CN115563543B - Channel quality multi-terminal electric quantity protection principle selection method, device and system - Google Patents
Channel quality multi-terminal electric quantity protection principle selection method, device and system Download PDFInfo
- Publication number
- CN115563543B CN115563543B CN202211336269.8A CN202211336269A CN115563543B CN 115563543 B CN115563543 B CN 115563543B CN 202211336269 A CN202211336269 A CN 202211336269A CN 115563543 B CN115563543 B CN 115563543B
- Authority
- CN
- China
- Prior art keywords
- protection
- protection principle
- current
- principle
- transmission line
- 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
- 238000010187 selection method Methods 0.000 title abstract description 10
- 238000004891 communication Methods 0.000 claims abstract description 137
- 230000005540 biological transmission Effects 0.000 claims abstract description 86
- 238000012549 training Methods 0.000 claims description 59
- 238000013528 artificial neural network Methods 0.000 claims description 52
- 238000004422 calculation algorithm Methods 0.000 claims description 46
- 230000006870 function Effects 0.000 claims description 29
- 238000004364 calculation method Methods 0.000 claims description 11
- 238000002474 experimental method Methods 0.000 claims description 6
- 238000005457 optimization Methods 0.000 claims description 5
- 238000010606 normalization Methods 0.000 claims description 3
- 239000012528 membrane Substances 0.000 claims description 2
- 238000000034 method Methods 0.000 description 23
- 230000008569 process Effects 0.000 description 6
- 230000008859 change Effects 0.000 description 3
- 230000007774 longterm Effects 0.000 description 3
- 210000002569 neuron Anatomy 0.000 description 3
- 238000013473 artificial intelligence Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000009434 installation Methods 0.000 description 2
- 238000005070 sampling Methods 0.000 description 2
- 230000003044 adaptive effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000002955 isolation Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 230000035800 maturation Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000007430 reference method Methods 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02H—EMERGENCY PROTECTIVE CIRCUIT ARRANGEMENTS
- H02H7/00—Emergency protective circuit arrangements specially adapted for specific types of electric machines or apparatus or for sectionalised protection of cable or line systems, and effecting automatic switching in the event of an undesired change from normal working conditions
- H02H7/26—Sectionalised protection of cable or line systems, e.g. for disconnecting a section on which a short-circuit, earth fault, or arc discharge has occured
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W24/00—Supervisory, monitoring or testing arrangements
- H04W24/08—Testing, supervising or monitoring using real traffic
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Mobile Radio Communication Systems (AREA)
Abstract
The invention relates to the technical field of relay protection of power systems, and particularly discloses a channel quality multi-terminal electric quantity protection principle-based selection method, which comprises the following steps: acquiring communication quality information of a current wireless communication channel and a current rapid protection principle of a power transmission line; judging whether to send out an instruction for switching the current fast protection principle according to the communication quality information of the current wireless communication channel and the current fast protection principle of the power transmission line; and if an instruction for switching the current rapid protection principle is sent, switching the current rapid protection principle of the power transmission line. The invention also discloses a device and a system for selecting the multi-terminal electric quantity protection principle based on the channel quality. The channel quality multi-terminal electric quantity protection principle selection method provided by the invention can select the protection principle according to the optimal communication quality of the wireless communication channel, reduce the probability of protection failure caused by unstable wireless signals, and improve the protection reliability and the protection performance.
Description
Technical Field
The invention relates to the technical field of relay protection of power systems, in particular to a channel quality multi-terminal electric quantity protection principle selection method, a channel quality multi-terminal electric quantity protection principle selection device and a channel quality multi-terminal electric quantity protection principle selection system comprising the channel quality multi-terminal electric quantity protection principle selection device.
Background
The power distribution network line protection based on high-speed wireless communication needs to collect electric quantity of different points in a power grid and transmits the electric quantity to the protection device on the side through a communication channel, and when the power grid breaks down, the protection device completes fault isolation according to a rapid protection principle. The adopted fast protection principle is usually a differential protection principle for transmitting analog quantity or a longitudinal switch value protection principle for transmitting switch value (comprising longitudinal direction protection, network topology protection, digital overcurrent protection and the like).
The differential protection channel needs to transmit sampling value information of currents at different points in the power grid in real time, the transmission information quantity is large, the technical requirements on communication time delay and the like are high, and an external clock is needed to synchronize time. The pilot switch quantity protection transmits logic signals in the communication channel, the transmission information quantity is less, the requirement on communication delay is lower, and the time synchronization is not required. But differential protection has advantages in aspects of protection speed, reliability, sensitivity and the like under power grid faults. Considering that wireless communication is subject to a plurality of external interference factors, in order to avoid the failure of differential protection caused by fluctuation of communication quality and influence on protection reliability, the principle of differential protection or longitudinal switch value protection is adopted according to the communication quality of the installation part of the protection device. The method generally depends on manual experience, cannot realize real-time control on switching of the protection principle, is difficult to cope with abrupt change of communication quality of a wireless channel, and cannot guarantee reliability of the protection device in the operation process.
Disclosure of Invention
Aiming at the defects and shortcomings in the prior art, the invention provides a channel quality multi-terminal-based electric quantity protection principle selection method, a channel quality multi-terminal-based electric quantity protection principle selection device and a channel quality multi-terminal-based electric quantity protection principle selection system comprising the channel quality multi-terminal-based electric quantity protection principle selection device, which are used for solving the problems that the real-time control on the switching of the protection principle cannot be realized by generally relying on manual experience in the prior art, the abrupt change of the communication quality of a wireless channel cannot be dealt with, and the reliability of the protection device in the operation process cannot be ensured.
As a first aspect of the present invention, there is provided a channel quality multi-terminal based electric quantity protection principle selection method, including:
step S1: acquiring communication quality information of a current wireless communication channel and a current rapid protection principle of a power transmission line;
step S2: judging whether to send out an instruction for switching the current quick protection principle according to the communication quality information of the current wireless communication channel and the current quick protection principle of the power transmission line;
step S3: and if an instruction for switching the current rapid protection principle is sent out, switching the current rapid protection principle of the power transmission line.
Further, the method further comprises the following steps:
inputting the communication quality information of the current wireless communication channel and the current rapid protection principle of the power transmission line into a training model to output the target rapid protection principle of the power transmission line;
judging whether the target rapid protection principle of the power transmission line is consistent with the current rapid protection principle, if not, sending out an instruction for switching the current rapid protection principle so as to switch the current rapid protection principle of the power transmission line into the target rapid protection principle.
As another aspect of the present invention, there is provided a channel quality multi-terminal electric quantity protection principle selection device including:
the acquisition module is used for acquiring the communication quality information of the current wireless communication channel and the current rapid protection principle of the power transmission line;
the judging module is used for judging whether to send out an instruction for switching the current quick protection principle according to the communication quality information of the current wireless communication channel and the current quick protection principle of the power transmission line; and if an instruction for switching the current rapid protection principle is sent out, switching the current rapid protection principle of the power transmission line.
As another aspect of the present invention, a channel quality multi-terminal electric quantity protection principle selection system is provided, including two protection devices, the two protection devices are a master side protection device and a slave side protection device, the master side protection device and the slave side protection device are respectively connected with a current transformer and a voltage transformer, the master side protection device and the slave side protection device are connected through a wireless communication channel, wherein the master side protection device includes a first protection criterion operation module, an agent and a first communication device, the slave side protection device includes a second protection criterion operation module and a second communication device, the first communication device and the second communication device communicate through the wireless communication channel, and the agent includes the channel quality multi-terminal electric quantity protection principle selection device described above;
collecting electric quantity at two ends of a power transmission line through current transformers and voltage transformers at two sides, transmitting the electric quantity at two ends of the power transmission line through communication equipment and a wireless communication channel, and finally calculating through protection criterion operation modules at two sides, so that a rapid protection function for the power transmission line is realized;
acquiring communication quality information of a current wireless communication channel and a current quick protection principle of a power transmission line through an intelligent agent, and judging whether to send out an instruction for switching the current quick protection principle;
if the intelligent agent judges to send the instruction for switching the current rapid protection principle, the instruction for switching the current rapid protection principle is respectively sent to a first protection criterion operation module and a second protection criterion operation module, wherein the intelligent agent directly sends the instruction for switching the current rapid protection principle to the first protection criterion operation module, and meanwhile, the intelligent agent sends the instruction for switching the current rapid protection principle to the second protection criterion operation module through the wireless communication channel, and the first protection criterion operation module and the second protection criterion operation module switch the current rapid protection principle of the power transmission line to the target rapid protection principle after receiving the instruction for switching the current rapid protection principle.
Further, the first protection criterion operation module transmits the electric quantity at two ends of the electric transmission line on the main side to the second protection criterion operation module on the secondary side through the wireless communication channel, the second protection criterion operation module transmits the electric quantity at two ends of the electric transmission line on the secondary side to the first protection criterion operation module on the main side through the wireless communication channel, and the second protection criterion operation module on the secondary side and the first protection criterion operation module on the main side calculate the electric quantity at two ends of the electric transmission line received, so that the rapid protection function of the electric transmission line is realized;
assuming that the current rapid protection principle of the power transmission line is a longitudinal switch value protection principle, wherein the target rapid protection principle is a differential protection principle, acquiring communication quality information of a current wireless communication channel and the current rapid protection principle of the power transmission line through an intelligent agent, and judging whether to send out an instruction for switching the current rapid protection principle;
if the intelligent agent judges to send the instruction for switching the current rapid protection principle, the instruction for switching the current rapid protection principle is respectively sent to a first protection criterion operation module and a second protection criterion operation module, and after receiving the instruction for switching the current rapid protection principle, the first protection criterion operation module and the second protection criterion operation module switch the longitudinal switching value protection principle of the power transmission line into a differential protection principle.
Further, the communication quality information of the current wireless communication channel and the current rapid protection principle of the power transmission line are input into the training model of the intelligent body so as to output the target rapid protection principle of the power transmission line;
the intelligent agent judges whether the target rapid protection principle of the power transmission line is consistent with the current rapid protection principle, if not, an instruction for switching the current rapid protection principle is sent out, so that the current rapid protection principle of the power transmission line is switched into the target rapid protection principle.
