Intelligent box transformer remote monitoring and optimization control method based on electric pole system

Abstract

The application discloses an intelligent box transformer remote monitoring and optimization control method based on an electric Hong system, which comprises the following steps: (1) hardware construction: the hardware comprises an electric Hong operating system terminal, a wireless communication module, a plurality of electric Hong sensors, a protection control device and a cloud server, and each electric Hong sensor and the protection control device are arranged in a box-type substation; (2) data acquisition and data transmission; (3) cloud preprocessing; (4) cloud modeling and decision: after comparison, whether real-time electric parameter data, non-electric temperature data and non-electric atlas data are risk data is judged; (5) control execution and feedback: the cloud server issues the generated optimization control strategy to the electric Hong operating system terminal, remotely modifies action parameters of a local protection device and controls the local protection device to execute opening and closing operation, and operation risks are predicted in advance. The application can optimize and remotely control, improves power supply reliability and efficiency, and reduces manual operation and maintenance time.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power equipment, and particularly relates to an intelligent box transformer remote monitoring and optimization control method based on an electric Hong system. BACKGROUND

[0002] The box transformer station (referred to as "box transformer") as an important power distribution facility in the power system is widely used in the process of power transmission and distribution. The core components of the box transformer station include a high-voltage ring network cabinet, a transformer room, a low-voltage cabinet and an environmental monitoring assembly. The traditional box transformer station monitoring and control method mainly relies on local equipment and manual inspection, although it can realize the monitoring and control of the equipment state to a certain extent, but still has many shortcomings:

[0003] (1) There are problems of untimely monitoring, slow fault response, inability to perform global optimization, etc., resulting in low maintenance efficiency, high operation and maintenance cost, and slow recovery time;

[0004] (2) The monitoring and control method mainly relies on manual on-site monitoring parameters and local control, and manual monitoring is slow, time-consuming, single technical means, low accuracy, non-domestic independent technical property and controllable, etc.;

[0005] (3) When the abnormality or failure of each key device in the box transformer station occurs, it cannot be monitored, maintained, analyzed and diagnosed in real time, but can only be identified through regular manual maintenance. SUMMARY

[0006] The technical problem to be solved by the present application is to provide an intelligent box transformer remote monitoring and optimization control method based on an electric Hong system, which can optimize and remotely control, improve power supply reliability and efficiency, reduce manual operation and maintenance time, and strengthen the technical property of domestic independent safety and controllability.

[0007] In order to solve the above technical problems, the technical scheme adopted by the present application is as follows:

[0008] An intelligent box transformer remote monitoring and optimization control method based on an electric Hong system, characterized by comprising the following steps:

[0009] (1) Hardware construction: the hardware includes an electric Hong operating system terminal, a wireless communication module, a plurality of electric Hong sensors, a protection control device and a cloud server, and each electric Hong sensor and protection control device is arranged in the box transformer station;

[0010] (2) Data acquisition and data transmission: the electric Hong operating system terminal communicates with each electric Hong sensor in real time through the wireless communication module, continuously acquires the electric parameter data, non-electric temperature data and non-electric atlas data of the box-type substation; and the electric Hong operating system terminal stores, transmits and displays the acquired electric parameter data, non-electric temperature data and non-electric atlas data locally, and uploads them to the cloud server in real time through the network;

[0011] (3) Cloud preprocessing: after the cloud server receives the electric parameter data, non-electric temperature data and non-electric atlas data, it preprocesses the electric parameter data, non-electric temperature data and non-electric atlas data, presets alarm data with upper and lower floating values for the electric parameter data and non-electric temperature data, and presets alarm atlas for non-electric atlas data;

[0012] (4) Cloud modeling and decision-making: the cloud server pre-establishes a box transformer operation model, and then monitors the preprocessed electric parameter data, non-electric temperature data and non-electric atlas data in real time, compares the real-time collected electric parameter data, non-electric temperature data with the preset alarm data, compares the real-time collected non-electric atlas data with the preset alarm atlas, and judges whether the real-time electric parameter data, non-electric temperature data and non-electric atlas data are risk data after comparison:

[0013] (4-1) If the real-time electric parameter data and non-electric temperature data are within the upper and lower floating value range of the alarm data, the real-time electric parameter data and non-electric temperature data are not risk data;

[0014] If the real-time non-electric atlas data is consistent with the alarm atlas, the real-time non-electric atlas data is not risk data;

[0015] And these real-time electric parameter data, non-electric temperature data and non-electric atlas data are re-sent to the box transformer operation model for optimization training, so that the box transformer operation model continuously learns these data without risk;

[0016] (4-2) If the real-time electric parameter data and non-electric temperature data exceed or are lower than the alarm data, the real-time electric parameter data and non-electric temperature data are risk data;

[0017] If the real-time non-electric atlas data is inconsistent with the alarm atlas, the real-time non-electric atlas data is risk data;

[0018] After determining the risk data, the future operation state and possible fault risk of the box-type substation are determined, and the corresponding optimization control strategy is generated and step (5) is executed;

[0019] (5) Control execution and feedback: the cloud server will issue the generated optimization control strategy to the electric Hong operating system terminal, remotely modify the action parameters of the local protection device and control the local protection device to perform the closing and opening operation, predict the operation risk in advance, and prevent equipment damage.

