CN112883634B - DC measurement system state prediction method and system based on multi-dimensional analysis - Google Patents
DC measurement system state prediction method and system based on multi-dimensional analysis Download PDFInfo
- Publication number
- CN112883634B CN112883634B CN202110063456.2A CN202110063456A CN112883634B CN 112883634 B CN112883634 B CN 112883634B CN 202110063456 A CN202110063456 A CN 202110063456A CN 112883634 B CN112883634 B CN 112883634B
- Authority
- CN
- China
- Prior art keywords
- direct current
- voltage
- value
- current
- data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
Abstract
The invention discloses a method and a system for predicting the state of a direct current measurement system based on multi-dimensional analysis, which relate to the technical field of direct current measurement system monitoring, wherein the method comprises the steps of obtaining the monitoring value and direct current key operation parameters of the direct current measurement system; judging whether the collected data of the direct current measurement system is abnormal or not according to a set first criterion based on the monitoring value and the direct current key operation parameter; and under the condition that the acquired data are not abnormal, judging whether the direct current power transmission system is abnormal according to a set second judgment datum. From the aspect of engineering application, on the basis of the existing operation and maintenance threshold diagnosis, the method adopts a transverse comparison and time sequence trend analysis method to diagnose the driving current and the data level of direct current measurement and identify the abnormality. The algorithm is simple and effective to apply.
Description
Technical Field
The invention relates to the technical field of monitoring of direct current measurement systems, in particular to a method and a system for predicting states of a direct current measurement system based on multi-dimensional analysis.
Background
In an extra-high voltage direct current transmission project, a direct current measurement system provides input signals for control and protection, and the measurement accuracy directly influences the safe and stable operation of a high-voltage direct current transmission system. At present, a great amount of photoelectric transformers are applied to direct current transmission systems of southern power grids, and when large deviation occurs in measurement of driving current or voltage, actions of a protection system can be caused, so that normal operation of a converter station is influenced. The direct current measurement system fault is one of the main problems faced by the operation of the direct current transmission system of the southern power grid in recent years. Patent 202010818172.5 proposes a measuring system detecting device, which detects the performance of the laser power supply loop of the high voltage direct current measuring system and detects the unit with fault, thereby realizing fault location. In the daily operation and maintenance work of the converter station, the running state of the direct current measurement system is mainly monitored by means of a data monitoring disc of operation and maintenance personnel, an abnormal alarm and the like. However, under the influence of factors such as the ambient temperature of the cabinet, the external environment, the defects of the optical fibers and the like, the direct current measurement current may have abnormal fluctuation, the abnormal alarm based on the threshold value alone has certain limitation, and when the alarm value is not exceeded, the abnormality (such as a slow increase trend) of the direct current measurement data may be covered in the fluctuation of the normal monitoring signal, and is difficult to identify under the condition of manual monitoring. There is no corresponding research for the prediction and analysis of the direct current measurement data trend. Therefore, it is necessary to perform prediction and analysis on the trend of the dc measurement data to provide an auxiliary basis for the state evaluation of the dc measurement system.
Disclosure of Invention
Aiming at the defects in the prior art, the invention provides a method and a system for predicting the state of a direct current measurement system based on multi-dimensional analysis, so as to realize the state evaluation of the direct current measurement system.
In order to achieve the purpose, the technical scheme of the invention is as follows:
a DC measurement system state prediction method based on multi-dimensional analysis is used for a DC power transmission system and comprises the following steps:
acquiring a monitoring value and a direct current key operation parameter of a direct current measuring system;
judging whether the collected data of the direct current measurement system is abnormal or not according to a set first criterion based on the monitoring value and the direct current key operation parameter;
and under the condition that the acquired data are not abnormal, judging whether the direct current power transmission system is abnormal according to a set second judgment datum.
The method for predicting the state of the direct current measurement system based on the multidimensional analysis further comprises the steps of monitoring the driving current, the data level and the ambient temperature, wherein the direct current key operation parameters comprise direct current voltage, direct current power, a trigger angle, a tap gear and direct current system impedance.
The method for predicting the state of the direct current measurement system based on the multidimensional analysis further includes:
calculating a theoretical value of the driving current, calculating deviation between a monitoring value of the driving current and the theoretical value, comparing the deviation with a set first threshold value, and judging whether the sampling of the driving current of a measuring point is abnormal or not;
and calculating a theoretical value of the voltage, calculating the deviation of the monitoring value of the voltage and the theoretical value, comparing the deviation with a set second threshold value, and judging whether the voltage sampling of the measuring point is abnormal or not.
