An unattended substation intelligent monitoring method and system
By acquiring infrared radiation grayscale images and load current in real time in unattended substations, and using gradient operators and thermal field coherence analysis, a temperature rise inertia index is constructed. Combined with time series monotonicity discrimination, the problem of high false alarm rate of traditional monitoring methods in complex environments is solved, and early and accurate warning of potential equipment failures is achieved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- JINAN SUN K ELECTRIC POWER EQUIP CO LTD
- Filing Date
- 2026-03-18
- Publication Date
- 2026-05-22
AI Technical Summary
In unattended substations, traditional monitoring methods are susceptible to dynamic load fluctuations and weather interference in complex environments, resulting in inaccurate assessment of equipment operating status, high false alarm rates, and an inability to effectively identify hidden thermal defects in sub-healthy states.
By acquiring real-time infrared radiation grayscale images and load current of the equipment area, and using gradient operators and thermal field coherence analysis, a temperature rise inertia index is constructed. Combined with a time-series monotonicity discrimination mechanism, load fluctuations and actual equipment heating are decoupled, accurately capturing minute initial heating defects, eliminating environmental reflection pseudo-hot spots, and achieving early warning.
It significantly improves the sensitivity and accuracy of early warning, reduces the false alarm rate, ensures the stable operation of the power grid, and can accurately identify potential equipment failures under load fluctuations and environmental interference.
Smart Images
Figure CN121863686B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of power system monitoring technology. More specifically, this invention relates to an intelligent monitoring method and system for unattended substations. Background Technology
[0002] In the modern power industry, unmanned substations are an important means to realize the intelligence of the power grid and reduce operating costs. Substations contain a variety of primary equipment, such as disconnect switches, circuit breakers, current transformers and surge arresters, and their operational safety is directly related to the stability of the power grid system. In order to achieve all-weather monitoring, substations are usually deployed with dual-light monitoring units that integrate infrared thermal imaging and visible light high-definition imaging. This technology can give full play to the dual advantages of thermal imaging temperature analysis and visible light high-resolution observation, and capture the thermal radiation distribution on the surface of equipment in real time.
[0003] However, in actual operation scenarios, outdoor substations are severely affected by environmental heat sources. Since most substations are located in environments with strong sunlight or drastic temperature changes, the surface temperature of the equipment is not only affected by Joule heating generated by the internal load current, but also by the coupling interference of external meteorological factors such as solar radiation, background reflection, and wind speed heat dissipation. Traditional power system monitoring methods mostly use fixed temperature thresholds, which cannot decouple the actual equipment fault temperature rise from the complex environmental background. Especially during periods of large load fluctuations, the tiny transient heat accumulation caused by the deterioration of contact resistance is easily masked by environmental thermal noise, leading to an increased false alarm rate.
[0004] Furthermore, existing monitoring technologies often analyze individual sampling points in isolation, neglecting the thermophysical correlation characteristics formed between power equipment through conductors and supporting structures. The actual physical defect heating exhibits a continuous gradient diffusion pattern in space, while environmental hot spots usually show a chaotic distribution. Without a comprehensive evaluation of the spatial topology and dynamic evolution characteristics of the thermal field, it will be impossible to provide early warning of hidden thermal defects in a sub-healthy state. Summary of the Invention
[0005] To address the technical problem of inaccurate equipment operation status assessment in unattended substations due to dynamic load fluctuations and complex meteorological environments, this invention provides solutions in the following aspects.