[0020] The above-mentioned electric Hong operating system terminal and electric Hong sensor both use domestic self-controlled electric Hong operating system, support plug and play and wireless communication.

[0021] In the preferred scheme, in step (2), the electric parameter data includes: voltage, current, phase angle, frequency, power, and partial discharge data of the high-voltage chamber; voltage, current, phase angle, frequency, and power data of the low-voltage chamber; incoming and outgoing line current, grounding current, and partial discharge data of the transformer chamber; non-electric quantity temperature data includes: cable head temperature data of the high-voltage chamber; incoming and outgoing line busbar temperature data of the low-voltage chamber; incoming and outgoing line busbar temperature, winding temperature, and contact temperature data of the transformer chamber; non-electric quantity map data includes: SF6 gas density data of the high-voltage chamber; oil level information and overall vibration data of the transformer chamber.

[0022] The above-mentioned electric parameter data mainly reflects the electrical characteristics (such as voltage and current size, phase relationship, power transmission, and partial discharge) in the operation of the box-type substation. The above-mentioned non-electric quantity temperature data and non-electric quantity map data reflect the physical state of the equipment (such as temperature, gas density, and liquid level), which is crucial for the safe operation of the equipment, although these parameters do not directly participate in the electrical energy transmission.

[0023] In the preferred scheme, in step (3), the electric parameter data and non-electric quantity temperature data are preprocessed, which only needs to directly obtain the electric parameter data and non-electric quantity temperature data and compare them with the preset alarm data.

[0024] In the preferred scheme, in step (3), the non-electric quantity map data is preprocessed as follows:

[0025] (3-1) Data cleaning and equipment characteristic adaptation: according to the characteristics of the SF6 gas density data of the high-voltage chamber, the oil level information of the transformer chamber, and the overall vibration data, identify and correct data missing, abnormal fluctuations, and repeated records; for linear change data, use the device physical characteristic model for filling to ensure that the data conforms to the actual operation law of the equipment, and provide a true basis for subsequent comparison with the alarm map;

[0026] (3-2) Data integration and map conversion: merge the cleaned data according to the equipment type and monitoring dimension; and convert it into a standardized map format, so that it is comparable with the preset alarm map;

[0027] (3-3) Feature extraction and risk dimension mapping: extract key features from the graphed data, and map these key features one by one with the core discriminant dimensions of the preset alarm graph, providing quantitative basis for subsequent real-time graph and alarm graph comparison;

[0028] (3-4) Data scaling and standardization calibration: dimension unification is performed on the extracted key features, and standardization calibration is performed according to the differences between device models, ensuring that non-electric quantity graph data of different transformers can be compared with alarm graphs based on unified standards, while adapting to the input requirements of transformer operation models;

[0029] (3-5) Data diversity and scene matching division: the standardized data is divided into training data set, verification data set and test data set; wherein, the training data set is used to optimize the preset alarm graph; the verification data set is used to check the learning effect of the transformer operation model on the non-electric quantity graph data; the test data set is used to simulate real-time monitoring scene, verify whether the preprocessed data can accurately support risk data judgment, and finally provide high-quality data input for the comparison and model training of the following step (4).

[0030] The standardized graph format in the above step (3-2) can convert the change of SF6 density over time into a trend graph, convert the transformer room vibration signal from time domain to frequency domain spectrum graph, and convert the oil level information into a liquid level ratio graph associated with the oil tank volume. The above alarm graph can be an SF6 density drop graph, an abnormal vibration spectrum, and an oil level below threshold graph.

[0031] In a preferred embodiment, in step (4), the transformer operation model is constructed based on electric quantity data, non-electric quantity temperature data and non-electric quantity graph data during normal operation of the transformer substation, including high-voltage chamber electric quantity and cable head temperature model, transformer chamber non-electric quantity model, low-voltage chamber electric quantity and cable head temperature model. These three models are adapted to the core monitoring requirements of high-voltage chamber, transformer chamber and low-voltage chamber, respectively.

[0032] In a further preferred embodiment, in step (4), the high-voltage chamber electric quantity and cable head temperature model is used to train and compare electric quantity data, non-electric quantity temperature data cable head temperature, non-electric quantity graph data SF6 density and partial discharge;

[0033] The non-electric quantity model of the transformer chamber is used to train and compare incoming and outgoing line busbar temperature, winding temperature, contact temperature and overall vibration.