The method for predicting the state of the direct current measurement system based on the multidimensional analysis further comprises the following steps of:
taking the monitoring value of the driving current as a dependent variable, and taking the direct current power and the current of the corresponding pole as independent variables;
constructing a relational expression of independent variables and dependent variables through polynomial fitting;
based on the historical data of the direct current power, the current and the driving current of the direct current power transmission system, the relation y (x) is obtained by fitting with a least square method1,x2)；
And obtaining a theoretical value of the driving current based on the historical data of the direct current power, the current and the driving current of the direct current power transmission system according to the relational expression.
The method for predicting the state of the direct current measurement system based on the multidimensional analysis further comprises the following steps of:
rectifying side:
inverting side:
U1and U2Respectively calculating the effective values of the air load voltage at the valve sides of the converter transformers of the rectifier station and the inverter station according to the bus voltage and the gear of the converter transformer tap, wherein alpha is the trigger angle of the rectifier station, gamma is the extinction angle of the inverter station, and X is the trigger angle of the rectifier stationr1And Xr2Commutation reactance, I, for a rectifier station and an inverter station, respectivelydIs direct current;
calculating the deviation of the monitored value of the voltage from the theoretical value
Ud_err＝Ud_cal-Ud_means
Wherein, Ud_errFor calculating the deviation of the voltage from the measured voltage, Ud_calFor calculating the voltage, Ud_meansFor measuring voltage, the measuring voltage of the high-end valve bank is UdH-UdM, and the measuring voltage of the low-end valve bank is UdM-UdN; UdH shows the high voltage dc bus voltage, UdM shows the medium voltage bus voltage between the high and low end valve trains, and UdN shows the neutral dc bus voltage.
The method for predicting the state of the direct current measurement system based on the multidimensional analysis further includes:
criterion one is as follows: at the same time, transversely comparing the drive current and data level of the same high and low ends, and when the deviation difference is larger than the threshold value delta1If so, judging the condition as abnormal;
criterion two: longitudinally analyzing the drive current and data level, calculating a sequence of samples X (X)1,x2,...xt) Is driven toThe value of the potential change, i.e., ∑ (x)t-xt-1) When the trend value is larger than the threshold value, δ2Early warning, i.e. sigma (x)t-xt-1)＞δ2. A trend-free fluctuation analysis method is considered, and the fact that the difference value between a monitored value and a sample mean value does not exceed a fixed value is used as sequence data fluctuation diagnosis. When the difference between the monitored value and the mean value exceeds the range of a multiplied by sigma, the result is judged to be abnormal, namely
Criterion three: and when the driving current trend predicted value is larger than the threshold value, judging that the driving current trend predicted value is abnormal. The prediction algorithm is implemented by LSTM.
The method for predicting the state of the direct current measurement system based on the multidimensional analysis further comprises the following steps of performing a driving current trend prediction value through a driving current prediction model, wherein the driving current prediction model comprises an LSTM model, and the method comprises the following steps:
predicting the driving current according to the driving current, the data level, the ambient temperature and the direct current power and through the relation among the ambient temperature, the direct current power, the data level and the driving current which are fitted by a driving current prediction model;
wherein the drive current prediction model comprises:
selecting the driving current, the direct current power, the environment temperature and the data level at the current moment as input parameters, outputting the driving current data at the next moment, and constructing a data sample set;
detecting and cleaning data in the data set by utilizing a gray correlation degree and a K-means clustering method, and dividing a data sample set into a training set and a testing set according to a set proportion;
normalizing the selected training set data to be used as LSTM model input, setting the number of neurons input by the LSTM model to be 4, wherein the number of neurons comprises direct current power, data level, environment temperature and current driving current, namely { x1,x2,x3,x4The output variable is the driving current at the next moment; the neural network selects 2 layers of LSTM, and is connected with a full connection layer at lastAnd taking an activation function as a final output.
If the average absolute error of the training is larger than the threshold value in the training process, updating the weights and the offsets W, b and a, and continuously interacting with the input to calculate; and if the error is smaller than the threshold value, outputting the result as another initialization parameter, wherein the calculation formula of the average absolute error is as follows:
and (4) carrying out error back propagation algorithm (BP) neural network training, namely carrying out back propagation calculation on errors at all times after t, updating the weight matrix until an optimal global parameter matrix is obtained, and finishing model training.