Further, the training model of the intelligent agent is a BP_AdaBoost strong classifier model, wherein the BP_AdaBoost strong classifier model is a strong classifier obtained by integrating a plurality of BP neural network weak classifiers through an Adaboost algorithm by taking a BP neural network as a weak classifier;
simulating a large number of communication state scenes in a movable mould experiment, quantifying an operation result of the movable mould experiment, and obtaining a large number of discrete data, wherein the discrete data comprises: the characteristic quantity of the communication quality of the wireless communication channel and the current rapid protection principle are represented by the on-off of the channel, the error rate of the channel, the delay jitter value and the frame loss number; the discrete data are used as training data of the BP_AdaBoost strong classifier model, input into the BP_AdaBoost strong classifier model and output as a target fast protection principle of a power transmission line;
inputting the training data into the BP_AdaBoost strong classifier model for classification training to obtain a trained BP_AdaBoost strong classifier model;
the trained intelligent agent can judge whether the current quick protection principle needs to be switched according to the communication quality information of the current wireless communication channel and the current quick protection principle of the power transmission line, and if so, a corresponding instruction is sent out.
Further, the Adaboost algorithm includes the following steps:
(1) Selecting a learning algorithm as a weak learning algorithm, and giving the weak learning algorithmDetermining sample space (X, Y), finding m groups of data from the sample space as training data, and setting initial weight of each group of training data as
(2) Carrying out T times of iterative operation by using a weak learning algorithm, and updating the weight of each group of training data according to the classification result after each time of operation;
(3) The weak classifier obtains a classification function sequence f through repeated calculation 1 ,f 2 …f T Each classification function is given a weight;
(4) After T iterations, the final strong classification function F is weighted by the weak classification function.
Further, the steps of the bp_adaboost algorithm are as follows:
(1) Randomly selecting m groups of training data from training data samples, and initial weightsDetermining the structure of the BP neural network according to the dimensionality of the input and output data, and initializing the weight and the threshold of the BP neural network;
(2) When training the kth BP neural network, training the BP neural network by using training data and predicting the output of the training data, wherein the sum of errors of the prediction output is e k Error sum e k The calculation formula of (2) is
Wherein g (k) is a predicted result and y is a desired result;
(3) Based on the error sum e of the prediction output k Calculating weight a of BP neural network k The weight calculation formula is as follows
(4) According to the calculated weight a of the kth BP neural network k To adjust the weight of the training sample of the next round, the specific relation is that
In the above, B k Is a normalization factor;
(5) After T-round training, weak classification function f (g) of the T-group BP neural network is obtained k ,a k ) From the T groups of weak classification functions f (g k ,a k ) Combining to obtain a strong classification function
Further, combining an MEA (membrane electrode assembly) algorithm and an LM (model-based algorithm) algorithm to improve the BP neural network, optimizing an initial weight and a threshold of the BP neural network by the MEA algorithm, and giving the optimized weight and threshold to the BP neural network to obtain an optimized BP neural network; and then training the optimized BP neural network according to an LM algorithm, thereby obtaining the MEA_LM_BP neural network subjected to two-step optimization.
The channel quality multi-terminal-based electric quantity protection principle selection method provided by the invention has the following advantages:
(1) Based on an artificial intelligence algorithm, the method can realize real-time judgment of the quality of communication signals and select a proper protection principle, and compared with a method of selecting the protection principle in advance by means of artificial experience, the method can adapt to the change of the quality of a wireless channel in real time, optimize the selection of the protection principle, effectively solve the problem of protection failure caused by unstable signals, and simultaneously improve the protection speed and reliability as much as possible;
(2) Taking long-term related data of the protected place as a data set to train the intelligent agent offline; the accuracy and reliability of the judgment of the intelligent agent are ensured by a large amount of training data, and compared with the judgment by means of manual experience, the judgment method has higher judgment precision and accuracy; with the rapid development of artificial intelligence algorithms, the method has wide application prospect.
Drawings
The accompanying drawings are included to provide a further understanding of the invention, and are incorporated in and constitute a part of this specification, illustrate the invention and together with the description serve to explain, without limitation, the invention.
Fig. 1 is a flow chart of a selection method based on the channel quality multi-terminal electric quantity protection principle.
Fig. 2 is a block diagram of a channel quality multi-terminal electric quantity protection principle selection device.
Fig. 3 is a block diagram of a system for selecting a channel quality multi-terminal electric quantity protection principle.
Fig. 4 is a flowchart of the AdaBoost algorithm provided by the invention.
Fig. 5 is a flowchart of calculation of the BP neural network provided by the present invention.
Detailed Description
It should be noted that, without conflict, the embodiments of the present invention and features of the embodiments may be combined with each other. The invention will be described in detail below with reference to the drawings in connection with embodiments.
In order that those skilled in the art will better understand the present invention, a technical solution in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in which it is apparent that the described embodiments are only some embodiments of the present invention, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the present invention without making any inventive effort, shall fall within the scope of the present invention.
It should be noted that the terms "first," "second," and the like in the description and the claims of the present invention and the above figures are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate in order to describe the embodiments of the invention herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations 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.
In this embodiment, a method for selecting a channel quality multi-terminal based electric quantity protection principle is provided, and fig. 1 is a flowchart of the method for selecting a channel quality multi-terminal based electric quantity protection principle provided by the present invention, as shown in fig. 1, where the method for selecting a channel quality multi-terminal based electric quantity protection principle includes:
step S1: acquiring communication quality information of a current wireless communication channel and a current rapid protection principle of a power transmission line;
step S2: judging whether to send out an instruction for switching the current quick protection principle according to the communication quality information of the current wireless communication channel and the current quick protection principle of the power transmission line;
step S3: and if an instruction for switching the current rapid protection principle is sent out, switching the current rapid protection principle of the power transmission line.
The invention provides a channel quality multi-terminal electric quantity protection principle selection method, which is based on the characteristic that longitudinal switch quantity protection and differential protection have different requirements on communication technology, and based on the characteristic quantities of delay, jitter, frame loss, error rate and the like of transmission signals, the characteristic quantities are used as the basis for judging the quality of the communication of a wireless communication channel, and the acquired historical data information of a protection device is used as a data set for offline training of an intelligent body, so that the intelligent body can select the protection principle according to the communication quality of the communication channel at the installation position of the protection device. When the communication quality is in the boundary condition of two kinds of protection requirements, the method can accurately select a proper protection mode, and ensure the effectiveness of the protection function. The wireless communication is easy to be interfered by external environment in the propagation process, so that when the protection device selects differential protection with higher communication requirements, the situation that the communication quality cannot meet the requirements possibly occurs in the operation process. According to the method, a protection principle is selected according to the communication quality optimization of the communication channel, the probability of protection failure caused by unstable wireless signals is reduced, the protection reliability is improved, the protection performance is improved, and a good foundation is laid for the application of wireless communication in multi-terminal electric quantity protection.
Preferably, the method further comprises:
inputting the communication quality information of the current wireless communication channel and the current rapid protection principle of the power transmission line into a training model to output the target rapid protection principle of the power transmission line;
judging whether the target rapid protection principle of the power transmission line is consistent with the current rapid protection principle, if not, sending out an instruction for switching the current rapid protection principle so as to switch the current rapid protection principle of the power transmission line into the target rapid protection principle.
As another embodiment of the present invention, there is provided a channel quality multi-terminal electric quantity protection principle selection device, as shown in fig. 2, wherein the channel quality multi-terminal electric quantity protection principle selection device includes:
the acquisition module is used for acquiring the communication quality information of the current wireless communication channel and the current rapid protection principle of the power transmission line;
the judging module is used for judging whether to send out an instruction for switching the current quick protection principle according to the communication quality information of the current wireless communication channel and the current quick protection principle of the power transmission line; and if an instruction for switching the current rapid protection principle is sent out, switching the current rapid protection principle of the power transmission line.
As another embodiment of the present invention, a channel quality multi-terminal electric quantity protection principle selection system is provided, as shown in fig. 3, the channel quality multi-terminal electric quantity protection principle selection system includes protection devices including two sides, the two sides of the protection devices are a master side protection device and a slave side protection device, the master side protection device and the slave side protection device are respectively connected with a current transformer and a voltage transformer, the master side protection device and the slave side protection device are connected through a wireless communication channel, wherein the master side protection device includes a first protection criterion operation module, an agent and a first communication device, the slave side protection device includes a second protection criterion operation module and a second communication device, the first communication device and the second communication device communicate through the wireless communication channel, and the agent includes the channel quality multi-terminal electric quantity protection principle selection device described above;
collecting electric quantity at two ends of a power transmission line through current transformers and voltage transformers at two sides, transmitting the electric quantity at two ends of the power transmission line through communication equipment and a wireless communication channel, and finally calculating through protection criterion operation modules at two sides, so that a rapid protection function for the power transmission line is realized;
acquiring communication quality information of a current wireless communication channel and a current quick protection principle of a power transmission line through an intelligent agent, and judging whether to send out an instruction for switching the current quick protection principle;
if the intelligent agent judges to send the instruction for switching the current rapid protection principle, the instruction for switching the current rapid protection principle is respectively sent to a first protection criterion operation module and a second protection criterion operation module, wherein the intelligent agent directly sends the instruction for switching the current rapid protection principle to the first protection criterion operation module, and meanwhile, the intelligent agent sends the instruction for switching the current rapid protection principle to the second protection criterion operation module through the wireless communication channel, and the first protection criterion operation module and the second protection criterion operation module switch the current rapid protection principle of the power transmission line to the target rapid protection principle after receiving the instruction for switching the current rapid protection principle.
It should be noted that, for the time difference problem between the fast protection principles of the protection criterion operation modules on two sides, the time difference problem is solved by installing the time setting device, the first protection criterion operation module on the main side will not switch the current fast protection principle immediately after receiving the switching instruction, and the current fast protection principle is switched together with the second protection criterion operation module on the slave side.
Specifically, the protection criterion operation modules at two sides simultaneously switch the current rapid protection principle in the next whole second after the switching instruction is sent from the intelligent agent. The switching instruction is prevented from being sent as soon as the next whole second, so that if the switching is performed at the same time in the next whole second, there may be a case that the second protection criterion operation module at the slave side has not received the switching instruction yet, and in order to avoid this case, the current fast protection principle is selected to be switched at the same time in the next whole second.
If the intelligent body is installed only in the protection device on one side, the instruction of switching the fast protection principle in the protection devices on both sides is less likely to be different. In order to prevent the occurrence of such a small probability event, which causes the selection of the fast protection principle to be inconsistent and causes the failure to occur, the intelligent body is installed in the one-side protection device (i.e. in a master-slave mode, one side on which the intelligent body is installed is the master side, and the other side is the slave side), and the switching instruction sent by the intelligent body is transmitted to the other side through the wireless communication channel.
Specifically, an intelligent body is installed in the protection device, the communication quality information of the current wireless communication channel and the current rapid protection principle of the power transmission line are input to the protection device, and an instruction of whether to switch the rapid protection principle is sent after the intelligent body is operated. Because the differential protection channel needs to transmit sampling value information of currents at different points in the power grid in real time, the transmission information quantity is large, the technical requirements on communication time delay and the like are high, and the external clock is needed to synchronize time. The pilot switch quantity protection transmits logic signals in the communication channel, the transmission information quantity is less, the requirement on communication delay is lower, and the time synchronization is not required. Therefore, when the communication quality of the current wireless communication channel is good, differential protection is selected, and when the communication quality of the current wireless communication channel is poor, longitudinal switch value protection is selected.
Preferably, the first protection criterion operation module transmits the electric quantity at two ends of the electric transmission line on the main side to the second protection criterion operation module on the secondary side through the wireless communication channel, the second protection criterion operation module transmits the electric quantity at two ends of the electric transmission line on the secondary side to the first protection criterion operation module on the main side through the wireless communication channel, and the second protection criterion operation module on the secondary side and the first protection criterion operation module on the main side calculate the electric quantity at two ends of the electric transmission line received, so that a rapid protection function for the electric transmission line is realized;
assuming that the current rapid protection principle of the power transmission line is a longitudinal switch value protection principle, wherein the target rapid protection principle is a differential protection principle, acquiring communication quality information of a current wireless communication channel and the current rapid protection principle of the power transmission line through an intelligent agent, and judging whether to send out an instruction for switching the current rapid protection principle;
if the intelligent agent judges to send the instruction for switching the current rapid protection principle, the instruction for switching the current rapid protection principle is respectively sent to a first protection criterion operation module and a second protection criterion operation module, and after receiving the instruction for switching the current rapid protection principle, the first protection criterion operation module and the second protection criterion operation module switch the longitudinal switching value protection principle of the power transmission line into a differential protection principle.
Preferably, the communication quality information of the current wireless communication channel and the current rapid protection principle of the power transmission line are input into a training model of the intelligent body so as to output the target rapid protection principle of the power transmission line;
the intelligent agent judges whether the target rapid protection principle of the power transmission line is consistent with the current rapid protection principle, if not, an instruction for switching the current rapid protection principle is sent out, so that the current rapid protection principle of the power transmission line is switched into the target rapid protection principle.
A reference method for offline training of an agent is given below:
preferably, the training model of the intelligent agent is a BP_AdaBoost strong classifier model, and the BP_AdaBoost strong classifier model is a strong classifier obtained by integrating a plurality of BP neural network weak classifiers through an Adaboost algorithm by taking a BP neural network as a weak classifier;
1. a large number of communication state scenes are simulated in the movable mould experiment, and the operation result of the movable mould experiment is quantized according to long-term operation experience and related operation standards, so that whether the longitudinal switch value protection or the differential protection is selected under a certain communication quality condition is judged.
2. A large amount of discrete data can be obtained through the above simulation and calculation. The discrete data includes: the characteristic quantity of the communication quality of the wireless communication channel and the current rapid protection principle are represented by the on-off of the channel, the error rate of the channel, the delay jitter value and the frame loss number; the discrete data are used as training data of the BP_AdaBoost strong classifier model, are input into the BP_AdaBoost strong classifier model, and are output to be a target fast protection principle of a power transmission line, namely, are output to be a selected protection type (longitudinal switching value protection or differential protection);
3. inputting the training data into the BP_AdaBoost (backstPropagation_ Adaptive Boosting) strong classifier model for classification training to obtain a trained BP_AdaBoost strong classifier model;
the trained intelligent agent can judge whether the current fast protection principle needs to be switched according to the communication quality information of the current wireless communication channel and the current fast protection principle of the power transmission line, and if so, a corresponding instruction is sent out.
It should be noted that, a layer of comparison operation is added before the output of the intelligent agent, so as to compare whether the principle of rapid protection of the input of the intelligent agent is consistent with the principle of rapid protection of the output. If the current fast protection principle is inconsistent, an instruction for switching the current fast protection principle is sent out, so that the current fast protection principle is switched into the target fast protection principle.