[0034] The low-voltage chamber electric quantity and cable head temperature model is used to train and compare incoming and outgoing line temperature of non-electric quantity temperature data, and partial discharge of non-electric quantity graph data.

[0035] The comparison of the above-mentioned electrical parameter data is performed by using a short-circuit current formula and a low voltage formula; the comparison of the above-mentioned non-electrical temperature data is performed by using a temperature rise formula model; and the comparison of the above-mentioned non-electrical map data is performed by using a gap discharge model, a statistical model and an equivalent circuit model.

[0036] In a further preferred scheme, in the step (4), the risk data includes: overlarge load current, low voltage, low frequency, high frequency, negative sequence characteristics, low SF6 gas pressure, overhigh cable head temperature, overhigh low-voltage bus temperature, partial discharge alarm and overhigh indoor humidity in the low-voltage chamber and the high-voltage chamber; overlarge incoming and outgoing line current, overlarge grounding current, overhigh incoming and outgoing line bus temperature, partial discharge alarm and abnormal oil level in the transformer chamber. In the step (4), the box transformer operation model is the data of a normally operating box-type substation. For example, (1) risk determination of temperature data: comparing the real-time non-electrical temperature data with the characteristic values of the temperature rise formula model in the box transformer operation model, if the real-time non-electrical temperature data exceeds or is lower than the characteristic values of the temperature rise formula model, it is determined as risk data; (2) risk determination of partial discharge: extracting the characteristic values of the discharge map, comparing the real-time non-electrical map data with the characteristic values of the discharge models such as sharp-end discharge and gap discharge saved in the box transformer operation model, and determining whether the discharge type is met; (3) risk determination of voltage and current: comparing the real-time electrical parameter data with the set voltage and current change curve of the box transformer operation model itself, comparing with the real-time load of the season and the morning and evening time, and determining whether the change rate is exceeded; (4) risk determination of vibration data: extracting the characteristic values of the vibration curve, comparing the real-time non-electrical map data with the map of the transformer chamber vibration curve in the box transformer operation model, and determining whether the threshold value of the vibration curve is exceeded.

[0037] In a further preferred scheme, in the step (4), the optimization control strategy includes:

[0038] (4-1) when the real-time load current or power of the low-voltage side switch is continuously overlarge, performing load shedding operation according to the preset priority of the load, and when the load shedding operation cannot meet the requirements, performing complete tripping of the low-voltage side by tripping the entire group of switches in the low-voltage chamber;

[0039] (4-2) when the transformer chamber has partial discharge alarm, abnormal oil level or overlarge grounding current, controlling the switch connected to the transformer chamber on the high-voltage side to trip and cut off the load;

[0040] (4-3) adjusting the thermal protection parameters and the overload parameters under the safety margin according to the fluctuation of the real-time data of the seasonal operation.

[0041] In the step (5), the local protection device includes a high-voltage chamber protection device and a low-voltage chamber protection device; the high-voltage chamber protection device receives the control switch opening and closing signal and the optimized protection parameter setting according to the optimized control strategy generated by the cloud server; and the low-voltage chamber protection device receives the load shedding operation of the control switch and the optimized protection parameter setting according to the optimized control strategy generated by the cloud server. The high-voltage chamber protection device can be a safe controllable distribution network automation terminal. The low-voltage chamber protection device can be a molded case circuit breaker.

[0042] In the preferred solution, the step (6) further includes a risk elimination judgment: if the risk is eliminated, the process is ended; and if the risk is not eliminated, the risk is sent to the cloud server for manual intervention to eliminate the risk.

[0043] In the further preferred solution, in the step (6), the manual intervention is to eliminate the risk by the following two methods:

[0044] (6-1) cutting off the load at the end;

[0045] (6-2) removing the entire set of risk components.

[0046] If the following risks occur, manual intervention can be performed:

[0047] (1) when the temperature of the low-voltage chamber outlet line is continuously too high, and the temperature change rate exceeds the threshold value, the end load is usually too large, and an alarm can be performed first; if the above phenomenon continues, the low-voltage intelligent molded case switch (i.e. the protection device) is controlled to open and cut off the low-voltage load;

[0048] (2) when the local discharge or vibration of the transformer room is abnormal, on one hand, the voltage and current of the low-voltage chamber are combined to determine whether the load is too high: if it is not caused by the load being too high, it is a local abnormality, and an alarm can be performed first; if it is caused by the load being too high, the protection device of the high-voltage chamber outlet line to the transformer room is controlled to open and cut off the load of the transformer room; if the voltage and current of the low-voltage chamber are abnormal at the same time, the low-voltage intelligent molded case switch (i.e. the protection device) is controlled to open and cut off the low-voltage load;

[0049] (3) when the voltage and current risk is abnormal, the overload protection parameter needs to be adjusted according to the seasonal and morning and evening trend, and an alarm is issued; if the change trend continues to exceed the threshold value and is not processed, the load is cut off. The overload protection parameter is adjusted to consider the sensitivity of the overload protection, and the overload protection parameter is adjusted according to the season to ensure the sensitivity of the overload alarm and tripping.

[0050] Compared with the prior art, the present application has the following advantages:

[0051] The electric Hong operation system terminal (edge side) of the application is connected to the cloud server (cloud side) for data interaction, realizes data uploading and control instruction receiving, and is connected to each electric Hong sensor (end side) for real-time collection of electric parameter data and non-electric quantity data of the box-type substation. The application is based on the electric Hong Internet of Things operation system, constructs a "cloud-edge-end" collaborative architecture, makes the entire box-type substation system have real-time interaction, real-time intelligent diagnosis and risk prediction, has high identification accuracy, can remotely control and optimize protection parameters when abnormal operation parameters occur, improves the accuracy and sensitivity of line protection, and guarantees operation reliability.

[0052] (2) The application can accelerate the evolution of traditional box-type substations to intelligent, green and diversified directions, realizes value chain reconstruction through technology fusion and technological innovation, improves the work efficiency and service quality of operation and maintenance personnel, improves operation and maintenance efficiency, reduces cost, guarantees the stability of the power grid and the stability of the power supply system, and provides strong technical support for building more intelligent, green and efficient power infrastructure in the future. BRIEF DESCRIPTION OF DRAWINGS

[0053] Figure 1 is the control flowchart of the embodiment of the application. DETAILED DESCRIPTION

[0054] The application will be further described below in combination with the drawings and the preferred embodiments of the application.

[0055] As shown in the drawings, Figure 1 The intelligent box-type substation remote monitoring and optimization control method based on the electric Hong system in the embodiment includes the following steps:

[0056] (1) Hardware construction: the hardware includes an electric Hong operation system terminal, a wireless communication module, a plurality of electric Hong sensors, a protection control device and a cloud server, and each electric Hong sensor and the protection control device are arranged in the box-type substation;

[0057] (2) Data collection and data transmission: the electric Hong operation system terminal communicates with each electric Hong sensor in real time through the wireless communication module, continuously collects electric parameter data, non-electric quantity temperature data and non-electric quantity map data of the box-type substation, and the electric Hong operation system terminal locally stores, transmits and displays the collected electric parameter data, non-electric quantity temperature data and non-electric quantity map data, and uploads them to the cloud server in real time through the network;

[0058] (3) Cloud preprocessing: after receiving the electric parameter data, non-electric quantity temperature data and non-electric quantity map data, the cloud server pre-processes the electric parameter data, non-electric quantity temperature data and non-electric quantity map data, presets alarm data with upper and lower floating values for the electric parameter data and non-electric quantity temperature data, and presets an alarm map for the non-electric quantity map data;

[0059] (4) Cloud modeling and decision-making: The cloud server pre-establishes a box transformer operation model, and then monitors the pre-processed electrical parameter data, non-electrical temperature data and non-electrical atlas data in real time. The real-time collected electrical parameter data, non-electrical temperature data and preset alarm data are compared, and the real-time collected non-electrical atlas data and the preset alarm atlas are compared. After comparison, it is judged whether the real-time electrical parameter data, non-electrical temperature data and non-electrical atlas data are risk data:

[0060] (4-1) If the real-time electrical parameter data and non-electrical temperature data are within the upper and lower floating value range of the alarm data, the real-time electrical parameter data and non-electrical temperature data are not risk data;

[0061] If the real-time non-electrical atlas data is consistent with the alarm atlas, the real-time non-electrical atlas data is not risk data;

[0062] And these real-time electrical parameter data, non-electrical temperature data and non-electrical atlas data are re-sent to the box transformer operation model for optimization training, so that the box transformer operation model continuously learns these non-risk data;

[0063] (4-2) If the real-time electrical parameter data and non-electrical temperature data exceed or are lower than the alarm data, the real-time electrical parameter data and non-electrical temperature data are risk data;

[0064] If the real-time non-electrical atlas data is inconsistent with the alarm atlas, the real-time non-electrical atlas data is risk data;

[0065] After determining the risk data, the future operation state and possible fault risk of the box-type transformer substation are determined, and the corresponding optimization control strategy is generated and step (5) is executed;

[0066] (5) Control execution and feedback: The cloud server issues the generated optimization control strategy to the electric Hong operating system terminal, remotely modifies the action parameters of the local protection device, and controls the local protection device to perform opening and closing operations. Early prediction of operation risks can prevent equipment damage.

[0067] The above-mentioned electric Hong operating system terminal and electric Hong sensor both use domestic self-controlled electric Hong operating system, support plug and play and wireless communication.

[0068] In step (2), the electric parameter data includes: voltage, current, phase angle, frequency, power, and partial discharge data of the high-voltage chamber; voltage, current, phase angle, frequency, and power data of the low-voltage chamber; incoming and outgoing line current, grounding current, and partial discharge data of the transformer chamber; the non-electric quantity temperature data includes: cable head temperature data of the high-voltage chamber; incoming and outgoing line busbar temperature data of the low-voltage chamber; incoming and outgoing line busbar temperature, winding temperature, and contact temperature data of the transformer chamber; the non-electric quantity atlas data includes: SF6 gas density data of the high-voltage chamber; oil level information and overall vibration data of the transformer chamber.

[0069] The above electric parameter data mainly reflects the electrical characteristics (such as voltage and current size, phase relationship, power transmission, and partial discharge electrical phenomena) in the operation of the box-type substation. The above non-electric quantity temperature data and non-electric quantity atlas data reflect the physical state (such as temperature, gas density, and liquid level) of the equipment. These parameters do not directly participate in electrical energy transmission, but are crucial to the safe operation of the equipment.

[0070] In step (3), the electric parameter data and the non-electric quantity temperature data are preprocessed by directly obtaining the electric parameter data and the non-electric quantity temperature data and comparing them with the preset alarm data.

[0071] In step (3), the preprocessing steps of the non-electric quantity atlas data are as follows:

[0072] (3-1) Data cleaning and equipment characteristic adaptation: according to the characteristics of the SF6 gas density data of the high-voltage chamber, the oil level information of the transformer chamber, and the overall vibration data, identify and correct data missing, abnormal fluctuations, and repeated records; for linearly changing data, use the device physical characteristic model to fill in, ensure that the data conforms to the actual operation law of the equipment, and provide a true basis for subsequent comparison with the alarm atlas;

[0073] (3-2) Data integration and atlas conversion: merge the cleaned data by equipment type and monitoring dimension; and convert it into a standardized atlas format, so that it is comparable with the preset alarm atlas;

[0074] (3-3) Feature extraction and risk dimension mapping: extract key features from the atlas data, so that these key features are one-to-one mapped with the core discriminant dimensions of the preset alarm atlas, providing a quantitative basis for subsequent real-time atlas and alarm atlas comparison;

[0075] (3-4) Data scaling and standardization calibration: unify the dimensions of the extracted key features, and perform standardization calibration according to the differences in equipment models, to ensure that the non-electric quantity atlas data of different box transformers can be compared with the alarm atlas based on a unified standard, while adapting to the input requirements of the box transformer operation model;

[0076] (3-5) Data distribution and scenario matching division: the standardized data is divided into a training data set, a verification data set, and a test data set; wherein, the training data set is used to optimize the preset alarm atlas; the verification data set is used to check the learning effect of the box transformer operation model on the non-electric quantity atlas data; the test data set is used to simulate a real-time monitoring scenario, verify whether the preprocessed data can accurately support risk data judgment, and finally provide high-quality data input for the comparative analysis and model training of the following step (4).

[0077] The standardized atlas format in the above step (3-2) can convert the change of SF6 density over time into a trend atlas, convert the transformer room vibration signal from the time domain to the frequency domain spectrum atlas, and convert the oil level information into a liquid level ratio atlas associated with the oil tank volume. The above alarm atlas can be an SF6 density drop atlas, an abnormal vibration spectrum, and an oil level threshold atlas.

[0078] In step (4), the box transformer operation model is constructed based on the electric quantity data, non-electric quantity temperature data, and non-electric quantity atlas data of the normal operation of the box transformer substation, including the electric quantity and cable head temperature model of the high-voltage chamber, the non-electric quantity model of the transformer chamber, and the electric quantity and cable head temperature model of the low-voltage chamber. These three models are adapted to the core monitoring needs of the high-voltage chamber, the transformer chamber, and the low-voltage chamber, respectively.

[0079] In step (4), the electric quantity and cable head temperature model of the high-voltage chamber is used to train and compare the cable head temperature of the electric quantity data, non-electric quantity temperature data, and SF6 density and partial discharge of non-electric quantity atlas data.

[0080] The non-electric quantity model of the transformer chamber is used to train and compare the incoming and outgoing line busbar temperature, winding temperature, contact temperature, and overall vibration.

[0081] The electric quantity and cable head temperature model of the low-voltage chamber is used to train and compare the incoming and outgoing line temperature of the electric quantity data and non-electric quantity temperature data, and the partial discharge of non-electric quantity atlas data.

[0082] The comparison of the above electric quantity data uses short-circuit current formula and low-voltage formula; the comparison of the above non-electric quantity temperature data uses temperature rise formula model; and the comparison of the above non-electric quantity atlas data uses air gap discharge model, statistical model, and equivalent circuit model.

[0083] In step (4), the risk data includes: excessive load current, low voltage, low frequency, high frequency, negative sequence characteristics, low SF6 gas pressure, excessive cable head temperature, excessive low voltage busbar temperature, partial discharge alarm, and excessive indoor humidity in the low voltage and high voltage rooms; excessive incoming and outgoing line current, excessive grounding current, excessive incoming and outgoing busbar temperature, partial discharge alarm, and abnormal oil level in the transformer room. In step (4) above, the above-mentioned box-type substation operation model is the data of a normally operating box-type substation. For example, (1) Risk judgment of temperature data: compare the real-time non-electrical temperature data with the characteristic value of the temperature rise formula model in the transformer substation operation model. If the real-time non-electrical temperature data exceeds or falls below the characteristic value of the temperature rise formula model, it is judged as risk data; (2) Risk judgment of partial discharge: by extracting the characteristic value of the discharge spectrum, compare the real-time non-electrical spectrum data with the characteristic value of the discharge model such as tip discharge and gap discharge stored in the transformer substation operation model to determine whether the discharge type is met; (3) Risk judgment of voltage and current: compare the real-time electrical parameter data with the set voltage and current change curve of the transformer substation operation model itself, and compare it with the real-time seasonal and morning and evening loads to determine whether the rate of change is exceeded; (4) Risk judgment of vibration data: by extracting the characteristic value of the vibration curve, compare the real-time non-electrical spectrum data with the spectrum of the transformer room vibration curve in the transformer substation operation model to determine whether the threshold of the vibration curve is exceeded.

[0084] In step (4), the optimized control strategy includes:

[0085] (4-1) When the real-time load current or power of the low-voltage side switch is too large, the load shedding operation is performed according to the preset load priority. If the load shedding operation cannot meet the requirements, the entire low-voltage room is switched off to completely disconnect the low-voltage side.

[0086] (4-2) When a partial discharge alarm is triggered in the transformer room, the oil level is abnormal, or the grounding current is too large, the switch connected to the transformer room on the high-voltage side is tripped to disconnect the load.

[0087] (4-3) Adjust the thermal protection parameters and overload parameters within the safety margin based on the fluctuation of real-time seasonal operation data.

[0088] In step (5), the local protection device includes a high-voltage compartment protection device and a low-voltage compartment protection device. The high-voltage compartment protection device receives control switch opening and closing signals and optimized protection parameter settings based on the optimized control strategy generated by the cloud server. The low-voltage compartment protection device receives control switch load shedding operations and optimized protection parameter settings based on the optimized control strategy generated by the cloud server. The aforementioned high-voltage compartment protection device can be a safe and controllable distribution network automation terminal. The aforementioned low-voltage compartment protection device can be a molded case circuit breaker.

[0089] The embodiment also includes step (6) risk elimination judgment: if the risk is eliminated, end; if the risk is not eliminated, send the risk to the cloud server, and eliminate the risk by manual intervention.

[0090] In step (6), the manual intervention is to eliminate the risk by the following two methods:

[0091] (6-1) cutting off the load of the terminal load;

[0092] (6-2) removing the entire set of risk components.

[0093] If the following risks are encountered, manual intervention can be performed:

[0094] (1) When the temperature of the low-voltage chamber outlet is continuously too high for a short time, and the temperature change rate exceeds the threshold value, at this time, the terminal load is usually too large, and alarm can be performed first; if the above phenomenon continues, control the low-voltage intelligent plastic shell switch (i.e. protection device) to trip and cut off the low-voltage load;

[0095] (2) When the local discharge or vibration of the transformer room is abnormal, on the one hand, first determine whether it is caused by high load according to the voltage and current of the low-voltage chamber: if it is not caused by high load, it is a local abnormality, and alarm can be performed first; if it is caused by high load, control the protection device from the high-voltage chamber outlet to the transformer room to trip and cut off the transformer room load; if the voltage and current of the low-voltage chamber are abnormal at the same time, control the low-voltage intelligent plastic shell switch (i.e. protection device) to trip and cut off the low-voltage load;

[0096] (3) When the voltage and current risk is abnormal, the overload protection parameters need to be adjusted according to the seasonal and morning and evening trend changes, and an alarm is issued, if the change trend continues to exceed the threshold value and is not processed, the load is cut off. The above adjustment of overload protection parameters is to consider the sensitivity of overload protection, and the overload protection parameters are adjusted according to the season, which can ensure the sensitivity of overload alarm and tripping.