The method for predicting the state of the direct current measurement system based on the multidimensional analysis, further,
judging whether the acquired data of the direct current measurement system is abnormal according to a set first criterion, and if so, giving an alarm;
and judging whether the direct current transmission system is abnormal according to the set second judgment data, and if so, giving a warning.
A direct current measurement system is used for obtaining state quantity of a direct current transmission system and transmitting the state quantity to a control protection system, and the control protection system works according to the method.
As an alternative implementation, in some embodiments,
the direct current transmission system comprises a remote module, the remote module is provided with an AD conversion module, an energy management module and an LED module,
the direct current measurement system comprises a merging unit, the merging unit is provided with a laser driver and an I/O unit, wherein the AD conversion module collects primary voltage through a voltage divider, collects primary current through a current divider and transmits the primary current to the I/O unit through a data optical fiber, and the merging unit transmits the primarily processed data to the control protection system; the laser driver drives the energy management module to provide energy for the AD conversion module through an energy supply optical fiber.
As an alternative embodiment, in some embodiments, the remote module is connected to a secondary voltage divider plate that is connected to the high and low voltage arms of the dc link.
Compared with the prior art, the invention has the beneficial effects that: 1. from the aspect of engineering application, on the basis of the existing operation and maintenance threshold diagnosis, a transverse comparison and time sequence trend analysis method is adopted for diagnosing the driving current and the data level of direct current measurement and identifying abnormality. The algorithm is simple and effective to apply.
2. Aiming at trend prediction, actual converter station direct current measurement data are used as samples, a multi-input single-output LSTM driving current prediction model is constructed by combining direct current power, the same sample set is adopted, and compared with a traditional time sequence prediction model, the prediction accuracy is high.
Drawings
In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings needed to be used in the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present application, and it is obvious for those skilled in the art that other drawings can be obtained according to these drawings without creative efforts.
Fig. 1 is a structural diagram of a converter station direct current optical measurement system according to an embodiment of the present invention;
FIG. 2 is a flow chart of prediction using LSTM according to an embodiment of the present invention;
FIG. 3 is a measurement circuit of the DC voltage measurement device according to the embodiment of the present invention;
FIG. 4 shows the results of different model predictions according to an embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be described clearly and completely with reference to the accompanying drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present application, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
Example (b):
it should be noted that the terms "first," "second," and the like in the description and claims of the present invention and in the drawings described above are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used is interchangeable under appropriate circumstances such that the embodiments of the invention described herein are capable of operation in sequences other than those illustrated or described herein. Furthermore, the terms "comprises," "comprising," and any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed, but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.
Furthermore, unless expressly stated or limited otherwise, the terms "mounted," "connected," and "connected" are to be construed broadly, as they may be fixedly connected, detachably connected, or integrally connected, for example; can be mechanically or electrically connected; they may be connected directly or indirectly through intervening media, or they may be interconnected between two elements. The specific meanings of the above terms in the present invention can be understood in specific cases to those skilled in the art.
Referring to fig. 1 to 4, fig. 1 is a structural diagram of a converter station dc optical measurement system according to an embodiment of the present invention; FIG. 2 is a flow chart of prediction using LSTM according to an embodiment of the present invention; FIG. 3 is a measurement circuit of the DC voltage measurement device according to the embodiment of the present invention; FIG. 4 shows the results of different model predictions according to an embodiment of the present invention.
A DC measurement system state prediction method based on multi-dimensional analysis is used for a DC power transmission system and comprises the following steps:
acquiring a monitoring value and a direct current key operation parameter of a direct current measuring system;
judging whether the collected data of the direct current measurement system is abnormal or not according to a set first criterion based on the monitoring value and the direct current key operation parameter;
and under the condition that the acquired data are not abnormal, judging whether the direct current power transmission system is abnormal according to a set second judgment datum.
As an alternative implementation, in some embodiments, the monitored values include drive current, data level, ambient temperature, and the dc key operating parameters include dc voltage, dc current, dc power, firing angle, tap position, and dc system impedance.
