Power distribution station house equipment monitoring system
The power distribution station equipment monitoring system addresses the limitations of manual inspections by implementing a sensor-based, adaptive monitoring system with deep learning and energy harvesting, enhancing accuracy and reducing costs through real-time data management and adaptive data transfer.
Patent Information
- Application Number
- CN202411292421.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-14
- Publication Date
- 2024-12-20
AI Technical Summary
Traditional power distribution station management relies heavily on manual inspections, which are time-consuming and prone to human errors, leading to missed or false detections, resulting in equipment failures that cannot be timely addressed.
A power distribution station equipment monitoring system utilizing a perception layer with various sensors, a transmission layer for data transfer, and an application layer for data processing and storage, incorporating intelligent sensing technology, deep learning networks, and energy harvesting, to enable real-time monitoring and adaptive data management.
Enhances monitoring accuracy and reliability, reduces human error, optimizes resource allocation, and lowers operation and maintenance costs through real-time monitoring and decision-making support, while improving energy efficiency and system autonomy.
Smart Images
Figure CN119171627A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of power distribution station maintenance, and in particular to a power distribution station equipment monitoring system. Background Art
[0002] With the continuous development of power systems and the improvement of their intelligence levels, distribution stations, as a key link in power distribution, are crucial for ensuring the reliability of power supply through their safe and stable operation.
[0003] Traditional power distribution station management mainly relies on manual regular inspections, but this approach has many limitations. On the one hand, the manual inspection cycle is long, and it is difficult to grasp the operating status and environmental changes of the equipment in the power distribution station in real time; on the other hand, manual inspections are easily affected by human factors, and there may be problems such as missed inspections and false inspections, resulting in equipment failures that cannot be discovered and handled in a timely manner.
[0004] In recent years, with the rapid development of sensor technology, communication technology and computer technology, intelligent substation equipment monitoring system has gradually become a hot spot for research and application. By installing various sensors in the substation, collecting equipment operating status and environmental parameters in real time, and transmitting the data to the monitoring center for processing and analysis, it is possible to achieve remote real-time monitoring and intelligent management of the substation, timely discover equipment failures and environmental anomalies, and improve the operating reliability and safety of the substation.
[0005] In order to solve the above problems, the present invention proposes a distribution station equipment monitoring system. Summary of the invention
[0006] The purpose of the present invention is to provide a distribution station equipment monitoring system to solve the problems raised in the background technology: Traditional distribution station management and inspection cycles are long, making it difficult to grasp the operating status of equipment and environmental changes in the distribution station in real time. They are easily affected by human factors, and may result in problems such as missed detection and false detection, which may lead to equipment failures that cannot be discovered and handled in a timely manner.
[0007] In order to achieve the above object, the present invention adopts the following technical solutions: A distribution station equipment monitoring system, comprising: Perception layer: used to collect environmental parameters and equipment operating status information in the power distribution station based on various sensors and intelligent monitoring equipment; Transmission layer: used to transmit the data collected by the perception layer to the application layer. Both wired and wireless transmission methods can be used. The best transmission method is selected for data transmission based on the actual situation and needs of the power distribution station. Application layer: used for data integration, processing and storage, and monitoring the real-time operation of sensors and equipment in the distribution station; The perception layer includes: Environmental monitoring module: used to monitor the distribution station environment; Equipment monitoring module: used to monitor the equipment in the power distribution station; The environmental monitoring module includes a temperature monitoring unit, a humidity monitoring unit, a gas monitoring unit, a water level monitoring unit, a flood monitoring unit, a smoke monitoring unit, an access control monitoring unit, an open fire monitoring unit, an electronic fence monitoring unit, an infrared dual-detection monitoring unit, a beam monitoring unit, and a camera monitoring unit; The equipment monitoring module includes a transformer monitoring unit, a busbar and cable pile temperature monitoring unit, and a switch cabinet monitoring unit; The application layer includes: Data processing module: used to process, analyze and store data uploaded by the transport layer; Equipment monitoring module: used for real-time monitoring, alarm and analysis of equipment and environment in the power distribution station; Database: used to store monitoring data and system configuration information.
[0008] Preferably, the various sensors of the sensing layer perform self-diagnosis, self-calibration and self-compensation based on intelligent sensing technology; The intelligent sensing technology performs spectrum analysis on the output signal, working parameters and historical fault data of the sensor based on a deep learning network, and uses the spectrum features as the input of a convolutional neural network to achieve fault diagnosis of the sensor; The deep learning network is based on the CNN network, which incorporates a multi-head attention mechanism and a multi-scale feature extraction module, is combined with a multi-layer Bi-LSTM network, and performs parameter optimization based on an optimization algorithm; The CNN network extracts and fuses multi-scale features of input data based on convolution kernels of sizes 3×3, 5×5, and 7×7, performs feature dimensionality reduction based on a fully connected layer, and then inputs the features into the attention module; The attention module maps the reduced-dimensional multi-scale features to obtain the query vector matrix of the input features , key vector matrix Sum vector matrix :
[0009]
[0010]
[0011] Output of the attention module as follows:
[0012]
[0013] in, , and All are learning parameters; is the query vector matrix in the adjacency space self-attention mechanism , key vector matrix Sum vector matrix The dimension of is the transpose of a vector; For multi-head output; A number for each head; Input the features extracted by the CNN network into the multi-layer Bi-LSTM network to obtain the feature vector sequence corresponding to each Bi-LSTM network , No. The hidden layer output of a Bi-LSTM is expressed as:
[0014]
[0015]
[0016] in, Indicates The forward hidden layer state of a Bi-LSTM; For the The forward weight parameters of a Bi-LSTM; Indicates The hidden layer output state of a Bi-LSTM; For the The forward bias parameters of a Bi-LSTM; Indicates The backward hidden layer state of a Bi-LSTM; For the The backward weight parameters of a Bi-LSTM; For the The backward bias parameter of a Bi-LSTM; Indicates The hidden layer output state of a Bi-LSTM; is the forward weight parameter of Bi-LSTM; is the backward weight parameter of Bi-LSTM; is the bias parameter of Bi-LSTM; The hidden layer of the last Bi-LSTM is connected to the encoding layer, and the encoding layer encodes the output of the hidden layer of the last Bi-LSTM. The encoding process is as follows:
[0017] in, is the encoding output; is a symbolic function; is the transpose of a vector; is the weight parameter of the encoding layer; is the bias parameter; Finally, the output coding anomaly detection is performed based on the SVM vector machine; The learning parameters in the deep learning network are all obtained based on the improved particle swarm algorithm; The improved particle swarm algorithm is specifically as follows:
[0018]
[0019]
[0020] in, , Respectively The parameter The second change and The speed of the next change; , For the The parameter The second change and The value of the iteration; is the inertia weight; , is the acceleration factor; , for A random variable uniformly distributed between is the individual optimal value; is the global optimal value; is the maximum number of iterations; is a random variable of weight parameter, with a value range of ; The intelligent sensing technology also performs self-calibration of the sensor based on an improved particle swarm algorithm; The intelligent sensing technology is also based on a deep neural network, which takes environmental factors and sensor input as the network input, and the sensor output error as the network output. By training the deep neural network, the error characteristics of the sensor are learned to achieve sensor output compensation.
[0021] Preferably, the sensor also adopts dynamic power management technology and software-defined networking to dynamically adjust the supply voltage and operating frequency of the sensor according to the monitoring task.
[0022] Preferably, the sensor centrally manages and controls the network connection of the monitoring device through a software-defined network controller, and dynamically adjusts the network topology, bandwidth and data transmission path according to the monitoring task.
[0023] Preferably, various equipment in the distribution station building use piezoelectric energy harvesting technology and electromagnetic induction energy harvesting technology to collect the vibration energy generated during the operation of the equipment and the alternating electromagnetic field energy in the distribution station building, and convert them into electrical energy to power the low-power perception layer and transmission layer.
[0024] Preferably, each monitoring unit in the equipment monitoring module determines the critical value according to the safe and stable operating environment factors required for different types of equipment, and also determines the critical value according to the safe and stable operating environment factors of different models of equipment of the same type; and thereby builds a distribution station equipment status monitoring model to perform real-time monitoring and early warning of the equipment status in the distribution station.
[0025] Compared with the prior art, the present invention provides a power distribution station equipment monitoring system, which has the following beneficial effects: The present invention improves monitoring accuracy and reliability through self-diagnosis, calibration and compensation of intelligent sensing technology, as well as deep learning networks and algorithms. Through technologies such as dynamic power management and software-defined networks, the flexibility and adaptability of the system are enhanced. Energy harvesting technology improves energy utilization efficiency and system autonomy. Comprehensive real-time monitoring and guarantee of safe and stable operation of distribution stations can also provide decision-making basis for operation and maintenance management, reduce inspection workload, optimize resource allocation, reduce operation and maintenance costs and improve economic benefits. BRIEF DESCRIPTION OF THE DRAWINGS
[0026] Figure 1 This is a system block diagram mentioned in Example 1 of the present invention.
[0027] Meaning of the symbols in the figure: 1. Perception layer; 11. Environmental monitoring module; 1101. Temperature monitoring unit; 1102. Humidity monitoring unit; 1103. Gas monitoring unit; 1104. Water level monitoring unit; 1105. Water immersion monitoring unit; 1106. Smoke monitoring unit; 1107. Access control monitoring unit; 1108. Open fire monitoring unit; 1109. Electronic fence monitoring unit; 1110. Infrared dual-detection monitoring unit; 1111. Opposite-beam monitoring unit; 1112. Camera monitoring unit; 12. Equipment monitoring module; 1201. Transformer monitoring unit; 1202. Busbar and cable pile temperature monitoring unit; 1203. Switch cabinet monitoring unit; 2. Transport layer; 3. Application layer; 31. Data processing module; 32. Equipment monitoring module; 33. Database. DETAILED DESCRIPTION
[0028] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.
[0029] The present invention improves monitoring accuracy and reliability through self-diagnosis, calibration and compensation of intelligent sensing technology, as well as deep learning networks and algorithms. Through technologies such as dynamic power management and software-defined networks, the flexibility and adaptability of the system are enhanced. Energy harvesting technology improves energy utilization efficiency and system autonomy. Comprehensive real-time monitoring and guarantee of safe and stable operation of distribution stations can also provide decision-making basis for operation and maintenance management, reduce inspection workload, optimize resource allocation, reduce operation and maintenance costs and improve economic benefits. Specifically include the following contents.
[0030] Embodiment 1: See also Figure 1 The present invention provides a power distribution station equipment monitoring system, comprising: Perception layer 1: It is composed of various sensors and intelligent monitoring equipment, responsible for collecting environmental parameters and equipment operation status information in the distribution station, such as temperature sensors, humidity sensors, gas sensors, water level sensors, water immersion sensors, smoke sensors, access control systems, cameras, current transformers, voltage transformers, partial discharge sensors, etc.
[0031] The perception layer 1 includes an environment monitoring module 11 and an equipment monitoring module 12. The environment monitoring module 11 includes a temperature monitoring unit 1101, a humidity monitoring unit 1102, a gas monitoring unit 1103, a water level monitoring unit 1104, a flood monitoring unit 1105, a smoke monitoring unit 1106, an access control monitoring unit 1107, an open fire monitoring unit 1108, an electronic fence monitoring unit 1109, an infrared dual-detection monitoring unit 1110, a beam monitoring unit 1111, and a camera monitoring unit 1112. The equipment monitoring module 12 includes a transformer monitoring unit (1201), a busbar and cable pile temperature monitoring unit (1202), and a switch cabinet monitoring unit (1203). The specific data of each monitoring unit can be referred to Table 1: Table 1 Parameters of each monitoring unit
[0032] Based on the deep neural network, the sensor of the monitoring unit is self-diagnosed, self-calibrated and self-compensated, and training and testing are performed based on multiple groups of data. The CNN network is used as model one, the CNN network integrating the multi-head attention mechanism and the multi-scale feature extraction module is used as model two, the CNN network + LSTM network integrating the multi-head attention mechanism and the multi-scale feature extraction module is used as model three, and the model of this embodiment is model four. The errors of the four models are statistically compared, and the specific results can be referred to Table 2: Table 2 Error results of different models
[0033] It can be seen from the table that model 4 of this embodiment has lower error and higher accuracy, and is more suitable for application in the diagnosis, self-calibration and self-compensation processes of the sensor.