[0097] In addition, it should be noted that the specific embodiments described in the specification can have different names and the like, and any equivalent or simple changes made according to the structure, features and principles of the patent concept described herein are included in the protection scope of the patent. Those skilled in the art can make various modifications or supplements to the described specific embodiments or use similar ways to replace them, as long as they do not deviate from the structure of the invention or exceed the scope defined by the claims, and they should belong to the protection scope of the invention.

Claims

1. A method for remote monitoring and optimized control of intelligent prefabricated substations based on a power supply system, characterized in that... Includes the following steps: (1) Hardware construction: The hardware includes the Dianhong operating system terminal, wireless communication module, multiple Dianhong sensors, protection and control devices and cloud server. All Dianhong sensors and protection and control devices are installed inside the box-type substation. (2) Data acquisition and data transmission: The Dianhong operating system terminal communicates with each Dianhong sensor in real time through the wireless communication module to continuously collect electrical parameter data, non-electrical temperature data and non-electrical spectrum data of the box-type substation; and the Dianhong operating system terminal stores, transmits and displays the collected electrical parameter data, non-electrical temperature data and non-electrical spectrum data locally, and uploads them to the cloud server in real time through the network. (3) Cloud preprocessing: After receiving electrical parameter data, non-electrical temperature data and non-electrical spectrum data, the cloud server preprocesses the electrical parameter data, non-electrical temperature data and non-electrical spectrum data, presets alarm data with upper and lower fluctuation values ​​for electrical parameter data and non-electrical temperature data, and presets alarm spectrum for non-electrical spectrum data. (4) Cloud-based modeling and decision-making: The cloud server pre-establishes a transformer substation operation model, and then monitors the pre-processed electrical parameter data, non-electrical temperature data, and non-electrical graph data in real time. The real-time collected electrical parameter data and non-electrical temperature data are compared with the preset alarm data, and the real-time collected non-electrical graph data is compared with the preset alarm graph. After comparison, it is determined whether the real-time electrical parameter data, non-electrical temperature data, and non-electrical graph data are risk data. (4-1) If the real-time electrical parameter data and non-electrical temperature data are within the range of the alarm data fluctuation, then the real-time electrical parameter data and non-electrical temperature data are not risk data. If the real-time non-electrical quantity data is consistent with the alarm data, then the real-time non-electrical quantity data is not risky data. These real-time electrical parameter data, non-electrical temperature data, and non-electrical spectrum data are then resent to the transformer substation operation model for optimization training, enabling the transformer substation operation model to continuously learn from these risk-free data. (4-2) If the real-time electrical parameter data and non-electrical temperature data exceed or fall below the alarm data, then the real-time electrical parameter data and non-electrical temperature data are risk data. If the real-time non-electrical quantity data is inconsistent with the alarm data, then the real-time non-electrical quantity data is considered risky data. After the risk data is determined, the future operating status and possible fault risks of the prefabricated substation are predicted, and corresponding optimization control strategies are generated and steps (5) are executed. (5) Control execution and feedback: The cloud server will send the generated optimized control strategy to the Dianhong operating system terminal to remotely modify the action parameters of the local protection device and control the local protection device to perform opening and closing operations, predict operational risks in advance, and prevent equipment damage.

2. The intelligent transformer substation remote monitoring and optimization control method based on the Dianhong system as described in claim 1, characterized in that: In step (2), the electrical parameter data includes: voltage, current, phase angle, frequency, power, and partial discharge of the high-voltage chamber; voltage, current, phase angle, frequency, and power of the low-voltage chamber; incoming and outgoing line current, grounding current, and partial discharge of the transformer chamber; non-electrical quantity temperature data includes: cable head temperature data of the high-voltage chamber; incoming and outgoing busbar temperature data of the low-voltage chamber; incoming and outgoing busbar temperature, winding temperature, and contact temperature of the transformer chamber; non-electrical quantity spectrum data includes: SF6 gas density data of the high-voltage chamber; oil level information and overall vibration of the transformer chamber.

3. The intelligent prefabricated substation remote monitoring and optimization control method based on the Dianhong system as described in claim 1, characterized in that: In step (3), the electrical parameter data and non-electrical temperature data are preprocessed. It is only necessary to directly obtain the electrical parameter data and non-electrical temperature data and compare them with the preset alarm data.