As an optional implementation manner, in some embodiments, the first criterion determination includes:
calculating a theoretical value of the driving current, calculating deviation between a monitoring value of the driving current and the theoretical value, comparing the deviation with a set first threshold value, and judging whether the sampling of the driving current of a measuring point is abnormal or not;
and calculating a theoretical value of the voltage, calculating the deviation of the monitoring value of the voltage and the theoretical value, comparing the deviation with a set second threshold value, and judging whether the voltage sampling of the measuring point is abnormal or not.
As an alternative implementation, in some embodiments, the theoretical value of the driving current is obtained by:
taking the monitoring value of the driving current as a dependent variable, and taking the direct current power and the current of the corresponding pole as independent variables;
constructing a relational expression of independent variables and dependent variables through polynomial fitting;
based on the historical data of the direct current power, the current and the driving current of the direct current power transmission system, the relation y (x) is obtained by fitting with a least square method1,x2)；
And obtaining a theoretical value of the driving current based on the historical data of the direct current power, the current and the driving current of the direct current power transmission system according to the relational expression.
As an alternative implementation, in some embodiments, the theoretical value of the voltage is determined by:
rectifying side:
inverting side:
U1and U2Respectively calculating the effective values of the air load voltage at the valve sides of the converter transformers of the rectifier station and the inverter station according to the bus voltage and the gear of the converter transformer tap, wherein alpha is the trigger angle of the rectifier station, gamma is the extinction angle of the inverter station, and X is the trigger angle of the rectifier stationr1And Xr2Commutation reactance, I, for a rectifier station and an inverter station, respectivelydIs direct current;
calculating the deviation of the monitored value of the voltage from the theoretical value
Ud_err＝Ud_cal-Ud_means
Wherein, Ud_errFor calculating the deviation of the voltage from the measured voltage, Ud_calFor calculating the voltage, Ud_meansFor measuring voltage, the measuring voltage of the high-end valve bank is UdH-UdM, and the measuring voltage of the low-end valve bank is UdM-UdN; UdH shows the high voltage dc bus voltage, UdM shows the medium voltage bus voltage between the high and low end valve trains, and UdN shows the neutral dc bus voltage.
As an optional implementation manner, in some embodiments, the second criterion includes:
criterion one is as follows: at the same time, transversely comparing the drive current and data level of the same high and low ends, and when the deviation difference is larger than the threshold value delta1If so, judging the condition as abnormal;
criterion two: longitudinally analyzing the drive current and data level, calculating a sequence of samples X (X)1,x2,...xt) The overall trend change value of, i.e., ∑ (x)t-xt-1) When the trend value is larger than the threshold value, δ2Early warning, i.e. sigma (x)t-xt-1)＞δ2. Considering a trend-removing fluctuation analysis method, using the difference value between the monitoring value and the sample mean value not exceeding the fixed value as the sequence data fluctuation diagnosis. When the difference between the monitored value and the mean value exceeds the range of a multiplied by sigma, the result is judged to be abnormal, namely
Criterion three: and when the driving current trend predicted value is larger than the threshold value, judging that the driving current trend predicted value is abnormal. The prediction algorithm is implemented by LSTM.
As an alternative implementation, in some embodiments, the driving current trend prediction value is performed by a driving current prediction model, the driving current prediction model including an LSTM model, which includes:
predicting the driving current according to the driving current, the data level, the ambient temperature and the direct current power and through the relation among the ambient temperature, the direct current power, the data level and the driving current which are fitted by a driving current prediction model;
wherein the drive current prediction model comprises:
selecting the driving current, the direct current power, the environment temperature and the data level at the current moment as input parameters, outputting the driving current data at the next moment, and constructing a data sample set;
detecting and cleaning data in the data set by utilizing a gray correlation degree and a K-means clustering method, and dividing a data sample set into a training set and a testing set according to a set proportion;
normalizing the selected training set data to be used as LSTM model input, setting the number of neurons input by the LSTM model to be 4, wherein the number of neurons comprises direct current power, data level, environment temperature and current driving current, namely { x1，x2，x3，x4The output variable is the driving current at the next moment; the neural network selects 2 layers of LSTM, and finally connects a full connection layer to be used as final output through an activation function.
If the average absolute error of the training is larger than the threshold value in the training process, updating the weights and the offsets W, b and a, and continuously interacting with the input to calculate; and if the error is smaller than the threshold value, outputting the result as another initialization parameter, wherein the calculation formula of the average absolute error is as follows:
and (4) carrying out error back propagation algorithm (BP) neural network training, namely carrying out back propagation calculation on errors at all times after t, updating the weight matrix until an optimal global parameter matrix is obtained, and finishing model training.