[0006] In a first aspect, the present invention provides an intelligent monitoring method for unattended substations, comprising: acquiring in real time an infrared radiation grayscale image of the area of the inspected equipment, the temperature of the monitoring point of the inspected equipment, and the load current of the circuit; for any monitoring point: determining the temperature rise inertia at the current moment based on the difference between the temperature change of adjacent sampling points within the observation window at the current moment and the square of the load current of the circuit; obtaining the spatial gradient vector of each pixel in the infrared radiation grayscale image where the monitoring point is located using a gradient operator; determining the thermal field coherence at the current moment based on the similarity between the spatial gradient vector of the monitoring point and the spatial gradient vector of each neighboring pixel of the monitoring point; determining the deviation energy value at the current moment based on the maximum value among the temperature rise inertia, thermal field coherence, and historical temperature rise inertia of the monitoring point at the current moment; and determining that the monitoring point has a fault at the current moment and triggering an unsafe warning for the substation in response to the sequence of deviation energy values of the monitoring point at the current moment and multiple historical moments before the current moment satisfying a monotonically increasing trend.
[0007] This invention effectively solves the problem of false alarms caused by environmental interference in unattended substations through multi-dimensional parameter alignment and dynamic evolution analysis. By constructing a temperature rise inertia index, it decouples load fluctuations from actual equipment heating and accurately captures minute initial heating defects. It uses thermal field coherence to identify spatial diffusion patterns and effectively filters out environmental reflection pseudo-hot spots. Finally, through a temporal monotonicity discrimination mechanism, it eliminates random thermal noise interference, achieving a leap from instantaneous detection to evolutionary trend monitoring, significantly improving early warning sensitivity and accuracy, and ensuring the stability of power grid operation.
[0008] Preferably, the temperature of the monitoring point of the device under test is obtained by taking the pixel at the center of the infrared radiation grayscale image of the area of the device under test as the monitoring point of the device under test; and converting the grayscale value of the monitoring point into a temperature value through the infrared detector calibration curve.
[0009] Preferably, the observation window refers to a fixed duration for backtracking to the past based on the current moment.
[0010] Preferably, the temperature rise inertia at the current moment satisfies the expression: In the formula, For the first The temperature rise inertia at each monitoring point at the current moment; For the first The monitoring point is the first one within the current observation window. The sampling point and the first Temperature at each sampling point; , The first one in the observation window at the current moment The sampling point and the first The loop load current at each sampling point; This is the rated current of the circuit; To observe the index value and total number of sampling points within the observation window; To avoid zero parameters; To take the absolute value.
[0011] This invention utilizes Joule's law to measure the physical relationship between current fluctuations and temperature rise, introduces rated current standardization to eliminate dimensional interference, and can decouple the true physical health of the equipment even under severe load fluctuations, accurately capture transient heat accumulation caused by contact resistance deterioration, and improve the ability to identify hidden defects.
[0012] Preferably, the gradient operator refers to the Sobel gradient operator.
[0013] Preferably, the thermal field coherence at the current moment satisfies the expression: In the formula, For the first The thermal field coherence of each monitoring point at the current moment; For the first Spatial gradient vector of each monitoring point; For the first The first monitoring point The spatial gradient vector of each neighboring pixel; These are the index values and total number of neighboring pixels; The modulus symbol; It is a small positive number.
[0014] This invention utilizes gradient vector consistency to measure thermal field laws. By leveraging the characteristic that physical heating follows radial diffusion, it effectively identifies and eliminates pseudo-heat sources in a cluttered and discontinuous environment. This spatial dimension constraint breaks the limitations of isolated point analysis and enhances the system's anti-interference performance in complex backgrounds from the underlying physical logic.
[0015] Preferably, the neighboring pixels are obtained as follows: in the infrared radiation grayscale image where the monitoring point is located, with the monitoring point as the center and a fixed pixel spacing as the radius, according to... Discrete sampling is performed at angular intervals to obtain eight pixels as the neighboring pixels of the monitoring point.
[0016] Preferably, the deviation energy value at the current moment satisfies the expression: In the formula, , , For the first The deviation energy value, temperature rise inertia, and thermal field coherence of each monitoring point at the current moment; For the first The maximum value of the historical temperature rise inertia at each monitoring point.