4. The intelligent transformer substation remote monitoring and optimization control method based on the Dianhong system as described in claim 1, characterized in that: In step (3), the preprocessing steps for the non-electrical quantity spectrum data are as follows: (3-1) Data cleaning and equipment feature adaptation: Based on the characteristics of SF6 gas density data in the high-voltage room, oil level information in the transformer room, and overall vibration data, identify and correct missing data, abnormal fluctuations, and duplicate records; for linearly changing data, use the equipment physical characteristic model to fill in the data to ensure that the data conforms to the actual operating law of the equipment and provide a real basis for subsequent comparison with alarm spectrum. (3-2) Data integration and graph conversion: The cleaned data is merged according to the equipment type and monitoring dimension; and converted into a standardized graph format so that it is comparable to the preset alarm graph. (3-3) Feature extraction and risk dimension mapping: Extract key features from the graph data and map these key features one by one with the core discrimination dimensions of the preset alarm graph, so as to provide a quantitative basis for subsequent comparison between the real-time graph and the alarm graph; (3-4) Data scaling and standardization calibration: The key features extracted are standardized in terms of dimensions, and the non-electrical quantity spectrum data of different transformer substations are standardized according to the differences in equipment models to ensure that the non-electrical quantity spectrum data of different transformer substations can be compared with the alarm spectrum based on a unified standard, while adapting to the input requirements of the transformer substation operation model; (3-5) Data set division and scenario matching: The standardized data is divided into training dataset, validation dataset and test dataset; among them, the training dataset is used to optimize the preset alarm map; the validation dataset is used to verify the learning effect of the transformer substation operation model on non-electrical power map data; The test dataset is used to simulate real-time monitoring scenarios, verify whether the preprocessed data can accurately support risk data judgment, and ultimately provide high-quality data input for the comparative analysis and model training in the following step (4).

5. The intelligent prefabricated substation remote monitoring and optimization control method based on the Dianhong system as described in claim 1, characterized in that: In step (4), the box-type substation operation model is constructed based on the electrical parameter data, non-electrical quantity temperature data and non-electrical quantity spectrum data during normal operation of the box-type substation. It includes the electrical parameter and cable head temperature model of the high-voltage room, the non-electrical quantity model of the transformer room, and the electrical parameter and cable head temperature model of the low-voltage room. These three models are adapted to the core monitoring requirements of the high-voltage room, the transformer room and the low-voltage room respectively.

6. The intelligent prefabricated substation remote monitoring and optimization control method based on the Dianhong system as described in claim 5, characterized in that: In step (4), the electrical parameters of the high-voltage chamber and the cable head temperature model are used to train and compare the cable head temperature of electrical parameter data and non-electrical temperature data, and the SF6 density and partial discharge of non-electrical spectrum data. The non-electrical quantity model of the transformer room is used to train and compare the temperature of the incoming and outgoing busbars, winding temperature, contact temperature, and overall vibration. The electrical parameters and cable head temperature models of the low-voltage chamber are used to train and compare the electrical parameter data, the incoming and outgoing line temperatures of non-electrical quantity temperature data, and the partial discharge of non-electrical quantity spectrum data.

7. The intelligent prefabricated substation remote monitoring and optimization control method based on the Dianhong system as described in claim 5, characterized in that: In step (4), the risk data includes: excessive load current, low voltage, low frequency, high frequency, negative sequence characteristics, low SF6 gas pressure, excessive cable head temperature, excessive low voltage bus temperature, partial discharge alarm and excessive indoor humidity in the low voltage and high voltage rooms; excessive incoming and outgoing current, excessive grounding current, excessive incoming and outgoing bus temperature, partial discharge alarm and abnormal oil level in the transformer room.

8. The intelligent prefabricated substation remote monitoring and optimization control method based on the Dianhong system as described in claim 5, characterized in that: In step (4), the optimized control strategy includes: (4-1) When the real-time load current or power of the low-voltage side switch is too large, the load shedding operation is performed according to the load preset priority. If the load shedding operation cannot meet the requirements, the entire low-voltage room is switched off to completely disconnect the low-voltage side. (4-2) When a partial discharge alarm is triggered in the transformer room, the oil level is abnormal, or the grounding current is too large, the switch connected to the transformer room on the high-voltage side is tripped to disconnect the load. (4-3) Adjust the thermal protection parameters and overload parameters within the safety margin based on the fluctuation of real-time seasonal operation data.

9. The intelligent prefabricated substation remote monitoring and optimization control method based on the Dianhong system as described in claim 1, characterized in that: In step (5), the local protection device includes a high-voltage chamber protection device and a low-voltage chamber protection device; the high-voltage chamber protection device receives the control switch opening and closing signals and the optimized protection parameters according to the optimized control strategy generated by the cloud server. The low-pressure room protection device receives and sends control switch load shedding operations and optimized protection parameter settings based on the optimized control strategy generated by the cloud server.

10. The intelligent prefabricated substation remote monitoring and optimization control method based on the Dianhong system as described in claim 1, characterized in that: It also includes step (6) risk elimination judgment: if the risk is eliminated, the process ends; if the risk is not eliminated, the risk is sent to the cloud server for manual intervention to eliminate the risk. In step (6), human intervention is used to eliminate the risk through the following two methods: (6-1) Disconnect the load at the end of the load; (6-2) Remove the entire group of risky components.

Images