As an alternative implementation, in some embodiments,
judging whether the acquired data of the direct current measurement system is abnormal according to a set first criterion, and if so, giving an alarm;
and judging whether the direct current transmission system is abnormal according to the set second judgment data, and if so, giving a warning.
A direct current measurement system is used for obtaining state quantity of a direct current transmission system and transmitting the state quantity to a control protection system, and the control protection system works according to the method.
As an alternative implementation, in some embodiments,
the direct current transmission system comprises a remote module, the remote module is provided with an AD conversion module, an energy management module and an LED module,
the direct current measurement system comprises a merging unit, the merging unit is provided with a laser driver and an I/O unit, wherein the AD conversion module collects primary voltage through a voltage divider, collects primary current through a current divider and transmits the primary current to the I/O unit through a data optical fiber, and the merging unit transmits the primarily processed data to the control protection system; the laser driver drives the energy management module to provide energy for the AD conversion module through an energy supply optical fiber.
As an alternative embodiment, in some embodiments, the remote module is connected to a secondary voltage divider plate that is connected to the high and low voltage arms of the dc link.
The embodiment collects the actual data of the direct current measurement of a certain converter station as a sample set. 180 measuring points are covered in the bipolar direct current transmission system. The sample set covers the operation and maintenance copying data from 12 months in 2018 to 11 months in 2019, and the sample interval is 1 hour. The state quantities include drive current, data level, ambient temperature, and dc power. And analyzing the transverse comparison deviation and the trend quantization value of the driving current and the data level, wherein the monitoring value of the driving current and the data level at the same measuring point does not exceed the thresholds of 65mA and 70mV, the trend quantization monitoring value is less than the thresholds of 100mA and 150mV, and the system state is normal. And analyzing the difference value between the monitoring value and the average value, when the difference value between the monitoring value and the average value exceeds 3 times of the standard difference value, performing early warning by the system, calculating the conductivity of the four groups of valve cooling systems, and if the results do not exceed 3 times of the standard difference value, driving the current index to be normal.
On the basis of sample data collection of a certain converter station, an LSTM algorithm model and a simulation environment are built, the LSTM algorithm model comprises a Python3.7 environment and function libraries such as Keras, Scikit-lern and Tensorflow, offline training is conducted on data, and the trained model is packaged and embedded into an early warning system. In the example, the learning rate is set to 0.01, the input dimension is 4, the output node is 1, all the matrixes are initialized, the error threshold value is set to 2 x 10 < -2 >, the adam optimization algorithm is selected for optimization, the iteration is performed for 80 times, when the iteration is performed for about 20 times, the training sample is smaller than the error threshold value, the output driving current value is obtained, and the training is finished.
The test set of 11 months in 2019 is input into the trained model, and the obtained test result and the result pair of the true values in the test set are shown in fig. 4. It can be seen from the figure that the coincidence ratio of the real value and the predicted value curve is high, the predicted value is slightly lower than the real value at the peak, the predicted value is slightly higher than the real value at the trough, and the overall prediction error is 2.6089%.
In order to verify the superiority of the LSTM prediction algorithm in this embodiment, the same sample set is used in this embodiment to perform comparative analysis on the regression prediction model based on time series and the prediction model based on LSTM. As can be seen from fig. 4, both prediction models can simulate the variation trend of the driving current, but when the actual value of the driving current fluctuates and continuously decreases, the time series model has a larger deviation between the prediction result and the actual value. The MAE evaluation index calculation results of the two models on the test set are respectively 36mA and 3mA, which shows that the prediction accuracy based on the LSTM is higher. The prediction model of the LSTM can be better fitted with the variation trend of the actual driving current result value.
In the description herein, references to the description of the term "one embodiment," "some embodiments," "an example," "a specific example," or "some examples," etc., mean that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the invention. In this specification, the schematic representations of the terms used above are not necessarily intended to refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. Furthermore, various embodiments or examples and features of different embodiments or examples described in this specification can be combined and combined by one skilled in the art without contradiction.
The above embodiments are only for illustrating the technical concept and features of the present invention, and the purpose thereof is to enable those skilled in the art to understand the contents of the present invention and implement the present invention accordingly, and not to limit the protection scope of the present invention accordingly. All equivalent changes or modifications made in accordance with the spirit of the present disclosure are intended to be covered by the scope of the present disclosure.