[0017] This invention uses dynamic weighting of temperature rise and spatial characteristics to generate an amplification effect to improve sensitivity when features match and a strong suppression effect when features do not match, compressing outliers to the background level and ensuring that real operational risks can be captured.
[0018] Preferably, the determination logic of the monotonic growth trend is as follows: sort the sequence of deviation energy values according to time sequence, calculate the first-order difference sequence of the sorted sequence, and if all values in the difference sequence are greater than 0 and the deviation energy value of the monitoring point at the current time is greater than the maximum value among the deviation energy values of multiple historical times before the current time, then it is considered to satisfy the time-series monotonic growth trend; otherwise, it is considered not to satisfy the time-series monotonic growth trend.
[0019] This invention is based on first-order difference trend determination logic, focusing on the dynamic evolution process of equipment degradation. By identifying the trend of continuous enhancement of energy sequence, it eliminates non-continuous interference such as environmental changes or sensor jitter, ensuring that the early warning has clear physical directionality and realizing early and accurate early warning before the fault reaches its peak.
[0020] Secondly, the present invention provides an intelligent monitoring system for unattended substations, including a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the above-mentioned intelligent monitoring method for unattended substations is implemented.
[0021] By adopting the above technical solution, a computer program for the above-mentioned unattended substation intelligent monitoring method is generated and stored in a memory for loading and execution by a processor. Terminal equipment is then created based on the memory and processor for convenient use.
[0022] The beneficial effects of this invention are as follows:
[0023] This invention constructs a multi-dimensional optimized monitoring system based on heat generation patterns, spatial distribution, and temporal evolution. Based on the dynamic decoupling method of temperature rise inertia, it solves the technical problem of coupling between large load fluctuations and meteorological environmental interference. At the same time, through thermal field coherence analysis, it endows the system with visual logic to identify false heat sources, enabling the monitoring system to evolve from simple numerical comparison to a comprehensive evaluation of the physical characteristics of the thermal field, significantly reducing the false alarm rate and ensuring the safe and stable operation of unattended substations. Attached Figure Description
[0024] Figure 1 This is a flowchart illustrating an unattended substation intelligent monitoring method according to the present invention;
[0025] Figure 2 This is a schematic diagram showing the temperature comparison of intelligent monitoring equipment in a substation. Detailed Implementation
[0026] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0027] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.
[0028] This invention discloses an intelligent monitoring method for unattended substations, referring to... Figure 1 This includes steps S1-S5:
[0029] S1. Real-time acquisition of infrared radiation grayscale image of the area of the equipment under inspection, temperature of the monitoring point of the equipment under inspection, and load current of the circuit.
[0030] It should be noted that the monitoring of unattended substations requires multi-dimensional alignment of thermal and electrical parameters. Various sensors installed in key locations are used to acquire load data such as current and voltage. At the same time, dual-light multi-view equipment with environmental compensation function is used to acquire infrared and visible light fused images to provide raw data support for decoupling environmental interference.
[0031] Specifically, the high-precision reset turntable deployed at the site is controlled to drive the dual-light acquisition unit to perform a fixed sampling frequency cruise scan on the inspected equipment parts such as disconnect switches and circuit breaker contacts, to obtain a fused image of infrared thermal image and visible light image. The visible light image is used to identify the physical outline of the inspected equipment, and the inspected equipment area in the infrared thermal image is identified according to the dual-light axis alignment parameters. The infrared radiation grayscale image of the area is extracted, and the pixel at the center of the grayscale image is used as the monitoring point of the inspected equipment.
[0032] Furthermore, the system retrieves the loop load current sequence of the station-end monitoring system in real time via industrial Ethernet at a fixed sampling frequency.
[0033] It should be added that the system uses a unified timing clock as a reference and sets the sampling frequency to 1Hz to ensure that the system synchronously collects the circuit load current and infrared radiation grayscale image in each sampling period.
[0034] At this point, the synchronized status data has been obtained.
[0035] S2. For any monitoring point: Determine the temperature rise inertia at the current moment based on the temperature change of adjacent sampling points within the observation window at the current moment and the difference between the square of the load current of the circuit.
[0036] It should be noted that when the load increases, the heating rate and heat dissipation rate of a normal electrical connector will reach a new equilibrium in a short time. If contact resistance is generated inside the connector due to oxidation, its temperature rise rate will be distorted with the step change of the load. Therefore, the physical health of the equipment in the dynamic process can be judged by analyzing the temperature change inertia caused by the fluctuation of unit joule heat power.
[0037] Specifically, by using the infrared detector calibration curve, the infrared radiation grayscale value is converted into a temperature value, and the temperature of any monitoring point at the current moment is obtained.
[0038] Obtain all temperatures and all circuit load currents within the observation window of any monitoring point at the current moment. The observation window refers to a fixed time period for backtracking from the current moment to the past.
[0039] It should be added that, in this embodiment of the invention, the fixed duration is 5 minutes to ensure that the segment contains the complete recent change process. The implementer can adjust it according to the rated load or thermal time constant of the equipment under test.
[0040] Based on the temperature change of any monitoring point within the observation window at the current moment and the difference between the square of the loop load, the temperature rise inertia of any monitoring point at the current moment is determined; the temperature rise inertia satisfies the expression:
[0041]
[0042] In the formula, For the first The temperature rise inertia at each monitoring point at the current moment; For the first The monitoring point is the first one within the current observation window. The sampling point and the first Temperature at each sampling point; , The first one in the observation window at the current moment The sampling point and the first The loop load current at each sampling point; This is the rated current of the circuit; To observe the index value and total number of sampling points within the observation window; To avoid zero parameters; To take the absolute value.
[0043] in, Indicates the first The degree of temperature jump at the sampling point within the observation window of each monitoring point at the current moment reflects the macroscopic feedback intensity of the corresponding equipment to external thermal disturbances. The intensity of external thermal power source fluctuations reflecting the heat generated by the equipment within the observation window at the current moment is determined by Joule's law, which states that the heat generated is proportional to the square of the current. The increase in heat production caused by load changes within the current observation window is standardized by dividing by the circuit's rated current to eliminate dimensional effects; if The larger the value, the more likely it is that the monitoring point has experienced a temperature jump that exceeds the normal physical condition under the same intensity of power fluctuation at the current moment. This indicates that the internal physical structure of the equipment has deteriorated, resulting in a surge in local thermal resistance. As a result, the heat cannot be effectively dissipated at the moment of generation, leading to transient accumulation.
[0044] It should be added that the parameters Used to prevent computational overflow under extremely low load fluctuations, the value range is [value range missing]. In this invention, 0.2 is used, but the implementer can adjust it according to the actual situation.
[0045] At this point, the temperature rise inertia of each monitoring point at the current moment has been obtained.
[0046] S3. Use the gradient operator to obtain the spatial gradient vector of each pixel in the infrared radiation grayscale image where the monitoring point is located; determine the thermal field coherence at the current moment based on the similarity between the spatial gradient vector of the monitoring point and the spatial gradient vector of each neighboring pixel of the monitoring point.
[0047] It should be noted that in the substation field, the heat between equipment mainly affects each other through physical conduction and spatial radiation. The heat diffusion of a real internal fault heat source of equipment will follow the law of spatial continuity and appear as a regular radial gradient distribution on the infrared image. In contrast, the heat field distribution of pseudo heat sources generated by environmental reflection often has discontinuous edge or messy pixel jump characteristics. Therefore, by analyzing the thermal field coherence between the monitoring point and its neighboring pixels, non-physical environmental interference can be effectively eliminated.
[0048] Specifically, the Sobel gradient operator is used to obtain the spatial gradient vector of each pixel in the infrared radiation grayscale image of any monitoring point.
[0049] To obtain the neighboring pixels of any monitoring point, the method is as follows: in the infrared radiation grayscale image where the monitoring point is located, with the monitoring point as the center and a fixed pixel spacing as the radius, according to... Discrete sampling is performed at angular intervals to obtain eight pixels as the neighboring pixels of the monitoring point.
[0050] It should be added that the fixed pixel spacing refers to the spatial displacement between the monitoring point and the neighboring sampling points. Since the heat generated by the device exhibits a radial conduction law from the inside to the outside in physics, selecting neighboring points with a certain spacing can more effectively capture the spatial temperature gradient characteristics between the monitoring point and its surrounding area. The fixed pixel spacing is 3 to 5 pixels. In this embodiment of the invention, the fixed pixel spacing is 4. The implementer can adjust it as needed.
[0051] The thermal field coherence of any monitoring point at the current moment is determined based on the similarity between the spatial gradient vector of any monitoring point and the spatial gradient vector of each neighboring pixel of that monitoring point; the thermal field coherence satisfies the expression:
[0052]
[0053] In the formula, For the first The thermal field coherence of each monitoring point at the current moment; For the first Spatial gradient vector of each monitoring point; For the first The first monitoring point The spatial gradient vector of each neighboring pixel; These are the index values and total number of neighboring pixels; The modulus symbol; It is a small positive number.
[0054] in, Reflecting the The spatial gradient vector of the monitoring point and the first monitoring point The first monitoring point The consistency of the spatial gradient vector change direction among neighboring pixels indicates that, when a real physical overheating source exists, heat is uniformly conducted to the surrounding area at the monitoring point, and the gradient direction exhibits a high degree of spatial consistency. The value approaches 1; the coherent features of all directions of all neighboring pixels of the monitoring point are fused, if... A value approaching 1 indicates that the thermal field at the monitoring point conforms to the radial diffusion topology of an ideal point source; if The extremely small value indicates that the thermal evolution trajectories at this monitoring point are uncorrelated. This spatial discontinuity suggests that the temperature rise is caused by non-physical interference from complex background reflection or non-uniform environmental heating. The range of values is To avoid the denominator being 0, 0.005 is used in this invention.
[0055] At this point, the thermal field coherence of each monitoring point at the current moment is obtained.
[0056] S4. Based on the maximum value among the temperature rise inertia, thermal field coherence, and historical temperature rise inertia of the monitoring point at the current moment, determine the deviation energy value at the current moment.
[0057] It should be noted that a real physical defect point must simultaneously meet two characteristics: abnormally sensitive thermal feedback to load fluctuations and regular conduction and diffusion of heat from the center to the periphery. In order to achieve highly reliable automatic early warning, the characteristics of the two need to be integrated to construct a multi-dimensional deviation energy value, thereby accurately identifying the real operational risks in a complex background.
[0058] Specifically, based on the maximum value among the temperature rise inertia, thermal field coherence, and historical temperature rise inertia at any monitoring point at the current moment, the deviation energy value of any monitoring point at the current moment is determined; the multidimensional deviation energy value satisfies the expression:
[0059]
[0060] In the formula, For the first The deviation energy value of each monitoring point at the current moment; For the first The temperature rise inertia at each monitoring point at the current moment; For the first The maximum value of the historical temperature rise inertia at each monitoring point; For the first The thermal field coherence of each monitoring point at the current moment.
[0061] in, Reflects the first Abnormal heat accumulation at the equipment connection points of a monitoring point due to increased contact resistance during load fluctuations; when the heat generation at this monitoring point conforms to the radial diffusion law of a real physical source, at this time... Approaching 1, The value approaching 2 indicates that the temperature rise of the equipment at this monitoring point has a clear physical source, thus effectively amplifying the effect. The contribution of this significantly improves the system's sensitivity to capturing real degradation trends; conversely, if the temperature rise is caused by irregular environmental reflections, It may increase due to instantaneous temperature jumps, but due to the thermal field coherence... Approaching -1 Approaching 0, The inhibitory effect will lead to the final The numerical values are significantly compressed to the background level, thus effectively filtering out this non-physical anomaly; considering that actual substation equipment will always have a thermal field during power supply, The value is greater than 0.
