A monitoring method for the operating state of a power station

Abstract

The present invention discloses a monitoring method for the operation status of a power station, which relates to the technical field of power stations and includes: deploying device operation status sensors and indoor environment status sensors in the power station to obtain various sensing data in real time; deploying a local memory in the power station to store the data of each sensor and preset each alarm threshold; deploying an edge data processing system in the power station to analyze each sensing data, predict potential hazards, and send the sensing data that has exceeded the alarm threshold and the potential hazards to a remote server deployed in the central monitoring room; the remote server pushes the sensing data that has exceeded the alarm threshold and the potential hazards to the user terminal. The present invention effectively improves the safety and operation efficiency of the power station.

Description

Technical Field

[0001] The present invention relates to the technical field of power stations, and particularly to a method for monitoring the operation status of a power station. Background Art

[0002] With the development of the power industry, as an important place for power production, the safe and stable operation of power stations is crucial to the reliability of the power grid.

[0003] Currently, the widely implemented monitoring of power stations mostly relies on personnel to regularly check the sensing data of power stations. Existing research on information-based and automated monitoring technologies mostly focuses on the monitoring of equipment status, without considering the monitoring of the indoor environment of power stations. At the same time, it mostly focuses on the research of threshold-triggered alarm mechanisms, while ignoring the importance of taking precautions. Therefore, it is particularly important to design a method that can monitor the operation status of power stations in real time and take precautions. Summary of the Invention

[0004] Aiming at the above deficiencies in the prior art, a method for monitoring the operation status of a power station provided by the present invention solves the problems of incomplete monitoring and untimely early warning in the existing power station operation status monitoring method.

[0005] In order to achieve the above invention purpose, the technical solution adopted by the present invention is as follows:

[0006] A method for monitoring the operation status of a power station, comprising the following steps:

[0007] S1. Deploy equipment operation status sensors and indoor environment status sensors in the power station to obtain each sensing data in real time;

[0008] S2. Deploy a local memory in the power station to store each sensor data and preset each alarm threshold;

[0009] S3. Deploy an edge data processing system in the power station to analyze each sensing data, predict potential hazards, and alarm the sensing data that has exceeded the alarm threshold and the potential hazards to a remote server deployed in the central monitoring room;

[0010] S4. The remote server pushes the sensing data that has exceeded the alarm threshold and the potential hazards to the user terminal.

[0011] Further, the edge data processing system includes:

[0012] A trend analysis module, which is used to predict the trend value of each sensing data through a high-order smoothing algorithm;

[0013] A threshold judgment module, which is used to judge whether each sensing data and its trend value exceed each alarm threshold;

[0014] A fault prediction module, which is used to analyze each sensing data through a pre-trained classifier to predict equipment faults.

[0015] The potential hazards described in steps S3 and S4 refer to the trend values of the sensing data exceeding the alarm threshold and the equipment faults predicted by the fault prediction module.

[0016] Further, the equipment operating state sensors include: a voltage sensor, an electric power sensor, a frequency meter, an electric field intensity sensor, and a first temperature sensor.

[0017] The indoor environment state sensors include: a second temperature sensor and a humidity sensor.

[0018] Further, the sensing data includes: the effective value of the AC voltage, the electric power value, the AC frequency, the three-dimensional matrix of the electric field vector, the equipment temperature value, the ambient temperature value, and the ambient humidity value;

[0019] The alarm thresholds include: an undervoltage threshold, an overvoltage threshold, an upper limit of the electric power, a frequency jitter limit, an upper limit of the equipment temperature, a lower limit of the ambient temperature, an upper limit of the ambient temperature, and an upper limit of the ambient humidity.

[0020] Further, the trend analysis module predicts the equipment temperature trend value, the ambient temperature trend value, and the ambient humidity trend value according to the equipment temperature value, the ambient temperature value, and the ambient humidity value through a high-order smoothing algorithm;

[0021] The expression of the high-order smoothing algorithm is:

[0022] ,

[0023] where y(t + 1) is the predicted trend value at time t + 1, y(t - k) is the previously predicted trend value at time t - k, x(t - k) is the true value at time t - k, and α k is the weighting coefficient of the k-th order true value, β k is the weighting coefficient of the k-th order error, and N is the order.

[0024] Further, the fault prediction module analyzes the three-dimensional matrix of the electric field vector through a pre-trained classifier to predict partial discharge faults.