Claims (6)
1. A DC measurement system state prediction method based on multi-dimensional analysis is used for a DC power transmission system and is characterized by comprising the following steps:
acquiring a monitoring value and a direct current key operation parameter of a direct current measuring system;
judging whether the collected data of the direct current measurement system is abnormal or not according to a set first criterion based on the monitoring value and the direct current key operation parameter;
judging whether the direct current transmission system is abnormal according to a set second judgment datum under the condition that the acquired data are not abnormal, wherein,
the first criterion judgment comprises:
calculating a theoretical value of the driving current, calculating deviation between a monitoring value of the driving current and the theoretical value, comparing the deviation with a set first threshold value, and judging whether the sampling of the driving current of a measuring point is abnormal or not;
calculating a theoretical value of the voltage, calculating the deviation of the monitoring value of the voltage and the theoretical value, comparing the deviation with a set second threshold value, and judging whether the voltage sampling of the measuring point is abnormal or not;
the theoretical value of the driving current is obtained by the following method:
taking the monitoring value of the driving current as a dependent variable, and taking the direct current power and the current of the corresponding pole as independent variables;
constructing a relational expression of independent variables and dependent variables through polynomial fitting;
based on historical data of direct current power, current and driving current of a direct current power transmission system, a least square method is adopted to fit to obtain a relation y ═ f (x)1,x2)；
Obtaining a theoretical value of the driving current based on historical data of the direct current power, the current and the driving current of the direct current power transmission system according to the relational expression;
the theoretical value of the voltage is obtained by the following method:
rectifying side:
inverting side:
U1and U2Respectively calculating the effective values of the air load voltage at the valve sides of the converter transformers of the rectifier station and the inverter station according to the bus voltage and the gear of the converter transformer tap, wherein alpha is the trigger angle of the rectifier station, and X is the trigger angle of the rectifier stationr1And Xr2Commutation reactance, I, for a rectifier station and an inverter station, respectivelydIs direct current;
calculating the deviation of the monitored value of the voltage from the theoretical value
Ud_err＝Ud_cal-Ud_means
Wherein, Ud_errFor calculating the deviation of the voltage from the measured voltage, Ud_calFor calculating the voltage, Ud_meansFor measuring voltage, the measuring voltage of the high-end valve bank is UdH-UdM, and the measuring voltage of the low-end valve bank is UdM-UdN; UdH, UdM, and UdN, wherein the high-voltage direct-current bus voltage is represented, the medium-voltage bus voltage between the high-low end valve group is represented, and the neutral direct-current bus voltage is represented;
the second criterion includes:
criterion one is as follows: at the same time, transversely comparing the drive current and data level of the same high and low ends, when the deviation is larger than the threshold value delta1If so, judging the condition as abnormal;
criterion two: longitudinally analyzing the drive current and data level, calculating a sequence of samples X (X)1,x2,...xt) The overall trend change value of, i.e., ∑ (x)t-xt-1) When the trend value is larger than the threshold value, δ2Early warning, i.e. sigma (x)t-xt-1)＞δ2(ii) a Considering a trend-removing fluctuation analysis method, and using a difference value between a monitored value and a sample mean value which does not exceed a fixed value as sequence data fluctuation diagnosis; when the difference between the monitored value and the mean value exceeds the range of a multiplied by sigma, the result is judged to be abnormal, namely
Criterion three: when the driving current trend predicted value is larger than the threshold value, judging that the driving current trend predicted value is abnormal; the prediction algorithm is implemented by LSTM;
the driving current trend prediction value is carried out through a driving current prediction model, and the driving current prediction model comprises
An LSTM model comprising:
predicting the driving current according to the driving current, the data level, the ambient temperature and the direct current power and through the relation among the ambient temperature, the direct current power, the data level and the driving current which are fitted by a driving current prediction model;
wherein the drive current prediction model comprises:
selecting the driving current, the direct current power, the environment temperature and the data level at the current moment as input parameters, outputting the driving current data at the next moment, and constructing a data sample set;
detecting and cleaning data in the data set by utilizing a gray correlation degree and a K-means clustering method, and dividing a data sample set into a training set and a testing set according to a set proportion;
normalizing the selected training set data to be used as LSTM model input, setting the number of neurons input by the LSTM model to be 4, wherein the number of neurons comprises direct current power, data level, environment temperature and current driving current, namely { x1,x2,x3,x4The output variable is the driving current at the next moment; the neural network selects 2 layers of LSTM, and finally a full connection layer is connected to serve as final output through an activation function;
if the average absolute error of the training is larger than the threshold value in the training process, updating the weights and the offsets W, b and a, and continuously interacting with the input to calculate; and if the error is smaller than the threshold value, outputting the result as another initialization parameter, wherein the calculation formula of the average absolute error is as follows:
and (4) carrying out error back propagation algorithm (BP) neural network training, namely carrying out back propagation calculation on errors at all times after t, updating the weight matrix until an optimal global parameter matrix is obtained, and finishing model training.