The BP_AdaBoost strong classifier model is described below:
preferably, as shown in fig. 4, the idea of the AdaBoost algorithm is to integrate the outputs of multiple weak classifier (BP neural network) models to generate an effective strong classifier, and the following steps of the AdaBoost algorithm are as follows:
(1) Selecting one learning algorithm as weak learning algorithm, giving sample space (X, Y), finding m groups of data from the sample space as training data, and setting initial weight of each group of training data as
(2) Carrying out T times of iterative operation by using a weak learning algorithm, and updating the weight of each group of training data according to the classification result after each time of operation; (greater weight is given to training data that failed classification, i.e., more emphasis is given to these training data at the next iteration).
(3) The weak classifier obtains a classification function sequence f through repeated calculation 1 ,f 2 …f T Each classification function is given a weight (the better the classification result, the greater its corresponding weight);
(4) After T iterations, the final strong classification function F is weighted by the weak classification function.
Preferably, in connection with fig. 5, the steps of the bp_adaboost algorithm are as follows:
(1) Randomly selecting m groups of training data from training data samples, and initial weightsDetermining the structure of the BP neural network according to the dimensionality of the input and output data, and initializing the weight and the threshold of the BP neural network;
(2) When training the kth BP neural network, training the BP neural network by using training data and predicting the output of the training data, wherein the sum of errors of the prediction output is e k Error sum e k The calculation formula of (2) is
Wherein g (k) is a predicted result and y is a desired result;
(3) Based on the error sum e of the prediction output k Calculating weight a of BP neural network k The weight calculation formula is as follows
(4) According to the calculated weight a of the kth BP neural network k To adjust the weight of the training sample of the next round, the specific relation is that
In the above, B k Is a normalization factor;
(5) After T-round training, weak classification function f (g) of the T-group BP neural network is obtained k ,a k ) From the T groups of weak classification functions f (g k ,a k ) Combining to obtain a strong classification function
The traditional BP neural network training has a slower convergence speed and is easy to fall into local optimum. The method comprises the steps of combining an MEA algorithm and an LM algorithm to improve a BP neural network, optimizing an initial weight and a threshold of the BP neural network by the MEA algorithm, and giving the optimized weight and threshold to the BP neural network to obtain an optimized BP neural network; and then training the optimized BP neural network according to an LM algorithm, thereby obtaining the MEA_LM_BP neural network subjected to two-step optimization. The algorithm framework structure is shown in table 1 below.
TABLE 1
The specific flow of the MEA algorithm for improving the BP neural network is as follows:
1. and initializing parameters according to the sample data. The main parameters comprise population size, number of winning sub-population, number of temporary sub-population, sub-population size, number of input layer neurons, number of hidden layer neurons, number of output layer neurons and iteration times.
2. An initial population is generated. The initial population includes a winning sub-population and a provisional sub-population.
3. The populations converge. Judging whether each population meets the dissimilarity condition by using a population maturation discriminant function.
4. If the condition is satisfied and the number of iterations has not reached the set iteration maximum, then a dissimilation operation is performed and a new sub-population is replenished.
5. And outputting an optimal individual after the algorithm meets the iteration stop condition, and performing decoding operation on the optimal individual to obtain the initial weight and the threshold of the BP neural network after optimization.
The LM algorithm further improves the BP neural network, specifically as follows:
the core of the LM algorithm is to adjust the weight size by a scaling factor μ, when μ=0, the LM algorithm is identical to the newton algorithm; when μ is a larger value, the LM algorithm is the same as the gradient descent algorithm.
The LM algorithm corrects the weight of the BP neural network according to the following
ω(n+1)＝ω(n)-[J T J+μI] -1 J T e
Where e is the error variable and I is the input variable. J is a jacobian matrix that contains the first derivative of the error performance function.
The LM algorithm has a fast convergence speed and avoids falling into local optima, and requires relatively fewer iterations.
In the embodiment of the invention, when the protection principle is frequently changed, the protection failure and other problems are caused. In order to prevent the agent from frequently sending a switching instruction due to large fluctuation in communication quality, a discrimination program is added to the agent. The discrimination program in the intelligent agent compares the received data with the communication quality data of the previous time, and if the fluctuation of the data is large (different judgment standards for each communication quality index exist and can be formulated according to the actual situation), the intelligent agent stops the protection principle switching.
In addition, considering that the pilot switching value protection has lower communication requirement, the intelligent agent is blocked for a longer time when switching from the differential protection with high communication quality requirement to the pilot protection; when switching from pilot protection mode to differential protection mode, the agent is blocked for a short period of time. The specific lock-out time is set by relevant regulations and long-term operating experience.
The method can effectively prevent the problem of faults caused by frequent replacement of the rapid protection principle due to large fluctuation of communication quality.
It is to be understood that the above embodiments are merely illustrative of the application of the principles of the present invention, but not in limitation thereof. Various modifications and improvements may be made by those skilled in the art without departing from the spirit and substance of the invention, and are also considered to be within the scope of the invention.
Claims (4)
1. The utility model provides a system is selected to electric quantity protection principle based on passageway quality multiport, its characterized in that includes the protection device of both sides, and the protection device of both sides is main side protection device and from side protection device respectively, main side protection device and from side protection device all connect current transformer and voltage transformer respectively, connect through wireless communication channel between main side protection device and the side protection device, wherein, main side protection device includes first protection criterion operation module, agent and first communication equipment, from side protection device includes second protection criterion operation module and second communication equipment, pass through between first communication equipment and the second communication equipment wireless communication channel communication, the agent includes and selects the device based on passageway quality multiport electric quantity protection principle, select the device based on passageway quality multiport electric quantity protection principle includes: the acquisition module is used for acquiring the communication quality information of the current wireless communication channel and the current rapid protection principle of the power transmission line; the judging module is used for judging whether to send out an instruction for switching the current quick protection principle according to the communication quality information of the current wireless communication channel and the current quick protection principle of the power transmission line; if an instruction for switching the current rapid protection principle is sent out, switching the current rapid protection principle of the power transmission line;
collecting electric quantity at two ends of a power transmission line through current transformers and voltage transformers at two sides, transmitting the electric quantity at two ends of the power transmission line through communication equipment and a wireless communication channel, and finally calculating through protection criterion operation modules at two sides, so that a rapid protection function for the power transmission line is realized;
acquiring communication quality information of a current wireless communication channel and a current quick protection principle of a power transmission line through an intelligent agent, and judging whether to send out an instruction for switching the current quick protection principle;
if the intelligent agent judges to send the instruction for switching the current rapid protection principle, the instruction for switching the current rapid protection principle is respectively sent to a first protection criterion operation module and a second protection criterion operation module, wherein the intelligent agent directly sends the instruction for switching the current rapid protection principle to the first protection criterion operation module, and meanwhile, the intelligent agent sends the instruction for switching the current rapid protection principle to the second protection criterion operation module through the wireless communication channel, and the first protection criterion operation module and the second protection criterion operation module switch the current rapid protection principle of the power transmission line to the target rapid protection principle after receiving the instruction for switching the current rapid protection principle;
the communication quality information of the current wireless communication channel and the current rapid protection principle of the power transmission line are input into the training model of the intelligent body so as to output the target rapid protection principle of the power transmission line;
the intelligent agent judges whether the target rapid protection principle of the power transmission line is consistent with the current rapid protection principle, if not, an instruction for switching the current rapid protection principle is sent out, so that the current rapid protection principle of the power transmission line is switched into the target rapid protection principle;
the training model of the intelligent agent is a BP_AdaBoost strong classifier model, wherein the BP_AdaBoost strong classifier model is a strong classifier obtained by integrating a plurality of BP neural network weak classifiers through an Adaboost algorithm by taking a BP neural network as a weak classifier;
simulating a large number of communication state scenes in a movable mould experiment, quantifying an operation result of the movable mould experiment, and obtaining a large number of discrete data, wherein the discrete data comprises: the characteristic quantity of the communication quality of the wireless communication channel and the current rapid protection principle are represented by the on-off of the channel, the error rate of the channel, the delay jitter value and the frame loss number; the discrete data are used as training data of the BP_AdaBoost strong classifier model, input into the BP_AdaBoost strong classifier model and output as a target fast protection principle of a power transmission line;
inputting the training data into the BP_AdaBoost strong classifier model for classification training to obtain a trained BP_AdaBoost strong classifier model;
the trained intelligent agent can judge whether the current quick protection principle needs to be switched according to the communication quality information of the current wireless communication channel and the current quick protection principle of the power transmission line, and if so, a corresponding instruction is sent out;
the first protection criterion operation module transmits the electric quantity at two ends of the power transmission line at the main side to the second protection criterion operation module at the secondary side through the wireless communication channel, the second protection criterion operation module transmits the electric quantity at two ends of the power transmission line at the secondary side to the first protection criterion operation module at the main side through the wireless communication channel, and the second protection criterion operation module at the secondary side and the first protection criterion operation module at the main side both calculate the electric quantity at two ends of the received power transmission line, so that the rapid protection function of the power transmission line is realized;
assuming that the current rapid protection principle of the power transmission line is a longitudinal switch value protection principle, wherein the target rapid protection principle is a differential protection principle, acquiring communication quality information of a current wireless communication channel and the current rapid protection principle of the power transmission line through an intelligent agent, and judging whether to send out an instruction for switching the current rapid protection principle;
if the intelligent agent judges to send the instruction for switching the current rapid protection principle, the instruction for switching the current rapid protection principle is respectively sent to a first protection criterion operation module and a second protection criterion operation module, and after receiving the instruction for switching the current rapid protection principle, the first protection criterion operation module and the second protection criterion operation module switch the longitudinal switching value protection principle of the power transmission line into a differential protection principle.
2. The channel quality multi-terminal electric quantity protection principle selection system according to claim 1, wherein the step of the Adaboost algorithm is as follows:
(1) Selecting one learning algorithm as weak learning algorithm, giving sample space (X, Y), finding m groups of data from the sample space as training data, and setting initial weight of each group of training data as
(2) Carrying out T times of iterative operation by using a weak learning algorithm, and updating the weight of each group of training data according to the classification result after each time of operation;
(3) The weak classifier obtains a classification function sequence f through repeated calculation 1 ,f 2 Lf T Each classification function is given a weight;
(4) After T iterations, the final strong classification function F is weighted by the weak classification function.
