An on-line monitoring and early warning system for operation state of main equipment of large transformer substation

Abstract

The present application relates to the field of large substation monitoring, and more particularly to a large substation main equipment operation state online monitoring and early warning system, comprising a data acquisition module, a switching detection module, a temperature section identification module, a thermal dynamic inversion module, a baseline comparison and attribution module, a station level horizontal comparison and early warning module, taking the natural cooling stage switching in the main equipment operation as a step excitation, combining with the cooling system image to confirm the switching event, intercepting the temperature section before and after the switching, inverting the equivalent heat exchange coefficient and thermal time constant, and comparing with the historical baseline; according to the parameter deviation direction, the temperature anomaly is classified as heat dissipation capacity attenuation or internal latent defect heat production, and after the station level horizontal comparison, the individual case is early warned. The present application solves the problem that the two types of root causes, heat dissipation capacity attenuation and internal latent defect heat production, are difficult to distinguish, and improves the root cause identification accuracy and early warning ability.

Description

Technical Field

[0001] This invention relates to the field of monitoring technology for large substations, specifically to an online monitoring and early warning system for the operating status of main equipment in large substations. Background Technology

[0002] The main equipment in large substations, such as power transformers and shunt reactors, may develop internal latent defects during operation due to insulation aging, partial discharge, and poor contact. These defects manifest as abnormal increases in oil and winding temperatures caused by additional heat sources. Simultaneously, the cooling system experiences problems such as filter blockage, reduced airflow, and contamination of heat exchange surfaces over time, leading to decreased heat dissipation capacity and similarly resulting in increased oil and winding temperatures. These two failure modes exhibit highly similar temperature characteristics and have long been confused, posing challenges to operation and maintenance.

[0003] To address these issues, the industry has proposed various monitoring and diagnostic solutions, but none have been able to directionally separate the two types of root causes in their early stages. Technical solutions that primarily rely on alarms based on a single temperature threshold or temperature rise rate. These solutions only use whether the oil temperature or winding temperature exceeds a set threshold as the basis for alarms, and cannot distinguish between two root causes: reduced heat dissipation capacity due to cooling system degradation and heat generation due to internal latent defects. This makes it difficult for maintenance personnel to handle the situation effectively, resulting in both false alarms and missed alarms.

[0004] Schemes that proactively inject cooling power to assess heat dissipation characteristics after an alarm has been issued include, for example, Chinese Patent Publication No. CN120928252A, which discloses a method for monitoring and warning of transformer operating status based on state index calculation. This method proactively inputs multiple sets of different cooling power and sets a monitoring time when the transformer is in an alarm state, collects the radiator outlet air temperature, and comprehensively assesses heat dissipation characteristics based on the changes in cooling power and outlet air temperature difference to determine a fault alarm. However, this scheme uses the proactive injection of multiple sets of cooling power as a diagnostic tool and only assesses heat dissipation characteristics after an alarm has been issued. It does not differentiate between directional attribution of heat dissipation attenuation and internal heat generation, nor does it utilize naturally occurring cooling stage switching events during operation as excitation, nor does it fit and invert the equivalent heat transfer coefficient and thermal time constant from the oil temperature and winding transient temperature trajectories.

[0005] Schemes based on steady-state temperature inversion for identifying overheating faults, such as the Chinese patent with publication number CN121859170A, disclose a method for identifying transformer overheating faults based on temperature inversion. This method obtains thermal characteristic parameters by solving an inverse problem of a steady-state top-level oil temperature calculation model using top-level oil temperature, ambient temperature, and load factor. Overheating faults are identified when the deviation between the theoretical load current and the measured current exceeds a threshold. However, this scheme focuses on identifying overheating faults rather than directionally separating the two root causes—heat dissipation attenuation and internal heat generation—and does not treat cooling stage switching as a step excitation.

[0006] In summary, existing technologies, whether relying on single-threshold alarms, using heat transfer-related quantities for cooling control / efficiency evaluation, or assessing heat dissipation characteristics through steady-state methods or active injection, fail to provide a systematic solution that utilizes the main equipment's own operating characteristics, requires no additional disturbance, and can directionally separate heat dissipation capacity decay from internal latent defects at an early stage, while also identifying whether it is an isolated case or a widespread phenomenon. Therefore, there is an urgent need for an online monitoring and early warning system for the operating status of main equipment in large substations to solve the above problems. Summary of the Invention

[0007] To address this, the present invention provides an online monitoring and early warning system for the operating status of main equipment in large substations. This system overcomes the shortcomings of existing technologies in monitoring the main equipment of large substations, which, in the scenario of heat generation due to latent defects inside the oil-immersed transformer, generally issue alarms for abnormal equipment temperature and use heat exchange-related quantities only for cooling control and efficiency evaluation, failing to directionally separate the two root causes—heat generation due to heat dissipation capacity attenuation and heat generation due to latent defects—in the early stages.