[0062] At this point, the deviation energy value of each monitoring point at the current moment is obtained.
[0063] S5. In response to the fact that the sequence of deviation energy values of the monitoring point at the current time and at multiple historical times before the current time satisfies a monotonically increasing trend, it is determined that there is a fault at the monitoring point at the current time, and an unsafe warning for the substation is triggered.
[0064] It should be noted that in the real-time monitoring of unattended substations, instantaneous energy deviations may originate from occasional environmental background thermal noise. In order to further eliminate non-continuous environmental interference and accurately capture the early physical structure degradation trend of the equipment, a monotonicity discrimination mechanism based on time-series logic is introduced. In the process of the formation of real physical defects, the heat generation power and the degree of thermal field distortion usually show a continuous increasing trend over time. By identifying the evolution logic of the energy sequence, the instantaneous noise and the physical evolution process of continuous degradation can be effectively distinguished.
[0065] Specifically, the deviation energy value of any monitoring point at the current moment is obtained, and the deviation energy values of the monitoring point at multiple historical moments before the current moment are retrieved from the historical database simultaneously. If the sequence of deviation energy values of any monitoring point at the current moment and multiple historical moments before the current moment satisfies a monotonically increasing trend, it indicates that the thermal state deviation of the monitoring point shows a clear evolutionary increasing trend. It is determined that there is a fault at the monitoring point at the current moment, and an unsafe warning for the substation is immediately triggered, indicating that there is a hidden danger at the monitoring point, which requires maintenance personnel to check.
[0066] The logic for determining a monotonically increasing trend is as follows: sort the sequence of deviation energy values according to time sequence, calculate the first-order difference sequence of the sorted sequence, and if all values in the difference sequence are greater than 0 and the deviation energy value of the monitoring point at the current moment is greater than the maximum value among the deviation energy values of multiple historical moments before the current moment, then it is considered to satisfy the time-series monotonically increasing trend; otherwise, it is considered not to satisfy the time-series monotonically increasing trend.
[0067] It should be added that the multiple historical moments determine the time observation window for trend determination. In this embodiment, the multiple historical moments are set to 5. In order to eliminate non-trend interference caused by ambient light flickering or random sensor jitter, and to ensure that the warning signal has a clear physical evolution direction, the implementers make dynamic adjustments according to the intensity of interference in the on-site environment.
[0068] For example, Figure 2This is a temperature comparison chart for intelligent monitoring of substation equipment. The horizontal axis represents monitoring time, and the vertical axis represents temperature value. Between 13:00 and 13:50, the temperature at the monitoring point exceeded the 40.0°C warning line due to interference from environmental heat sources, resulting in false anomalies. Traditional threshold methods, due to their single judgment dimension, triggered a large number of false alarms during this period. However, this invention, by calculating the temperature rise inertia and thermal field coherence, identified that the temperature rise lacked physical diffusion coherence, successfully eliminating environmental interference. In the later stage of actual fault evolution, this invention utilizes the monotonically increasing criterion of deviation energy to accurately locate the fault before the temperature reaches its peak, achieving a lower false alarm rate and higher early warning sensitivity than traditional methods.
[0069] This invention also discloses an intelligent monitoring system for unattended substations, including a processor and a memory. The memory stores computer program instructions, which, when executed by the processor, implement an intelligent monitoring method for unattended substations according to the present invention.
[0070] The system also includes other components well known to those skilled in the art, such as communication buses and communication interfaces, the settings and functions of which are known in the art and will not be described in detail here.