[0025] Further, the pre-trained classifier includes:

[0026] A modulo operation layer, which is used to perform a modulo operation on each element in the three-dimensional matrix of the electric field vector to obtain a three-dimensional matrix of the electric field scalar;

[0027] A difference operation layer, which is used to obtain the difference between the elements in the current three-dimensional matrix of the electric field scalar and the corresponding elements in the three-dimensional matrix of the electric field scalar at the previous moment to obtain a three-dimensional matrix of the electric field difference component;

[0028] A Reshape layer for rearranging the three-dimensional matrix of the electric field difference components into a sequence of electric field difference components;

[0029] At least one fully connected layer for processing the sequence of electric field difference components through a fully connected neural network operation to obtain an electric field feature vector;

[0030] An output layer for analyzing the electric field feature vector and predicting whether a partial discharge fault occurs.

[0031] Furthermore, the expression of the output layer of the pre-trained classifier is:

[0032] ,

[0033] where f out is the partial discharge fault prediction value obtained by the output layer of the classifier, with the data type being an unsigned integer of 1 bit. A value of 1 represents that a partial discharge fault will occur, and a value of 0 represents that no partial discharge fault will occur; tanh(∙) is the hyperbolic tangent function; w i is the i-th weight coefficient of the output layer; f in (i) is the i-th element of the electric field feature vector; b ias is the bias coefficient of the output layer; M is the length of the electric field feature vector.

[0034] Furthermore, the pre-trained classifier is pre-trained with multiple three-dimensional matrices of electric field vectors labeled with whether a partial discharge fault occurs as samples, and the loss function used is:

[0035] ,

[0036] where r is the label of whether a partial discharge fault occurs in the sample, with the data type being an unsigned integer of 1 bit. A value of 1 represents that a partial discharge fault will occur, and a value of 0 represents that no partial discharge fault will occur; f out is the partial discharge fault prediction value obtained by the output layer of the classifier, with the data type being an unsigned integer of 1 bit. A value of 1 represents that a partial discharge fault will occur, and a value of 0 represents that no partial discharge fault will occur; L is the loss function; log2(∙) is the logarithm function with base 2.

[0037] The beneficial effects of the present invention are:

[0038] (1)The present invention realizes real-time monitoring of the operating status of equipment and the environmental status in the power station; through automated data acquisition, storage, and processing, it reduces human operation errors and improves the accuracy of monitoring; it not only has a threshold-triggered alarm mechanism but also can predict alarm situations and prevent problems before they occur; and it has the functions of remote alarm pushing and remote storage, preventing data loss due to damage to the power station. It effectively improves the safety and operating efficiency of the power station.

[0039] (2)It sets multiple data monitoring including voltage, current, frequency, temperature, humidity, and electric field, covering the operating status of the power station three-dimensionally. In particular, it can timely detect and transmit alarm situations such as undervoltage, overvoltage, power limit exceedance, and frequency jitter exceedance, avoiding risks.

[0040] (3)For data such as equipment temperature, environmental temperature, and environmental humidity that have time-domain continuity and cumulativeness, the present invention sets a high-order smoothing algorithm based on smooth iteration to predict the development trend and give earlier warnings.

[0041] (4)Partial discharge is a discharge phenomenon in which only part of the insulation system discharges under the action of an electric field without forming a through-discharge channel. It will damage the molecular structure of equipment materials, cause overheating, corrosion, sludge accumulation, and accelerate the aging of insulation devices, leading to accidents. The present invention measures the electric field vectors at various locations in the three-dimensional space of the power station through electric field strength sensors, sets a classifier, extracts the amplitude of the vector through modulus operation, obtains the change information of the amplitude through differential operation, rearranges the three-dimensional matrix into a sequence, and finally obtains the prediction result of partial discharge through fully connected operation and binary logic output, so as to more comprehensively monitor the safety of the power station.

[0042] (5)The output layer of the classifier selects the hyperbolic tangent function as the activation function. Compared with the sigmoid function, it is not prone to the situation of gradient disappearance and is easier to train the parameters of the classifier. The present invention translates and scales the result of the hyperbolic tangent function to the interval of 0 and 1, and sets the bit width to an unsigned integer of 1 bit, restricting the value to the binary values of 0 and 1. Then, using the loss function designed by the present invention, through the training of samples with the same rule labels, it becomes a logical indication of whether there is a partial discharge fault and can accurately predict partial discharge faults. Description of the Drawings

[0043] Figure 1 It is a flowchart of a method for monitoring the operating status of a power station provided by an embodiment of the present invention;

[0044] Figure 2 It is a schematic diagram of equipment operating status sensors, indoor environmental status sensors, local memories, edge data processing systems, remote servers, and user terminals deployed in an embodiment of the present invention;

[0045] Figure 3 This is the structural diagram of the classifier pre-trained in the embodiments of the present invention. Specific Embodiments

[0046] The following describes the specific embodiments of the present invention to facilitate those skilled in the art of this technology to understand the present invention. However, it should be clear that the present invention is not limited to the scope of the specific embodiments. For those of ordinary skill in the art of this technology, as long as various changes are within the spirit and scope of the present invention defined and determined by the appended claims, these changes are obvious, and all inventions and creations using the concept of the present invention are within the scope of protection.

[0047] As Figure 1 shown, in an embodiment of the present invention, a method for monitoring the operating state of a power station includes the following steps:

[0048] S1. Deploy device operating state sensors and indoor environment state sensors in the power station to obtain various sensing data in real time.

[0049] The device operating state sensors, as Figure 2 shown, include: a voltage sensor, an electric power sensor, a frequency meter, an electric field intensity sensor, and a first temperature sensor. The indoor environment state sensors include: a second temperature sensor and a humidity sensor.

[0050] The various sensing data include: the effective value of the AC voltage, the electric power value, the AC frequency, the three-dimensional matrix of the electric field vectors formed by the electric field vectors at various locations in the three-dimensional space of the power station indoor, the device temperature value measured by the first temperature sensor, the environmental temperature value measured by the second temperature sensor, and the environmental humidity value.

[0051] S2. Deploy a local memory in the power station to store the data of each sensor and preset each alarm threshold.

[0052] The various alarm thresholds include: an undervoltage threshold, an overvoltage threshold, an upper limit of electric power, a frequency jitter limit, an upper limit of device temperature, a lower limit of environmental temperature, an upper limit of environmental temperature, and an upper limit of environmental humidity.

[0053] S3. Deploy an edge data processing system in the power station to analyze the various sensing data, predict potential hazards, and report the sensing data that has exceeded the alarm threshold and the potential hazards to the remote server deployed in the central monitoring room.

[0054] The edge data processing system, as Figure 2 shown, includes:

[0055] A trend analysis module for predicting the device temperature trend value, the environmental temperature trend value, and the environmental humidity trend value according to the device temperature value, the environmental temperature value, and the environmental humidity value through a high-order smoothing algorithm.

[0056] A threshold judgment module, which is used to judge whether each sensing data and its trend value exceed each alarm threshold;

[0057] A fault prediction module, which is used to analyze the three-dimensional matrix of electric field vectors through a pre-trained classifier to predict partial discharge faults.

[0058] The potential hazards in this embodiment refer to the trend values of sensing data that exceed the alarm threshold and the partial discharge faults predicted by the fault prediction module.

[0059] The expression of the high-order smoothing algorithm adopted by the trend analysis module is:

[0060] ,

[0061] where y(t + 1) is the predicted trend value at time t + 1, y(t - k) is the predicted trend value at time t - k in the past, x(t - k) is the true value at time t - k, and α k is the weighting coefficient of the k-th order true value, β k is the weighting coefficient of the k-th order error, and N is the order.

[0062] The present invention sets up a variety of data monitoring including voltage, current, frequency, temperature, humidity, and electric field, which comprehensively covers the operating state of the power station. In particular, it can timely detect and transmit alarms such as undervoltage, overvoltage, power limit exceeding, and frequency jitter exceeding, and avoid risks.

[0063] For data such as equipment temperature, ambient temperature, and ambient humidity, which have time-domain continuity and cumulativeness, the present invention sets up a high-order smoothing algorithm based on smooth iteration to predict the development trend and give earlier warnings.

[0064] In this embodiment, the initial value y(0) of the predicted trend value at time 0 is set to 0 to start smooth iteration and perform subsequent iterative updates.

[0065] It should be noted that the weighting coefficient of the true value of each order must be greater than the weighting coefficient of the error; the sum of the weighting coefficients of the true values of each order must be equal to 1.

[0066] Such as Figure 3 shown, the pre-trained classifier adopted by the fault prediction module includes:

[0067] A modulo operation layer, which is used to perform a modulo operation on each element in the three-dimensional matrix of electric field vectors to obtain a three-dimensional matrix of electric field scalars;

[0068] A difference operation layer, which is used to calculate the difference between each element in the current three-dimensional matrix of electric field scalars and the corresponding element in the three-dimensional matrix of electric field scalars at the previous moment to obtain a three-dimensional matrix of electric field difference components;

[0069] A Reshape layer for rearranging the three-dimensional matrix of the electric field difference components into a sequence of electric field difference components;

[0070] At least one fully connected layer for processing the sequence of electric field difference components through a fully connected neural network operation to obtain an electric field feature vector. In this embodiment, one fully connected layer is adopted;

[0071] An output layer for analyzing the electric field feature vector and predicting whether a partial discharge fault occurs.