[0008] To achieve the above objectives, the present invention provides an online monitoring and early warning system for the operating status of main equipment in large substations, comprising: The data acquisition module is used to collect the cooling stage switching status signal information of the target main equipment, as well as the top oil temperature timing data and winding temperature timing data of the main equipment. The switching detection module is used to acquire image data of the target main equipment cooling system and the cooling stage switching status signal to determine whether a cooling stage switching event has occurred. The temperature segment identification module is used to respond to the occurrence of a switching event by determining the steady-state temperature segment before switching and the transient temperature response segment after switching based on the top oil temperature timing data and the winding temperature timing data. The thermal dynamics inversion module is used to determine the equivalent heat transfer coefficient and thermal time constant of the target main equipment based on the transient temperature response segment; The baseline comparison and attribution module is used to compare the equivalent heat transfer coefficient and thermal time constant obtained by inversion with the historical baseline of a predetermined period, and to classify the current temperature anomaly according to the offset direction of the equivalent heat transfer coefficient and thermal time constant obtained by inversion relative to the historical baseline of the predetermined period. The station-level horizontal comparison module is used to perform a horizontal comparison query between the temperature anomaly of the target main equipment and the attribution result and similar data of several other main equipment in the large substation to determine whether the temperature anomaly is an isolated case. The early warning module is used to output early warning information corresponding to the root cause of the temperature anomaly based on the baseline comparison and the attribution results of the attribution module.

[0009] Furthermore, the process of the target main equipment cooling system image data identified by the switching detection module includes acquiring the image data of the target main equipment cooling system, performing image processing on the image data to determine the cooling fan blade image and the cooler status indicator light source image; Based on the image of the cooling fan blades, determine the start / stop and direction of rotation of the cooling fan blades; Based on the image of the cooler status indicator light source, the brightness state of the cooler status indicator light source is determined.

[0010] Furthermore, the process by which the switching detection module determines whether a cooling stage switching event has occurred includes: If the cooling fan blades are detected to change from stationary to rotating, the corresponding cooling stage is determined to be in operation; if the cooling fan blades are detected to change from rotating to stationary, the corresponding cooling stage is determined to be out of operation. If the status indicator light of the cooler changes from off to on, it is determined that the corresponding cooling stage is in operation. If the status indicator light of the cooler changes from lit to off, it is determined that the corresponding cooling stage has exited operation. If the cooling stage switching status signal changes from an unswitched state to a switched state, it indicates that the corresponding cooling stage is put into operation. If the cooling stage switching status signal changes from switched to unswitched, it indicates that the corresponding cooling stage has stopped operating.

[0011] Furthermore, the temperature range identification module A timing marker submodule is used to mark the switching time on the top oil temperature timing data and the winding temperature timing data in response to the occurrence of the cooling stage switching event; The temperature segment extraction submodule is used to extract temperature data for a preset time length before the switching time as the steady-state temperature segment before the switching time, and extract temperature data for a preset time length after the switching time as the transient temperature response segment after the switching time.

[0012] Furthermore, the process by which the thermal dynamic inversion module determines the equivalent heat transfer coefficient and thermal time constant of the target main equipment based on the transient temperature response segment includes: Extract the temperature trajectories of the top oil temperature and the winding temperature as a function of time within the transient temperature response segment; The temperature trajectory is fitted with least squares, and the equivalent heat transfer coefficient is determined based on the steady-state temperature value reached after the switching. The thermal time constant is determined based on the rate at which the temperature trajectory approaches the steady-state temperature value after the switching.

[0013] Furthermore, the baseline comparison and attribution module includes: The baseline establishment submodule is used to cluster historical data according to the combination of cooling stages, load rate, and ambient temperature, and to establish historical baselines for the equivalent heat transfer coefficient and the thermal time constant for each cluster of operating conditions. The comparison attribution module is used to select the historical baseline corresponding to the operating condition cluster that matches the current operating condition, compare the equivalent heat transfer coefficient and the thermal time constant obtained by inversion with the historical baseline, and classify the current temperature anomaly according to the offset direction.

[0014] Furthermore, the process by which the comparison attribution factor module classifies the current temperature anomaly includes: If the equivalent heat transfer coefficient obtained by inversion decreases relative to the historical baseline and the thermal time constant increases relative to the historical baseline, then the temperature anomaly is classified as a decrease in heat dissipation capacity. If the equivalent heat transfer coefficient obtained by inversion remains basically unchanged relative to the historical baseline and the top oil temperature continues to rise, then the temperature anomaly is attributed to internal latent defects generating heat. If the equivalent heat transfer coefficient and the thermal time constant obtained by inversion both fall within the tolerance band of the historical baseline, then the temperature anomaly is classified as a fluctuation under normal operating conditions.

[0015] Furthermore, the process by which the early warning module determines the early warning information corresponding to the root cause of the temperature anomaly based on the baseline comparison and the attribution classification results of the attribution module includes: Based on the results of attribution classification, the warning level, attribution category, and location of the anomaly in the early warning information are determined. Based on the attribution classification results, the following treatment recommendations are made: if the attribution classification result is a decrease in heat dissipation capacity, the treatment recommendations are directed to the maintenance of the cooling system and the overhaul of the filter and fan; if the attribution classification result is heat generation due to internal latent defects, the treatment recommendations are directed to further diagnosis of internal insulation or latent defects. The output includes the warning information, the attribution category, the location of the anomaly, and the handling recommendations.

[0016] Furthermore, the online monitoring and early warning system for the main equipment operation status of the large substation also includes: The thermal balance residual confirmation unit is used to perform secondary confirmation on the attribution classification results of the baseline comparison and attribution module, which determine that the heat is generated by internal latent defects, based on the thermal balance residual between the oil temperature data and the gas production rate data.