[0034] Transport layer 2: It is mainly responsible for transmitting the data collected by the perception layer to the application layer. It can adopt wired transmission (such as Ethernet, RS485 bus, etc.) and wireless transmission (such as Zigbee, LoRa, NB-IoT, etc.). According to the actual situation and needs of the power distribution station, select the appropriate transmission method to ensure the stability and reliability of data transmission.
[0035] The transmission layer 2 feed adopts a combination of wired (Ethernet with a transmission rate of 100Mbps) and wireless (LoRa with a transmission frequency of 470-510MHz). Critical data (such as alarm data) is transmitted in real time, and non-critical data is transmitted every 5 minutes.
[0036] Application layer 3: The core of the power distribution plant equipment monitoring system, mainly including data processing servers, monitoring software and databases. The data processing server is responsible for processing, analyzing and storing the data uploaded by the transport layer; Dell PowerEdge R740 servers can be used for data processing, which can process 1,000 data per second and has a storage capacity of 10TB; The monitoring software can realize real-time monitoring, alarm management, data analysis, report generation and other functions of the equipment and environment in the power distribution station based on the self-developed power distribution station equipment monitoring software; The database may adopt MySQL database, which has a longer storage time and can store monitoring data and system configuration information.
[0037] Based on the above system, the environmental parameters and equipment operating status in the power distribution station are monitored in real time, and the data is accurate and stable. When the environmental parameters or equipment operating status exceed the set critical value, the system can issue an alarm signal in time and notify the operation and maintenance personnel through SMS, email, etc. The operation and maintenance personnel can check the operation status of the power distribution station anytime and anywhere through the monitoring software, which improves the operation and maintenance efficiency and management level.
[0038] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.
Claims
1. A distribution station equipment monitoring system, characterized in that: include: Perception layer (1): used to collect environmental parameters and equipment operating status information in the power distribution station based on various sensors and intelligent monitoring equipment; Transmission layer (2): used to transmit the data collected by the perception layer (1) to the application layer (3). Both wired transmission and wireless transmission can be used. The best transmission method is selected for data transmission based on the actual situation and needs of the power distribution station. Application layer (3): used for data integration, processing and storage, and monitoring the real-time operation of sensors and equipment in the distribution station; The perception layer (1) comprises: Environmental monitoring module (11): used to monitor the environment of the power distribution station; Equipment monitoring module (12): used to monitor the equipment in the power distribution station; The environmental monitoring module (11) comprises a temperature monitoring unit (1101), a humidity monitoring unit (1102), a gas monitoring unit (1103), a water level monitoring unit (1104), a flood monitoring unit (1105), a smoke monitoring unit (1106), an access control monitoring unit (1107), an open fire monitoring unit (1108), an electronic fence monitoring unit (1109), an infrared dual-detection monitoring unit (1110), a beam monitoring unit (1111), and a camera monitoring unit (1112); The equipment monitoring module (12) comprises a transformer monitoring unit (1201), a busbar and cable pile temperature monitoring unit (1202), and a switch cabinet monitoring unit (1203); The application layer (3) comprises: Data processing module (31): used to process, analyze and store data uploaded by the transport layer; Equipment monitoring module (32): used for real-time monitoring, alarming and analysis of equipment and environment in the power distribution station; Database (33): used to store monitoring data and system configuration information.