2. The method of claim 1, wherein the monitored values include drive current, data level, ambient temperature, and the DC key operating parameters include DC voltage, DC current, DC power, firing angle, tap position, and DC system impedance.
3. The method for predicting the state of a DC measurement system based on multi-dimensional analysis according to claim 1,
judging whether the acquired data of the direct current measurement system is abnormal according to a set first criterion, and if so, giving an alarm;
and judging whether the direct current transmission system is abnormal according to the set second judgment data, and if so, giving a warning.
4. A dc measurement system for obtaining and transmitting state quantities of a dc transmission system to a control and protection system operating according to the method of any one of claims 1-3.
5. The direct current measurement system of claim 4,
the direct current transmission system comprises a remote module, the remote module is provided with an AD conversion module, an energy management module and an LED module,
the direct current measurement system comprises a merging unit, the merging unit is provided with a laser driver and an I/O unit, wherein the AD conversion module collects primary voltage through a voltage divider, collects primary current through a current divider and transmits the primary current to the I/O unit through a data optical fiber, and the merging unit transmits the primarily processed data to the control protection system; the laser driver drives the energy management module to provide energy for the AD conversion module through an energy supply optical fiber.
6. The direct current measurement system of claim 5, wherein the remote module is connected to a secondary voltage divider plate that is connected to a high voltage arm and a low voltage arm of the direct current line.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110063456.2A CN112883634B (en) | 2021-01-18 | 2021-01-18 | DC measurement system state prediction method and system based on multi-dimensional analysis |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110063456.2A CN112883634B (en) | 2021-01-18 | 2021-01-18 | DC measurement system state prediction method and system based on multi-dimensional analysis |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112883634A CN112883634A (en) | 2021-06-01 |
| CN112883634B true CN112883634B (en) | 2022-04-08 |
Family
ID=76049079
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110063456.2A Active CN112883634B (en) | 2021-01-18 | 2021-01-18 | DC measurement system state prediction method and system based on multi-dimensional analysis |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112883634B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114295883B (en) * | 2022-01-06 | 2023-08-22 | 南京大学 | Multi-dimensional calibration method for improving measurement accuracy of optical fiber current sensor |
| CN114489170B (en) * | 2022-04-15 | 2022-06-21 | 天津航天和兴科技有限公司 | System and method for predicting and adjusting temperature in sealed cabin |
| CN116502156B (en) * | 2023-06-30 | 2023-09-08 | 中国电力科学研究院有限公司 | Intelligent identification method and system for light CT abnormal state of converter station |
Family Cites Families (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2001141767A (en) * | 1999-11-09 | 2001-05-25 | Yokogawa Electric Corp | Jitter measuring circuit and ic tester using it |
| US7529069B1 (en) * | 2002-08-08 | 2009-05-05 | Weems Ii Warren A | Apparatus and method for ground fault detection and location in electrical systems |
| CN102185278A (en) * | 2011-05-03 | 2011-09-14 | 中国电力工程顾问集团中南电力设计院 | Extra-high-voltage direct-current power transmission converter station protecting method |
| CN103116107B (en) * | 2012-12-28 | 2015-12-02 | 广东志成冠军集团有限公司 | A kind of high-voltage DC power supply insulating monitor |
| KR101306283B1 (en) * | 2013-05-08 | 2013-09-09 | (주) 에이블리 | A power supply device for testing a semiconductor element |
| US9791495B2 (en) * | 2015-12-31 | 2017-10-17 | Abb Schweiz Ag | High impedance fault location in DC distribution systems |
| CN106771500B (en) * | 2017-01-22 | 2019-06-21 | 中国南方电网有限责任公司超高压输电公司检修试验中心 | A kind of HVDC transmission system difference channel current amount otherness detection method |