Claims
1. A method for intelligent monitoring of unattended substations, characterized in that, include: Real-time acquisition of infrared radiation grayscale images of the area of the equipment under inspection, temperature of the monitoring points of the equipment under inspection, and load current of the circuit; For any monitoring point: Determine the temperature rise inertia at the current moment based on the difference between the temperature change of adjacent sampling points within the observation window at the current moment and the square of the load current of the circuit; The spatial gradient vector of each pixel in the infrared radiation grayscale image where the monitoring point is located is obtained using the gradient operator; the thermal field coherence at the current moment is determined based on the similarity between the spatial gradient vector of the monitoring point and the spatial gradient vector of each neighboring pixel of the monitoring point. Based on the maximum value among the temperature rise inertia, thermal field coherence, and historical temperature rise inertia of the monitoring point at the current moment, the deviation energy value at the current moment is determined. If the sequence of deviation energy values of the monitoring point at the current time and at multiple historical times before the current time satisfies a monotonically increasing trend, it is determined that there is a fault at the monitoring point at the current time, triggering an unsafe warning for the substation.
2. The intelligent monitoring method for unattended substations according to claim 1, characterized in that, The temperature of the monitoring points of the tested equipment is obtained in the following way: The pixel at the center of the infrared radiation grayscale image of the area of the equipment under test is used as the monitoring point of the equipment under test; the grayscale value of the monitoring point is converted into a temperature value through the infrared detector calibration curve.
3. The intelligent monitoring method for unattended substations according to claim 1, characterized in that, The observation window refers to a fixed duration for tracing back to the past based on the current moment.
4. The intelligent monitoring method for unattended substations according to claim 1, characterized in that, The temperature rise inertia at the current moment satisfies the following expression: ； In the formula, For the first The temperature rise inertia at each monitoring point at the current moment; For the first The monitoring point is the first one within the current observation window. The sampling point and the first Temperature at each sampling point; , The first one in the observation window at the current moment The sampling point and the first The loop load current at each sampling point; This is the rated current of the circuit; To observe the index value and total number of sampling points within the observation window; To avoid zero parameters; To take the absolute value.
5. The intelligent monitoring method for unattended substations according to claim 1, characterized in that, The gradient operator referred to is the Sobel gradient operator.
6. The intelligent monitoring method for unattended substations according to claim 1, characterized in that, The thermal field coherence at the current moment satisfies the expression: ； In the formula, For the first The thermal field coherence of each monitoring point at the current moment; For the first Spatial gradient vector of each monitoring point; For the first The first monitoring point The spatial gradient vector of each neighboring pixel; These are the index values and total number of neighboring pixels; The modulus symbol; It is a small positive number.
7. The intelligent monitoring method for unattended substations according to claim 6, characterized in that, The method for obtaining the neighboring pixels is as follows: In the infrared radiation grayscale image where the monitoring point is located, with the monitoring point as the center and a fixed pixel spacing as the radius, according to... Discrete sampling is performed at angular intervals to obtain eight pixels as the neighboring pixels of the monitoring point.
8. The intelligent monitoring method for unattended substations according to claim 1, characterized in that, The deviation energy value at the current moment satisfies the expression: ； In the formula, , , For the first The deviation energy value, temperature rise inertia, and thermal field coherence of each monitoring point at the current moment; For the first The maximum value of the historical temperature rise inertia at each monitoring point.
9. The intelligent monitoring method for unattended substations according to claim 1, characterized in that, The logic for determining the monotonic growth trend is as follows: The sequence of deviation energy values is sorted according to time sequence. The first-order difference sequence of the sorted sequence is calculated. If all values in the difference sequence are greater than 0 and the deviation energy value of the monitoring point at the current time is greater than the maximum value among the deviation energy values of multiple historical times before the current time, then it is considered to satisfy the time-series monotonically increasing trend; otherwise, it is considered not to satisfy the time-series monotonically increasing trend.
10. An intelligent monitoring system for unattended substations, characterized in that, include: A processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, a method for intelligent monitoring of an unattended substation according to any one of claims 1-9 is implemented.