[0017] Furthermore, the process by which the thermal balance residual verification unit performs a secondary verification of the attribution classification result includes: Using the equivalent heat transfer coefficient obtained from the inversion and the current oil temperature, the thermal power corresponding to the theoretical gas production rate under steady state is calculated. The measured additional heat power was obtained by converting the measured gas production rate. The difference between the thermal power corresponding to the theoretical gas production rate and the measured additional thermal power is determined as the thermal balance residual. If the thermal balance residual exceeds a predetermined range and the equivalent heat transfer coefficient obtained by inversion does not decrease significantly relative to the historical baseline, then it is confirmed that the temperature anomaly originates from heat generation due to internal latent defects rather than a decrease in heat dissipation capacity. Compared with existing technologies, the system of this invention includes a data storage module, a switching detection module, a temperature segment identification module, a thermal dynamic inversion module, a baseline comparison and attribution module, a station-level lateral comparison module, and an early warning module. The data storage module acquires the cooling stage switching status signal of the target main equipment, as well as top-level oil temperature time-series data and winding temperature time-series data, providing a unified and continuous data foundation for subsequent inversion and attribution. The switching detection module combines the image data of the cooling system with the switching status signal to cross-reference and determine whether a cooling stage switching event has actually occurred. The temperature segment identification module responds to the switching event, automatically determining the steady-state temperature segment before switching and the transient temperature response segment after switching. The thermal dynamic inversion module further enhances the system's capabilities. The simulation module can obtain the equivalent heat transfer coefficient and thermal time constant of the target main equipment based on the transient temperature response segment and the lumped heat capacity model. The baseline comparison and attribution module can compare the obtained thermal dynamic parameters with the historical baseline and classify the directional attribution of temperature anomalies into heat dissipation capacity attenuation and heat generation from internal latent defects based on the offset direction of the temperature anomaly relative to the historical baseline. The station-level horizontal comparison module can compare the temperature anomaly and attribution results of this equipment with similar data of several other main equipment in the station to determine whether the temperature anomaly is a common phenomenon. The early warning module can output early warning information corresponding to the root cause based on the attribution results, realizing differentiated and early fault warning and handling guidance.

[0018] In particular, by using the naturally occurring cooling stage switching during operation as a free step excitation, and combining image detection of visible physical changes in the cooling system to confirm the switching event, the equivalent heat transfer coefficient and thermal time constant are inverted through the lumped heat capacity model, and the directional attribution is compared with the historical baseline of the predetermined period. Combined with the horizontal comparison of multiple devices at the station level to determine individual cases, early warnings are given respectively. This can separate the two types of temperature anomaly root causes that have been confused for a long time in the early stage without additional disturbance.

[0019] In particular, this invention addresses the problem that existing technologies rely on a single temperature threshold for alarms and cannot distinguish the root cause. Existing solutions only alarm based on whether the oil temperature or winding temperature exceeds the threshold. The reduced heat dissipation capacity and the heat generated by internal latent defects are highly similar in temperature performance, making it difficult for maintenance to handle them in a targeted manner. This invention utilizes the transient thermal response under switching step excitation to invert thermal dynamic parameters and makes directional attributions based on the direction of their offset relative to the historical baseline, thus separating the two types of root causes from the mechanism.

[0020] In particular, this invention addresses the problem that existing cooling capacity assessment and intelligent adjustment schemes rely on offline calibration for heat transfer coefficients and do not make directional attributions. This invention does not aim to control cooling, but passively utilizes the naturally occurring cooling stage switching during operation as a free stimulus, and inverts the equivalent heat transfer coefficient and thermal time constant from the transient temperature trajectory in real time, outputting two differentiated early warnings: one for heat dissipation capacity attenuation or heat generation due to internal latent defects. Attached Figure Description

[0021] Figure 1 This is a functional module block diagram of the online monitoring and early warning system for the main equipment operation status of a large substation, according to an embodiment of the present invention. Figure 2 This is a temperature transient response and inversion curve of the cooling stage switching as a step excitation in an embodiment of the present invention; Figure 3 This is a flowchart of the secondary confirmation of the oil temperature-gas production heat balance residual of the ultra-high voltage parallel reactor in a preferred embodiment of the present invention. Figure 4 This is a flowchart of an online monitoring and early warning method for the operating status of main equipment in a large substation, according to an embodiment of the present invention. Detailed Implementation

[0022] To make the objectives and advantages of the present invention clearer, the present invention will be further described below with reference to embodiments; it should be understood that the specific embodiments described herein are merely for explaining the present invention and are not intended to limit the present invention.

[0023] Preferred embodiments of the present invention will now be described with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0024] Please see Figure 1 This is a functional module block diagram of the online monitoring and early warning system for the main equipment operation status of a large substation according to an embodiment of the present invention; the system of the embodiment includes: Data acquisition module: It is used to acquire the cooling stage switching status signal of the target main equipment, as well as minute-level time-series data of top oil temperature and winding temperature through the main equipment online monitoring device; Switching detection module: It uses a visible light camera to acquire image data of the cooling system and cooling stage switching status signals to determine whether a cooling stage switching event has occurred; in the preferred implementation, image processing is performed on the image data to identify changes in the operating status of the cooling system to assist in the determination; the image data includes the start, stop and direction of the cooling fan blades, and the lighting and extinguishing of the cooler status indicator light. Temperature segment identification module: It is used to respond to the occurrence of a switching event, mark the time t0, and extract the steady-state temperature segment before t0 and the transient temperature response segment after t0; Thermal dynamics inversion module: It is used to fit and invert the equivalent heat transfer coefficient and thermal time constant of the target main equipment based on the transient temperature response segment using a lumped heat capacity model; Baseline Comparison and Attribution Module: It is used to compare the inverted thermal dynamic parameters with the historical baseline of a predetermined period, classify the causes of temperature anomalies according to the direction of offset, and further classify the severity levels. Station-level horizontal comparison module: It is used to compare the temperature anomaly of this equipment and its attribution results with several other main equipment in the station to determine whether the temperature anomaly is an isolated case. Early warning module: It is used to output early warning information based on the baseline comparison and the attribution results of the attribution module.