| CN107121639B (en) * | 2017-05-02 | 2019-09-06 | 广东电网有限责任公司揭阳供电局 | A kind of multi-Dimensional parameters direct current system accumulator cell management method and device |
| CN107727963B (en) * | 2017-10-10 | 2019-08-09 | 中国南方电网有限责任公司超高压输电公司昆明局 | The electrical quantities measurement fault distinguishing method verified based on historical data sample and algorithm |
| CN107680835B (en) * | 2017-10-13 | 2019-07-26 | 中国电力科学研究院 | A kind of breaker arcing contact ablation state evaluating method neural network based |
| CN109269557A (en) * | 2018-09-19 | 2019-01-25 | 中国南方电网有限责任公司超高压输电公司广州局 | A kind of change of current station equipment operating parameter and running environment intelligent monitor system and method |
| CN109813987B (en) * | 2019-03-13 | 2020-12-01 | 国网陕西省电力公司电力科学研究院 | Transformer direct-current magnetic bias state evaluation method based on BP neural network |
| CN110943543A (en) * | 2019-12-14 | 2020-03-31 | 启迪中电智慧能源科技(深圳)有限公司 | Monitoring system applied to data of high-voltage direct-current power transmission system |
| CN111734885B (en) * | 2020-06-30 | 2021-11-30 | 全球能源互联网研究院有限公司 | Converter valve on-line monitoring and evaluating method and system |
| CN111896224B (en) * | 2020-08-14 | 2022-04-12 | 南方电网科学研究院有限责任公司 | Laser power supply loop performance detection device and method and terminal equipment |
| CN112098758B (en) * | 2020-09-18 | 2022-06-24 | 国网湖南省电力有限公司 | Test platform and test method for extra-high voltage direct current deep well grounding electrode |
-
2021
- 2021-01-18 CN CN202110063456.2A patent/CN112883634B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN112883634A (en) | 2021-06-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112883634B (en) | DC measurement system state prediction method and system based on multi-dimensional analysis | |
| CN111624494B (en) | Battery analysis method and system based on electrochemical parameters | |
| CN110298455B (en) | Mechanical equipment fault intelligent early warning method based on multivariate estimation prediction | |
| CN113156917B (en) | Power grid equipment fault diagnosis method and system based on artificial intelligence | |
| CN105300692B (en) | A kind of bearing failure diagnosis and Forecasting Methodology based on expanded Kalman filtration algorithm | |
| KR102148761B1 (en) | Apparatus diagnosing fault of photovoltaic system | |
| CN105846780A (en) | Decision tree model-based photovoltaic assembly fault diagnosis method | |
| CN110311398B (en) | Connection topology, control system and method of novel energy storage battery system | |
| CN109813544B (en) | A kind of rotating machinery Incipient Fault Diagnosis method and system based on on-line monitoring | |
| CN111669123B (en) | Method and device for fault diagnosis of photovoltaic string | |
| CN113759868B (en) | Industrial Ethernet fault prediction method based on neural network | |
| CN114252103B (en) | Fusion power station operation fault prediction method | |
| CN112149877B (en) | Multi-source data driven fault prediction method and system for multi-element complex urban power grid | |
| CN110866366A (en) | XGboost algorithm-based island detection method for photovoltaic microgrid containing PHEV | |
| CN116308304A (en) | New energy intelligent operation and maintenance method and system based on meta learning concept drift detection | |
| CN105334472A (en) | Online remaining life prediction method for mining intrinsic safety power supply | |
| CN102095953B (en) | A kind of performance of accumulator charger online test method | |
| CN105741184A (en) | Transformer state evaluation method and apparatus | |
| CN107133674B (en) | Method and system for determining integral state of converter valve | |
| CN114499409A (en) | Photovoltaic module hot spot fault diagnosis method | |
| CN114400776A (en) | Substation automation equipment state diagnosis method and system based on digital mirror image | |
| CN113722977B (en) | Gas turbine rotor fault early warning method based on hybrid prediction | |
| CN117394409B (en) | Intelligent assessment method and system for equipment state of energy storage power station | |
| Liu et al. | State fusion evaluation and fusion compression method of multi-source sensors in power distribution Internet of things | |
| CN117607751A (en) | Superconducting cable state evaluation system based on vector machine and random forest model |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |