allowed to fluctuate within a larger range without triggering an abnormal judgment, thus avoiding misjudgment due to environmental noise. On the contrary, the smaller the value, the more sensitive the sensor signal is to signal deviation.

[0067] Step S7: constructing a signal difference amount at each moment based on the difference between each type of sensor signal and the fusion reference signal at each moment; performing insulator abnormality detection based on the signal difference amount and the compensated dynamic threshold.

[0068] Based on the difference between each type of sensor signal and the fusion reference signal at each moment, the signal difference between the original signal and the fusion reference signal at each moment is calculated. Preferably, in the embodiment of the present application, the expression of the signal difference at each moment can be: , where is the signal difference between the original signal and the fused reference signal at time t; 、 、 are the leakage current, temperature and ultrasonic signal at the tth moment respectively; is the fusion reference signal at the tth moment.

[0069] In other embodiments of the present application, the signal difference amount at each moment may be the average value of the absolute value of the difference between the leakage current, temperature, and ultrasonic signal at each moment and the fusion reference signal.

[0070] The fused reference signal is a comprehensive representation of multiple sensors. The difference between the original signal and the fused reference signal indicates the degree of deviation between the original signal and the fused reference signal. When a sensor becomes abnormal due to failure or interference, the difference between the original and reference signals will increase significantly, while other undisturbed sensors can still ensure the reliability of the reference signal. The signal difference is compared with the threshold to determine whether the insulator is abnormal.

[0071] Multi-sensor signals maintain a stable collaborative relationship due to physical correlation, and the signal difference is at a low level; when an insulator has an abnormality, one or more sensor signals deviate from the baseline, and the signal difference exceeds the threshold.

[0072] If the difference between the original signal and the fused reference signal at each moment is greater than the compensated dynamic threshold at that moment, it indicates that the insulator is abnormal, and the relevant maintenance personnel are notified to inspect and repair the insulator. If the difference between the original signal and the fused reference signal at each moment is less than or equal to the compensated dynamic threshold at that moment, it indicates that the insulator is normal.

[0073] The schematic diagram of the process of obtaining the dynamic weight coefficient of the temperature signal is as follows: Figure 2 shown.

[0074] In summary, the embodiments of the present application collect various sensor signals from the insulator at each moment, including leakage current signals, temperature signals, and ultrasonic signals; construct signal energy gradients for each type of sensor signal at each moment based on the change characteristics of each type of sensor signal in the time window at each moment; and construct dynamic weight coefficients for each type of sensor signal at each moment based on the correlation between the time series of each type of sensor signal and the insulator discharge current series, dynamically reflecting the difference in sensitivity of each sensor to fault characteristics; construct a fused reference signal at each moment based on the leakage current signals, temperature signals, and ultrasonic signals at each moment, and the corresponding dynamic weight coefficients; use the signal energy gradients of multiple sensors as weights, and utilize the physical complementarity of different signals to generate a fused signal, suppressing the noise impact of a single sensor, effectively improving the signal-to-noise ratio of the fused signal, and achieving multi-dimensional state perception; construct a dynamic threshold based on the material properties, real-time signal characteristics, and environmental impact of the insulator; construct a signal difference value at each moment based on the difference between each type of sensor signal and the fused reference signal at each moment; and determine whether the insulator has an abnormality based on the comparison of the signal difference value with the dynamic threshold. This avoids the problem of anomaly omissions and false alarms that are easily caused by fixed thresholds, reduces computational complexity, and improves the accuracy of insulator anomaly detection.

[0075] It should be noted that the order in which the embodiments of the present application are presented is for illustrative purposes only and does not necessarily represent the superiority or inferiority of the embodiments. Furthermore, the above descriptions are of specific embodiments of the present application. Furthermore, the processes depicted in the accompanying drawings do not necessarily require the specific order or sequential sequence shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0076] The various embodiments in this application are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on the differences from other embodiments.

[0077] The above-described embodiments are only used to illustrate the technical solutions of the present application, and not to limit them. Modifications to the technical solutions described in the aforementioned embodiments, or equivalent replacements of some of the technical features therein, do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application, and should all be included in the scope of protection of the present application.