[0025] In this embodiment, the thermal dynamics inversion module constructs a lumped heat capacity model and performs least squares fitting on the transient segment. The lumped heat model is determined by fitting based on oil temperature or winding temperature, ambient temperature, internal heat source power, equivalent heat transfer coefficient, and lumped heat capacity. In this embodiment, the thermal time constant is the ratio of lumped heat capacity to equivalent heat transfer coefficient.

[0026] The baseline comparison and attribution module clusters historical data based on three factors: cooling stage, load rate, and ambient temperature, establishing a historical baseline for each operating condition cluster at a predetermined period. During comparison, a matching operating condition cluster baseline is selected: if the equivalent heat transfer coefficient decreases and the thermal time constant increases, it is classified as a decrease in heat dissipation capacity; if the equivalent heat transfer coefficient remains essentially unchanged while the oil temperature rises and gas production increases, it is classified as heat generation due to internal latent defects; if the parameters fall within the baseline tolerance band, it is classified as normal operating condition fluctuation; and further, the severity level is determined based on the magnitude of the deviation.

[0027] Please see Figure 2This is a transient temperature response and inversion curve for cooling stage switching according to an embodiment of the present invention. In this embodiment, after the cooling stage switching event occurs, the oil temperature begins to change transiently from its steady-state value: when the heat dissipation capacity is normal, the equivalent heat transfer coefficient is large and the thermal time constant is small, and the temperature quickly tends to a new equilibrium; when the heat dissipation capacity decreases, the equivalent heat transfer coefficient decreases and the thermal time constant increases, and the rate at which the temperature tends to equilibrium is significantly slower. The thermal dynamic inversion module fits the transient segment after the switching event using the aforementioned lumped heat capacity model to obtain the equivalent heat transfer coefficient and thermal time constant, which are then used by the baseline comparison and attribution module for directional classification.

[0028] Specifically, the baseline comparison, directional attribution, and detailed classification process in this embodiment of the invention includes: the baseline comparison and attribution module obtains the equivalent heat transfer coefficient and thermal time constant of the current inversion; selects a historical baseline of the same operating condition that matches the current operating condition; compares the offset direction of the inversion parameters relative to the baseline; classifies them according to the aforementioned rules as heat dissipation capacity attenuation, heat generation from internal latent defects, or fluctuations in normal operating conditions, and classifies the severity level according to the offset magnitude; and then outputs the results to the early warning module and the station-level horizontal comparison module.

[0029] Specifically, the process of identifying individual cases as general phenomena through station-level horizontal comparison in this embodiment of the invention includes: the station-level horizontal comparison module horizontally compares and queries the temperature anomaly of this equipment and its attribution results with similar data of several other main equipment in a large substation; if only this equipment shows the abnormal attribution, while other similar equipment does not show the same anomaly, it is determined to be an individual case; if multiple similar equipment show the same attribution, it is determined to be a general phenomenon, indicating that there may be common operating conditions or systemic factors.

[0030] Please see Figure 3 This is a flowchart illustrating the secondary confirmation process for the oil temperature-gas production heat balance residual in this embodiment of the invention. The target main equipment in this embodiment is an ultra-high voltage parallel reactor. The data acquisition module additionally collects the dissolved gas production rate in the oil. The heat balance residual confirmation unit performs the following steps: using the equivalent heat transfer coefficient obtained through inversion combined with the current oil temperature, the theoretical heat power corresponding to the gas production under steady state is calculated; the measured additional heat power is obtained by converting the measured gas production rate; the difference between the two is the heat balance residual. When the heat balance residual exceeds the predetermined range and the equivalent heat transfer coefficient does not decrease significantly, it is confirmed that the temperature anomaly originates from heat generation due to internal latent defects, rather than a decrease in heat dissipation capacity, thus performing a secondary confirmation of the second root cause attribution.

[0031] Specifically, the baseline comparison and attribution module uses a sliding time window to continuously update the historical baseline of each operating condition cluster at a predetermined period to suppress baseline drift caused by natural aging of equipment. When determining the degradation of heat dissipation capacity, it first removes the slow drift component caused by aging, and only alarms for sudden changes and trend-based declines that exceed the aging trend. The system executes the aforementioned modules independently for multiple main devices in the substation, and the substation-level horizontal comparison module summarizes them into an overview of the operating status of the main devices at the substation level.

[0032] Each module runs as software on edge computing nodes within the substation that already possess data acquisition capabilities, reusing existing online monitoring data without altering primary equipment or the main protection system. The switching detection module's camera is installed in locations that allow observation of cooling fans, status indicator lights, and oil flow indicators; inversion and comparison operate on a minute-level cycle, resulting in low computational overhead and suitability for long-term online operation.

[0033] Please see Figure 4 This is a flowchart of an online monitoring and early warning method for the operating status of main equipment in a large substation according to an embodiment of the present invention; the method described in the embodiment includes: Step S1: Collect the cooling stage switching status signal information, top oil temperature timing data and winding temperature timing data of the target main equipment within a predetermined period, and analyze the gas generation rate data of dissolved gas in the oil. Step S2: Extract image data of the cooling system and cooling stage switching status signal information for the corresponding predetermined period. By performing image processing on the image data, identify the start and stop information of the cooling fan blade rotation and the lighting and extinguishing information of the cooler status indicator light. Combined with the cooling stage switching status signal, determine whether a cooling stage switching event has occurred. Step S3: If a cooling stage switching event occurs, the switching time node is marked based on the top oil temperature timing data and the winding temperature timing data, and the steady-state temperature segment before switching and the transient temperature response segment after switching are determined. Step S4: Based on the transient temperature response segment after switching, perform least squares fitting on the top oil temperature and winding temperature to invert and obtain the equivalent heat transfer coefficient and thermal time constant of the target main equipment. Step S5: The baseline comparison and attribution module compares the equivalent heat transfer coefficient and thermal time constant obtained by inversion with the predetermined cycle and the historical baseline under the same operating conditions. Based on the direction of the offset of the inversion parameters relative to the historical baseline, the temperature anomaly is classified into heat dissipation capacity decay, heat generation due to internal latent defects, or fluctuations under normal operating conditions, and the severity level is divided according to the offset magnitude. The historical baseline is continuously updated using a sliding time window to eliminate baseline drift caused by natural aging of the equipment. Step S6: Using the equivalent heat transfer coefficient obtained by inversion and the current oil temperature, the thermal balance residual confirmation unit calculates the theoretical gas production rate and compares it with the measured gas production rate to obtain the thermal balance residual. When the residual exceeds the predetermined range, the attribution result of the internal latent defect heat generation determined in step S5 is confirmed for the second time to distinguish between internal heat generation and heat dissipation attenuation. Step S7: The temperature anomaly of the target main equipment and its attribution results are compared and queried with similar data of several other main equipment in the large substation through the station-level horizontal comparison module to determine whether the temperature anomaly is an isolated case or a common phenomenon. In step S8, the early warning module outputs early warning information, including early warning level, attribution category, location of abnormality and handling suggestions, based on the baseline comparison and attribution results of the attribution module, and connects it to the substation centralized monitoring platform; the station-level horizontal comparison module summarizes the operating results of multiple main equipment to form an overview of the operating status of substation-level main equipment.

[0034] The embodiment processes image data to identify the start / stop information of cooling fan blade rotation and the on / off information of cooler status indicator light. Combined with the cooling stage switching status signal, the process of determining whether a cooling stage switching event has occurred includes: when the blades change from stationary to rotating, the cooling stage is determined to be in operation, i.e., a start event occurs; when the blades change from rotating to stationary, the cooling stage is determined to be out of operation, i.e., an exit event occurs; when the cooler status indicator light changes from off to on, the cooling stage is determined to be in operation, i.e., a start event occurs; when the indicator light changes from on to off, the cooling stage is determined to be out of operation, i.e., an exit event occurs; when the cooling stage switching status signal changes from off to on, it indicates that the cooling stage is in operation, and vice versa. The above three signals corroborate each other and serve as a reference for determining whether a switching event has actually occurred.

[0035] Through the cyclic execution of steps S1 to S8, the system uses the naturally occurring cooling stage switching during operation as a free step excitation, achieving directional separation and differentiated early warning of two root causes: heat dissipation capacity attenuation and heat generation from internal latent defects, without additional disturbance. In the scenario of UHV parallel reactors, the reliability of the judgment of heat generation from internal latent defects is improved by leveraging the oil temperature-gas generation heat balance residual.

[0036] Specifically, the process by which the cut detection module identifies the image data of the target main equipment cooling system includes: Image data of the target main equipment cooling system is acquired, and image processing is performed on the image data to determine the image of the cooling fan blades and the image of the cooler status indicator light source; Based on the image of the cooling fan blades, determine the start / stop and direction of rotation of the cooling fan blades; Based on the image of the cooler status indicator light source, the brightness state of the cooler status indicator light source is determined.

[0037] The switching detection image recognition process in the corresponding embodiment locates the fan blades and status indicator light source from the cooling system image and determines their start / stop, direction, and brightness status. This provides visible physical evidence independent of electrical contacts for determining cooling stage switching events. It not only eliminates the dependence on a single electrical signal and avoids missed judgments caused by signal loss or malfunction, but also uses macroscopic and microscopic physical changes of the equipment body for multi-source mutual verification. At the same time, image processing automatically completes the state jump detection, enabling the switching event to be reliably confirmed without human intervention. Ultimately, this improves the anti-interference capability of switching detection and the overall early warning reliability of the system.

[0038] Specifically, the process by which the switching detection module determines whether a cooling stage switching event has occurred includes: If the cooling fan blades are detected to change from stationary to rotating, the corresponding cooling stage is determined to be in operation; if the cooling fan blades are detected to change from rotating to stationary, the corresponding cooling stage is determined to be out of operation. If the status indicator light of the cooler changes from off to on, it is determined that the corresponding cooling stage is in operation. If the status indicator light of the cooler changes from lit to off, it is determined that the corresponding cooling stage has exited operation. If the cooling stage switching status signal changes from an unswitched state to a switched state, it indicates that the corresponding cooling stage is put into operation. If the cooling stage switching status signal changes from switched to unswitched, it indicates that the corresponding cooling stage has stopped operating.

[0039] The switching event determination process in this embodiment verifies the consistency of two types of image signals—blade start / stop and indicator light on / off—with the cooling stage switching status signal, providing an accurate and reliable switching time benchmark for temperature segment identification. This not only reduces the probability of false alarms or missed alarms from a single signal through multi-criteria cross-validation but also clarifies the directional semantics of engagement and disengagement. Furthermore, by using electrical signals as a reference, it suppresses misjudgments caused by image illumination interference, ensuring that subsequent transient temperature segment extraction is strictly anchored to the actual switching time. Ultimately, this improves the time alignment accuracy of the inversion input and the robustness of the attribution judgment. Specifically, the temperature segment identification module… A timing marker submodule is used to mark the switching time on the top oil temperature timing data and the winding temperature timing data in response to the occurrence of the cooling stage switching event; The temperature segment extraction submodule is used to extract temperature data for a preset time length before the switching time as the steady-state temperature segment before the switching time, and extract temperature data for a preset time length after the switching time as the transient temperature response segment after the switching time.

[0040] The temperature segment identification process in this embodiment provides a temperature trajectory window with clear boundaries and explicit physical meaning for thermal dynamic inversion by using the switching moment as the boundary and extracting steady-state and transient segments of preset lengths before and after the switching moment. This not only eliminates the interference of slow load and environmental drift before the switching moment, but also ensures the consistency and comparability of the extracted window length under each switching event. At the same time, it enables the transient segment to fully cover the response process of the temperature tending to a new equilibrium, allowing the inversion module to stably fit the lumped heat capacity model, and ultimately improves the accuracy and repeatability of the equivalent heat transfer coefficient and thermal time constant inversion.

[0041] Specifically, the process by which the thermal dynamic inversion module determines the equivalent heat transfer coefficient and thermal time constant of the target main equipment based on the transient temperature response segment includes: Extract the temperature trajectories of the top oil temperature and the winding temperature as a function of time within the transient temperature response segment; The temperature trajectory is fitted with least squares, and the equivalent heat transfer coefficient is determined based on the steady-state temperature value reached after the switching. The thermal time constant is determined based on the rate at which the temperature trajectory approaches the steady-state temperature value after the switching.

[0042] The thermal dynamics inversion process in the corresponding embodiment fits the transient temperature trajectory after switching using the least squares method, and calculates the equivalent heat transfer coefficient and thermal time constant from the steady-state temperature value after switching and the rate of approaching that value, respectively. This provides quantifiable and comparable thermal dynamic parameters for baseline comparison. It not only avoids the stringent requirements of the traditional steady-state thermal resistance method on the steady-state period, but also makes full use of the free step excitation of the cooling stage switching that occurs naturally during operation. At the same time, it provides two complementary parameters, the equivalent heat transfer coefficient and the thermal time constant, with a single model, so that the two root causes of heat dissipation capacity and internal heat generation can be separated within the same framework, ultimately improving the ability to identify the root causes of temperature anomalies.

[0043] Specifically, the baseline comparison and attribution module includes: The baseline establishment submodule is used to cluster historical data according to the combination of cooling stages, load rate, and ambient temperature, and to establish historical baselines for the equivalent heat transfer coefficient and the thermal time constant for each cluster of operating conditions. The comparison attribution module is used to select the historical baseline corresponding to the operating condition cluster that matches the current operating condition, compare the equivalent heat transfer coefficient and the thermal time constant obtained by inversion with the historical baseline, and classify the current temperature anomaly according to the offset direction.

[0044] The baseline establishment and comparison process in this embodiment establishes historical baselines for each operating condition cluster by clustering based on the combination of cooling level, load rate, and ambient temperature. Then, it selects a matching baseline based on the current operating condition for offset comparison, providing a comparable reference benchmark under the same operating condition for attribution classification. This not only eliminates misattribution caused by natural differences in thermal parameters under different loads and environmental conditions, but also enhances the model's tolerance for individual equipment differences by using an adaptive baseline for operating conditions. At the same time, it uses the offset direction rather than a fixed threshold as the criterion, which allows for directional differentiation between the two root causes of heat dissipation attenuation and internal heat generation, ultimately improving the accuracy and generalization ability of attribution classification.

[0045] Specifically, the process by which the comparison attribution module classifies the current temperature anomaly includes: If the equivalent heat transfer coefficient obtained by inversion decreases relative to the historical baseline and the thermal time constant increases relative to the historical baseline, then the temperature anomaly is classified as a decrease in heat dissipation capacity. If the equivalent heat transfer coefficient obtained by inversion remains basically unchanged relative to the historical baseline and the top oil temperature continues to rise, then the temperature anomaly is attributed to internal latent defects generating heat. If the equivalent heat transfer coefficient and the thermal time constant obtained by inversion both fall within the tolerance band of the historical baseline, then the temperature anomaly is classified as a fluctuation under normal operating conditions.

[0046] The attribution classification process in this embodiment classifies temperature anomalies using a combination rule of offset between equivalent heat transfer coefficient and thermal time constant, supplemented by a continuous rise in top oil temperature as evidence of internal heat generation. This provides the early warning module with a classification result with a clear root cause. It not only achieves directional separation between heat dissipation capacity attenuation and heat generation from internal latent defects, but also filters out false alarms caused by fluctuations in normal operating conditions by using tolerance band judgment. At the same time, it reduces misclassification caused by ambiguity of a single indicator by using multi-parameter joint criteria, enabling early warning information to accurately point to the corresponding handling direction, ultimately improving the pertinence of fault early warning and the efficiency of operation and maintenance.