[0072] The expression of the output layer is:

[0073] ,

[0074] where f out is the predicted value of the partial discharge fault obtained by the output layer of the classifier, and the data type is an unsigned integer of 1 bit. The value 1 represents that a partial discharge fault will occur, and the value 0 represents that no partial discharge fault will occur; tanh(∙) is the hyperbolic tangent function; w i is the i-th weight coefficient of the output layer; f in (i) is the i-th element of the electric field feature vector; b ias is the bias coefficient of the output layer; M is the length of the electric field feature vector.

[0075] The so-called pre-training means pre-training the parameters of the fully connected layer and the output layer. In this embodiment, pre-training is performed using multiple three-dimensional matrices of electric field vectors labeled with whether a partial discharge fault occurs as samples, and the loss function used is:

[0076] ,

[0077] where r is the label of whether a partial discharge fault occurs in the sample, and the data type is an unsigned integer of 1 bit. The value 1 represents that a partial discharge fault will occur, and the value 0 represents that no partial discharge fault will occur; f out is the predicted value of the partial discharge fault obtained by the output layer of the classifier, and the data type is an unsigned integer of 1 bit. The value 1 represents that a partial discharge fault will occur, and the value 0 represents that no partial discharge fault will occur; L is the loss function; log2(∙) is the logarithm function with base 2.

[0078] Partial discharge is a kind of discharge phenomenon in which, under the action of an electric field, only some regions in the insulation system have discharge without forming a through-discharge channel. It will damage the molecular structure of the equipment materials, cause overheating, corrosion, sludge accumulation, accelerate the aging of the insulation device, and trigger accidents. In the present invention, the electric field vectors at various points in the three-dimensional space of the power station are measured by an electric field intensity sensor, and a classifier is set. Through modulus operation, the amplitude of the vector is extracted, and the change information of the amplitude is obtained through differential operation. Then, the three-dimensional matrix is rearranged into a sequence, and finally, through fully connected operation and binary logic output, the prediction result of partial discharge is obtained, so as to more comprehensively monitor the safety of the power station.

[0079] The hyperbolic tangent function is selected as the activation function for the output layer. Compared with the sigmoid function, it is not easy to have the situation of gradient disappearance and is easier to train the parameters of the classifier. In the present invention, the result of the hyperbolic tangent function is translated and scaled to the interval of 0 and 1, and the bit width is set as an unsigned integer with 1 bit, so that the value is limited to the binary values of 0 and 1. Then, using the loss function designed by the present invention, through the training of samples with the same rule labels, it becomes a logical indication of whether there is a partial discharge fault and can accurately predict the partial discharge fault.

[0080] S4. The remote server pushes the sensing data and potential hazards that have exceeded the alarm threshold to the user terminal.

[0081] In summary, the present invention realizes the real-time monitoring of the operating state and environmental state of the equipment in the power station; through automated data collection, storage and processing, it reduces human operation errors and improves the accuracy of monitoring; it not only has a threshold-triggered alarm mechanism, but also can predict the alarm situation and take precautions; and it has the functions of remote alarm push and remote storage of the alarm situation to prevent data loss due to damage to the power station. It effectively improves the safety and operating efficiency of the power station.

[0082] In the present invention, specific embodiments are applied to elaborate on the principle and implementation manner of the present invention. The description of the above embodiments is only used to help understand the method and its core idea of the present invention; at the same time, for those of ordinary skill in the art, according to the idea of the present invention, there will be changes in the specific implementation manner and application scope. In summary, the content of this specification should not be construed as a limitation to the present invention.

[0083] Those of ordinary skill in the art will realize that the embodiments described here are for helping readers understand the principle of the present invention, and it should be understood that the protection scope of the present invention is not limited to such specific statements and embodiments. Those of ordinary skill in the art can make various other specific deformations and combinations that do not deviate from the essence of the present invention according to the technical revelations disclosed in the present invention, and these deformations and combinations are still within the protection scope of the present invention.