2. A distribution station equipment monitoring system according to claim 1, characterized in that: The various sensors of the sensing layer (1) perform self-diagnosis, self-calibration and self-compensation based on intelligent sensing technology; The intelligent sensing technology performs spectrum analysis on the output signal, working parameters and historical fault data of the sensor based on a deep learning network, and uses the spectrum features as the input of a convolutional neural network to achieve fault diagnosis of the sensor; The deep learning network is based on the CNN network, which incorporates a multi-head attention mechanism and a multi-scale feature extraction module, is combined with a multi-layer Bi-LSTM network, and performs parameter optimization based on an optimization algorithm; The CNN network extracts and fuses multi-scale features of input data based on convolution kernels of sizes 3×3, 5×5, and 7×7, performs feature dimensionality reduction based on a fully connected layer, and then inputs the features into the attention module; The attention module maps the reduced-dimensional multi-scale features to obtain the query vector matrix of the input features , key vector matrix Sum vector matrix : Output of the attention module as follows: in, , and All are learning parameters; is the query vector matrix in the adjacency space self-attention mechanism , key vector matrix Sum vector matrix The dimension of is the transpose of a vector; For multi-head output; A number for each head; Input the features extracted by the CNN network into the multi-layer Bi-LSTM network to obtain the feature vector sequence corresponding to each Bi-LSTM network , No. The hidden layer output of a Bi-LSTM is expressed as: in, Indicates The forward hidden layer state of a Bi-LSTM; For the The forward weight parameters of a Bi-LSTM; Indicates The hidden layer output state of a Bi-LSTM; For the The forward bias parameters of a Bi-LSTM; Indicates The backward hidden layer state of a Bi-LSTM; For the The backward weight parameters of the Bi-LSTM; For the The backward bias parameter of a Bi-LSTM; Indicates The hidden layer output state of a Bi-LSTM; is the forward weight parameter of Bi-LSTM; is the backward weight parameter of Bi-LSTM; is the bias parameter of Bi-LSTM; The hidden layer of the last Bi-LSTM is connected to the encoding layer, and the encoding layer encodes the output of the hidden layer of the last Bi-LSTM. The encoding process is as follows: in, is the encoding output; is a symbolic function; is the transpose of a vector; is the weight parameter of the encoding layer; is the bias parameter; Finally, the output coding anomaly detection is performed based on the SVM vector machine; The learning parameters in the deep learning network are all obtained based on the improved particle swarm algorithm; The improved particle swarm algorithm is specifically as follows: in, , Respectively The parameter The second change and The speed of the next change; , For the The parameter The second change and The value of the iteration; is the inertia weight; , is the acceleration factor; , for A random variable uniformly distributed between is the individual optimal value; is the global optimal value; is the maximum number of iterations; is a random variable of weight parameter, with a value range of ; The intelligent sensing technology also performs self-calibration of the sensor based on an improved particle swarm algorithm; The intelligent sensing technology is also based on a deep neural network, which takes environmental factors and sensor input as the network input, and the sensor output error as the network output. By training the deep neural network, the error characteristics of the sensor are learned to achieve sensor output compensation.
3. A distribution station equipment monitoring system according to claim 2, characterized in that: The sensor also uses dynamic power management technology and software-defined networking to dynamically adjust the sensor's supply voltage and operating frequency according to the monitoring task.
4. A distribution station equipment monitoring system according to claim 3, characterized in that: The sensor centrally manages and controls the network connection of the monitoring device through a software-defined network controller, and dynamically adjusts the network topology, bandwidth and data transmission path according to the monitoring task.
5. A distribution station equipment monitoring system according to claim 1, characterized in that: Various devices in the power distribution station room use piezoelectric energy collection technology and electromagnetic induction energy collection technology to collect vibration energy generated during the operation of the devices and alternating electromagnetic field energy in the power distribution station room, and convert them into electrical energy to power the low-power sensing layer (1) and transmission layer (2).
6. A distribution station equipment monitoring system according to claim 1, characterized in that: Each monitoring unit in the equipment monitoring module (12) determines a critical value based on the safe and stable operating environment factors required by different types of equipment, and also determines a critical value based on the safe and stable operating environment factors of different models of equipment of the same type; and thereby builds a distribution station equipment status monitoring model to perform real-time monitoring and early warning of the equipment status in the distribution station.