[0047] Specifically, the process by which the early warning module determines the early warning information corresponding to the root cause of the temperature anomaly based on the baseline comparison and the attribution classification results of the attribution module includes: Based on the results of attribution classification, the warning level, attribution category, and location of the anomaly in the early warning information are determined. Based on the attribution classification results, the following treatment recommendations are made: if the attribution classification result is a decrease in heat dissipation capacity, the treatment recommendations are directed to the maintenance of the cooling system and the overhaul of the filter and fan; if the attribution classification result is heat generation due to internal latent defects, the treatment recommendations are directed to further diagnosis of internal insulation or latent defects. The output includes the warning information, the attribution category, the location of the anomaly, and the handling recommendations.

[0048] The early warning information generation process in this embodiment determines the early warning level, attribution category, location of anomaly, and differentiated handling suggestions based on the attribution classification results and outputs them uniformly. This provides maintenance personnel with root cause-oriented and actionable decision-making basis. It not only maps the two root causes of heat dissipation attenuation and internal heat generation to differentiated maintenance and diagnostic paths, but also improves the readability of alarm information and the integrability between systems with structured fields. At the same time, it realizes tiered risk management through level classification, enabling maintenance resources to be accurately deployed as needed, and ultimately improving the practicality and timeliness of early warning for substation main equipment.

[0049] Specifically, the online monitoring and early warning system for the main equipment operation status of large substations also includes: The thermal balance residual confirmation unit is used to perform secondary confirmation on the attribution classification results of the baseline comparison and attribution module, which determine that the heat is generated by internal latent defects, based on the thermal balance residual between the oil temperature data and the gas production rate data.

[0050] The thermal balance residual verification unit in this embodiment provides physical property evidence independent of thermal dynamic parameters for the attribution of heat generation from internal latent defects by setting up a secondary verification step based on the thermal balance residual of oil temperature and gas generation rate. This not only makes up for the ambiguity of the judgment of a single thermal parameter in the case of heat dissipation attenuation and internal heat generation boundary, but also enhances the specific adaptation of the main equipment by taking advantage of the oil-gas thermal balance relationship unique to UHV parallel reactors. At the same time, the secondary verification mechanism reduces the risk of high-consequence false alarms, making the early identification of internal latent defects more reliable, and ultimately improving the system's confidence in the judgment of high-risk root causes.

[0051] Specifically, the process by which the thermal balance residual verification unit performs a secondary verification of the attribution classification result includes: Using the equivalent heat transfer coefficient obtained from the inversion and the current oil temperature, the thermal power corresponding to the theoretical gas production rate under steady state is calculated. The measured additional heat power was obtained by converting the measured gas production rate. The difference between the thermal power corresponding to the theoretical gas production rate and the measured additional thermal power is determined as the thermal balance residual. If the thermal balance residual exceeds the predetermined range and the equivalent heat transfer coefficient obtained by inversion does not decrease significantly relative to the historical baseline, then it is confirmed that the temperature anomaly originates from heat generation due to internal latent defects rather than a decrease in heat dissipation capacity.

[0052] The secondary confirmation process in this embodiment calculates the theoretical gas-generating heat power from the equivalent heat transfer coefficient and the current oil temperature, and compares it with the measured additional heat power to obtain the heat balance residual. The residual exceeding the limit and the heat transfer coefficient not decreasing significantly are used as confirmation conditions. This provides a quantitative and interpretable judgment boundary for attributing the heat generation of internal latent defects. It not only quantifies the intensity of the internal additional heat source by the deviation between the expected heat generation and the actual heat generation, but also eliminates the interference of heat dissipation attenuation under the premise that the heat transfer coefficient does not decrease significantly. At the same time, it makes the confirmation conclusion have clear physical semantics and auditability, so that high-risk internal defects can be identified with high confidence at an early stage, and ultimately improves the accuracy and reliability of the system's early warning of latent faults.

[0053] The technical solution of the present invention has been described above with reference to the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the scope of protection of the present invention.

Claims

1. An online monitoring and early warning system for the operating status of main equipment in a large substation, characterized in that, include: The data storage module is used to acquire the cooling stage switching status signal of the target main equipment, as well as the top oil temperature timing data and winding temperature timing data of the main equipment. The switching detection module is used to acquire image data of the target main equipment cooling system and the switching status signal information of the cooling stage in order to determine whether a cooling stage switching event has occurred. The temperature segment identification module is used to respond to the occurrence of a switching event by determining the steady-state temperature segment before switching and the transient temperature response segment after switching based on the top oil temperature timing data and the winding temperature timing data. The thermal dynamics inversion module is used to determine the equivalent heat transfer coefficient and thermal time constant of the target main equipment based on the transient temperature response segment; The baseline comparison and attribution module is used to compare the equivalent heat transfer coefficient and thermal time constant obtained by inversion with the historical baseline of a predetermined period. Based on the offset direction of the equivalent heat transfer coefficient and thermal time constant obtained by inversion relative to the historical baseline of the predetermined period, the current temperature anomaly is classified by attribution. The station-level horizontal comparison module is used to perform horizontal comparison and query between the temperature anomaly of the target main equipment and the attribution result and similar data of several other main equipment in the large substation, so as to determine the handling strategy for the temperature anomaly of the target main equipment. The early warning module is used to determine early warning information corresponding to the root cause of the temperature anomaly based on the baseline comparison and the attribution classification results of the attribution module.

2. The online monitoring and early warning system for the main equipment operation status of a large substation according to claim 1, characterized in that, The process by which the switching detection module identifies the image data of the target main equipment cooling system includes: Image data of the target main equipment cooling system is acquired, and image processing is performed on the image data to determine the image of the cooling fan blades and the image of the cooler status indicator light source; Based on the image of the cooling fan blades, determine the start / stop and direction of rotation of the cooling fan blades; Based on the image of the cooler status indicator light source, the brightness state of the cooler status indicator light source is determined.