Claims

1. A computer monitoring system abnormality monitoring method for large-scale power construction projects, characterized in that: The method comprises the following steps: Collect various sensor signals of insulators at all times, including leakage current signals, temperature signals and ultrasonic signals; collect humidity signals of the environment at all times; Set a time window at each moment, and construct the signal energy gradient of each sensor signal at each moment based on the change characteristics of each sensor signal in the time window at each moment; Conduct multiple discharge tests on insulators, collect the discharge current of the insulators during the discharge tests, and construct a fault severity sequence for each discharge test. Based on the correlation between the time series of various sensor signals and the fault severity sequence, combined with the signal energy gradient, a dynamic weight coefficient for each type of sensor signal at each moment is constructed. Based on the leakage current signal, temperature signal and ultrasonic signal at each moment, as well as the corresponding dynamic weight coefficient, a fusion reference signal at each moment is constructed; Based on the material properties of the insulator and the temperature change characteristics at each moment, combined with the signal values of multiple sensors at each moment, a dynamic threshold value at each moment is constructed; Compensating the dynamic threshold based on the ambient humidity signal and the voltage level of the insulator at each moment, and constructing a compensated dynamic threshold at each moment; A signal difference amount at each moment is constructed based on the difference between each type of sensor signal and the fusion reference signal at each moment; and insulator abnormality detection is performed based on the signal difference amount and combined with the compensated dynamic threshold.

2. The computer monitoring system abnormality monitoring method for large-scale electric power construction projects according to claim 1, characterized in that: The time window of each moment is a time period of a preset time length before each moment.

3. The computer monitoring system abnormality monitoring method for large-scale electric power construction projects according to claim 1, characterized in that: The expression of the signal energy gradient of the various sensor signals is: , where is the signal energy gradient of the temperature at moment t; 、 Respectively represent the temperature signals at the first and last moments in the time window at moment t; Indicates the length of the time window; Based on the leakage current signal and ultrasonic signal in each time window, the signal energy gradient of the leakage current and the signal energy gradient of the ultrasonic wave at each moment are calculated using the same calculation method as the signal energy gradient of the temperature at each moment.

4. The computer monitoring system abnormality monitoring method for large-scale electric power construction projects according to claim 1, characterized in that: The process of obtaining the fault severity sequence is as follows: normalizing the time series of the insulator discharge current, and the obtained sequence is used as the fault severity sequence.

5. The computer monitoring system abnormality monitoring method for large-scale electric power construction projects according to claim 1, characterized in that: The process of obtaining the dynamic weight coefficient of each type of sensor signal at each moment is as follows: Calculate the correlation between the time series of the temperature signal and each fault severity series, recorded as the first correlation; use the average of all the first correlations of the temperature signal as the correlation coefficient of the temperature signal; use the normalized value of the product of the correlation coefficient of the temperature signal and the signal energy gradient of the temperature at each moment as the dynamic weight coefficient of the temperature signal at each moment; Based on the time series of the leakage current signal and the ultrasonic signal and the corresponding signal energy gradient at each moment, the dynamic weight coefficients of the leakage current signal and the ultrasonic signal at each moment are obtained using the same acquisition method as the dynamic weight coefficient of the temperature signal at each moment.

6. The computer monitoring system abnormality monitoring method for large-scale electric power construction projects according to claim 1, characterized in that: The expression of the fusion reference signal at each moment is: , where is the fusion reference signal at time t; 、 、 are the dynamic weight coefficients of leakage current, temperature and ultrasonic signal at time t respectively; 、 、 are the leakage current, temperature and ultrasonic signal at the tth moment respectively.

7. The method for abnormality monitoring of a computer monitoring system for a large-scale electric power construction project according to claim 1, characterized in that: The expression of the dynamic threshold at each moment is: , where is the dynamic threshold at time t; K represents the thermal conductivity of the insulator; represents the specific heat capacity of the insulator; represents the temperature change rate at time t; is the leakage current signal at the tth moment; is the ultrasonic signal at the tth moment; is the normalization function; The temperature change rate at the tth moment is the ratio of the difference between the temperature at the tth moment and the temperature at the previous moment to the time interval between the two moments.

8. The method for abnormality monitoring of a computer monitoring system for a large-scale electric power construction project according to claim 1, characterized in that: The expression of the compensated dynamic threshold at each moment is: Where, represents the dynamic threshold after compensation at time t; is the dynamic threshold at time t; is the humidity compensation coefficient; represents the humidity of the environment at time t; Indicates the preset adjustment coefficient; Indicates the rated voltage of the insulator; Represents the logarithmic function with e as a real number.

9. The computer monitoring system abnormality monitoring method for large-scale electric power construction projects according to claim 1, characterized in that: The expression of the signal difference at each moment is: , where is the signal difference at time t; 、 、 are the leakage current, temperature and ultrasonic signal at the tth moment respectively; is the fusion reference signal at the tth moment.

10. The computer monitoring system abnormality monitoring method for large-scale electric power construction projects according to claim 1, characterized in that: The insulator abnormality detection is performed based on the signal difference and in combination with the compensated dynamic threshold, specifically: If the signal difference at each moment is greater than the corresponding compensated dynamic threshold, then the insulator is abnormal; otherwise, the insulator is not abnormal.

Images