Claims

1. A monitoring method for the operating state of a power station, characterized in that, It includes the following steps: S1. Deploy device operation status sensors and indoor environment status sensors in the power station to obtain various sensing data in real time; S2. Deploy a local memory in the power station to store the data of each sensor and preset each alarm threshold; S3. Deploy an edge data processing system in the power station to analyze various sensing data, predict potential hazards, and send the sensing data that has exceeded the alarm threshold and the potential hazards to the remote server deployed in the central monitoring room. Among them, the edge data processing system includes: a trend analysis module for predicting the trend values of various sensing data through a high-order smoothing algorithm, a threshold judgment module for judging whether various sensing data and their trend values exceed each alarm threshold, and a fault prediction module for analyzing various sensing data through a pre-trained classifier to predict equipment failures; S4. The remote server pushes the sensing data that has exceeded the alarm threshold and the potential hazards to the user terminal; The pre-trained classifier includes: A modulo operation layer for performing a modulo operation on each element in the three-dimensional matrix of the electric field vector to obtain a three-dimensional matrix of the electric field scalar; A difference operation layer for obtaining the difference between the elements in the current three-dimensional matrix of the electric field scalar and the corresponding elements in the three-dimensional matrix of the electric field scalar at the previous moment to obtain a three-dimensional matrix of the electric field difference component; A Reshape layer for rearranging the three-dimensional matrix of the electric field difference component into a sequence of the electric field difference component; At least one fully connected layer for processing the sequence of the electric field difference component through a fully connected neural network operation to obtain an electric field feature vector; An output layer for analyzing the electric field feature vector to predict whether a partial discharge fault occurs.

2. The monitoring method for the operating state of a power station according to claim 1, characterized in that, The device operation status sensors include: a voltage sensor, an electric power sensor, a frequency meter, an electric field intensity sensor, and a first temperature sensor; The indoor environment status sensors include: a second temperature sensor and a humidity sensor.

3. The monitoring method of the power station operation state according to claim 2, characterized in that, The various sensing data include: the effective value of the AC voltage, the electric power value, the AC frequency, the three-dimensional matrix of the electric field vector, the device temperature value, the environmental temperature value, and the environmental humidity value; The various alarm thresholds include: an undervoltage threshold, an overvoltage threshold, an upper limit of the electric power, a frequency jitter limit, an upper limit of the device temperature, a lower limit of the environmental temperature, an upper limit of the environmental temperature, and an upper limit of the environmental humidity.

4. The monitoring method for the operating state of a power station according to claim 3, characterized in that, The trend analysis module predicts the device temperature trend value, the environmental temperature trend value, and the environmental humidity trend value through a high-order smoothing algorithm according to the device temperature value, the environmental temperature value, and the environmental humidity value; The expression of the high-order smoothing algorithm is: ， Among them, y(t + 1) is the predicted trend value at time t + 1, y(t - k) is the previously predicted trend value at time t - k, x(t - k) is the true value at time t - k, and α k is the weighting coefficient of the k-th order true value, β k is the weighting coefficient of the k-th order error, and N is the order number.

5. The monitoring method of the power station operation state according to claim 3, wherein The fault prediction module analyzes the three-dimensional matrix of the electric field vector through a pre-trained classifier to predict a partial discharge fault.

6. The monitoring method for the operating state of a power station according to claim 1, characterized in that The expression of the output layer of the pre-trained classifier is: ， Among them, f out is the partial discharge fault prediction value obtained by the output layer of the classifier. The data type is an unsigned integer of 1 bit. The value 1 represents that a partial discharge fault will occur, and the value 0 represents that no partial discharge fault will occur; tanh(∙) is the hyperbolic tangent function; w i is the i-th weight coefficient of the output layer; f in (i) is the i-th element of the electric field feature vector; b ias is the bias coefficient of the output layer; M is the length of the electric field feature vector.

7. The monitoring method for the operating state of a power station according to claim 6, characterized in that, The pre-trained classifier is pre-trained with multiple three-dimensional matrices of the electric field vector labeled with whether a partial discharge fault occurs as samples, and the loss function used is: ， Among them, r is the label indicating whether the sample has a partial discharge fault. The data type is an unsigned integer of 1 bit. The value 1 represents that a partial discharge fault will occur, and the value 0 represents that no partial discharge fault will occur; f out is the predicted value of the partial discharge fault obtained by the output layer of the classifier. The data type is an unsigned integer of 1 bit. The value 1 represents that a partial discharge fault will occur, and the value 0 represents that no partial discharge fault will occur; L is the loss function; log2(∙) is the logarithmic function with base 2.

Images