3. The online monitoring and early warning system for the main equipment operation status of a large substation according to claim 2, characterized in that, The process by which the switching detection module determines whether a cooling stage switching event has occurred includes: If the cooling fan blades are detected to change from stationary to rotating, the corresponding cooling stage is determined to be in operation; if the cooling fan blades are detected to change from rotating to stationary, the corresponding cooling stage is determined to be out of operation. If the status indicator light of the cooler changes from off to on, it is determined that the corresponding cooling stage is in operation. If the status indicator light of the cooler changes from lit to off, it is determined that the corresponding cooling stage has exited operation. If the cooling stage switching status signal changes from an unswitched state to a switched state, it indicates that the corresponding cooling stage is put into operation. If the cooling stage switching status signal changes from switched to unswitched, it indicates that the corresponding cooling stage has stopped operating.

4. The online monitoring and early warning system for the operating status of main equipment in large substations according to claim 1, characterized in that, The temperature range identification module A timing marker submodule is used to mark the switching time on the top oil temperature timing data and the winding temperature timing data in response to the occurrence of the cooling stage switching event; The temperature segment extraction submodule is used to extract temperature data for a preset time length before the switching time as the steady-state temperature segment before the switching time, and extract temperature data for a preset time length after the switching time as the transient temperature response segment after the switching time.

5. The online monitoring and early warning system for the main equipment operation status of a large substation according to claim 1, characterized in that, The process by which the thermal dynamic inversion module determines the equivalent heat transfer coefficient and thermal time constant of the target main equipment based on the transient temperature response segment includes: Extract the temperature trajectories of the top oil temperature and the winding temperature as a function of time within the transient temperature response segment; The temperature trajectory is fitted with least squares, and the equivalent heat transfer coefficient is determined based on the steady-state temperature value reached after the switching. The thermal time constant is determined based on the rate at which the temperature trajectory approaches the steady-state temperature value after the switching.

6. The online monitoring and early warning system for the operating status of main equipment in large substations according to claim 1, characterized in that, The baseline comparison and attribution module includes: The baseline establishment submodule is used to cluster historical data according to the combination of cooling stages, load rate, and ambient temperature, and to establish historical baselines for the equivalent heat transfer coefficient and the thermal time constant for each cluster of operating conditions. The comparison attribution module is used to select the historical baseline corresponding to the operating condition cluster that matches the current operating condition, compare the equivalent heat transfer coefficient and the thermal time constant obtained by inversion with the historical baseline, and classify the current temperature anomaly according to the offset direction.

7. The online monitoring and early warning system for the operating status of main equipment in a large substation according to claim 6, characterized in that, The process by which the comparison attribution factor module classifies the current temperature anomaly includes: If the equivalent heat transfer coefficient obtained by inversion decreases relative to the historical baseline and the thermal time constant increases relative to the historical baseline, then the temperature anomaly is classified as a decrease in heat dissipation capacity. If the equivalent heat transfer coefficient obtained by inversion remains basically unchanged relative to the historical baseline and the top oil temperature continues to rise, then the temperature anomaly is attributed to internal latent defects generating heat. If the equivalent heat transfer coefficient and the thermal time constant obtained by inversion both fall within the tolerance band of the historical baseline, then the temperature anomaly is classified as a fluctuation under normal operating conditions.

8. The online monitoring and early warning system for the operating status of main equipment in large substations according to claim 1, characterized in that, The process by which the early warning module determines the early warning information corresponding to the root cause of the temperature anomaly based on the baseline comparison and the attribution classification results of the attribution module includes: Based on the results of attribution classification, the warning level, attribution category, and location of the anomaly in the early warning information are determined. Based on the attribution classification results, the following treatment recommendations are made: if the attribution classification result is a decrease in heat dissipation capacity, the treatment recommendations are directed to the maintenance of the cooling system and the overhaul of the filter and fan; if the attribution classification result is heat generation due to internal latent defects, the treatment recommendations are directed to further diagnosis of internal insulation or latent defects. The output includes the warning information, the attribution category, the location of the anomaly, and the handling recommendations.

9. The online monitoring and early warning system for the operating status of main equipment in a large substation according to claim 1, characterized in that, The online monitoring and early warning system for the main equipment operation status of large substations also includes: The thermal balance residual confirmation unit is used to perform secondary confirmation on the attribution classification results of the baseline comparison and attribution module, which determine that the heat is generated by internal latent defects, based on the thermal balance residual between the oil temperature data and the gas production rate data.

10. The online monitoring and early warning system for the operating status of main equipment in a large substation according to claim 9, characterized in that, The process of secondary verification of the attribution classification results by the thermal balance residual verification unit includes: Using the equivalent heat transfer coefficient obtained from the inversion and the current oil temperature, the thermal power corresponding to the theoretical gas production rate under steady state is calculated. The measured additional heat power was obtained by converting the measured gas production rate. The difference between the thermal power corresponding to the theoretical gas production rate and the measured additional thermal power is determined as the thermal balance residual. If the thermal balance residual exceeds the predetermined range and the equivalent heat transfer coefficient obtained by inversion does not decrease significantly relative to the historical baseline, then it is confirmed that the temperature anomaly originates from heat generation due to internal latent defects rather than a decrease in heat dissipation capacity.

Images