3. The channel quality multi-terminal electric quantity protection principle selection system according to claim 2, wherein the steps of the bp_adaboost algorithm are as follows:
(1) Randomly selecting m groups of training data from training data samples, and initial weightsDetermining the structure of the BP neural network according to the dimensionality of the input and output data, and initializing the weight and the threshold of the BP neural network;
(2) When training the kth BP neural network, training the BP neural network by using training data and predicting the output of the training data, wherein the sum of errors of the prediction output is e k Error sum e k The calculation formula of (2) is
Wherein g (k) is a predicted result and y is a desired result;
(3) Based on the error sum e of the prediction output k Calculating weight a of BP neural network k The weight calculation formula is as follows
(4) According to the calculated weight a of the kth BP neural network k To adjust the weight of the training sample of the next round, the specific relation is that
In the above, B k Is a normalization factor;
(5) After T-round training, weak classification function f (g) of the T-group BP neural network is obtained k ,a k ) From the T groups of weak classification functions f (g k ,a k ) Combining to obtain a strong classification function
4. The system for selecting the electric quantity protection principle based on the channel quality multi-terminal according to claim 3, wherein the BP neural network is improved by combining an MEA (membrane electrode assembly) algorithm and an LM (model building algorithm), an initial weight and a threshold value of the BP neural network are optimized by the MEA algorithm, and the optimized weight and threshold value are given to the BP neural network to obtain an optimized BP neural network; and then training the optimized BP neural network according to an LM algorithm, thereby obtaining the MEA_LM_BP neural network subjected to two-step optimization.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211336269.8A CN115563543B (en) | 2022-10-28 | 2022-10-28 | Channel quality multi-terminal electric quantity protection principle selection method, device and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211336269.8A CN115563543B (en) | 2022-10-28 | 2022-10-28 | Channel quality multi-terminal electric quantity protection principle selection method, device and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115563543A CN115563543A (en) | 2023-01-03 |
| CN115563543B true CN115563543B (en) | 2024-03-29 |
Family
ID=84767895
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211336269.8A Active CN115563543B (en) | 2022-10-28 | 2022-10-28 | Channel quality multi-terminal electric quantity protection principle selection method, device and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115563543B (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105680427A (en) * | 2016-03-16 | 2016-06-15 | 华中电网有限公司 | Multi-adaptive optical fiber differential protection method for power transmission line |
| CN111464465A (en) * | 2020-03-11 | 2020-07-28 | 重庆邮电大学 | Channel estimation method based on integrated neural network model |
| CN113810938A (en) * | 2021-06-29 | 2021-12-17 | 荣耀终端有限公司 | Channel switching method, electronic device and storage medium |
| CN114243898A (en) * | 2021-11-08 | 2022-03-25 | 国网河南省电力公司郑州供电公司 | Power distribution network differential protection system and protection method based on 5G communication network |
| CN114400626A (en) * | 2021-12-13 | 2022-04-26 | 北京四方继保工程技术有限公司 | Channel switching-based power transmission line triple main protection implementation method and system |
| CN114678860A (en) * | 2020-12-24 | 2022-06-28 | 中国科学院沈阳自动化研究所 | Power distribution network protection control method and system based on deep reinforcement learning |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US11544540B2 (en) * | 2019-05-10 | 2023-01-03 | Hewlett Packard Enterprise Development Lp | Systems and methods for neural network training and deployment for hardware accelerators |
-
2022
- 2022-10-28 CN CN202211336269.8A patent/CN115563543B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105680427A (en) * | 2016-03-16 | 2016-06-15 | 华中电网有限公司 | Multi-adaptive optical fiber differential protection method for power transmission line |
| CN111464465A (en) * | 2020-03-11 | 2020-07-28 | 重庆邮电大学 | Channel estimation method based on integrated neural network model |
| CN114678860A (en) * | 2020-12-24 | 2022-06-28 | 中国科学院沈阳自动化研究所 | Power distribution network protection control method and system based on deep reinforcement learning |
| CN113810938A (en) * | 2021-06-29 | 2021-12-17 | 荣耀终端有限公司 | Channel switching method, electronic device and storage medium |
| CN114243898A (en) * | 2021-11-08 | 2022-03-25 | 国网河南省电力公司郑州供电公司 | Power distribution network differential protection system and protection method based on 5G communication network |
| CN114400626A (en) * | 2021-12-13 | 2022-04-26 | 北京四方继保工程技术有限公司 | Channel switching-based power transmission line triple main protection implementation method and system |
Non-Patent Citations (3)
| Title |
|---|
| Influence of the channel model in the optimum switching points in a cognitive radio network combined with adaptive modulation;Ana Paula Teles Ribeiro da Silva;《2016 International Symposium on Networks, Computers and Communications (ISNCC)》;20160513;第1-6页 * |
| 不完整量测下网络化***故障检测与控制;万雄波;《中国博士学位论文全文数据库 信息科技辑》;20110915(第9期);第I140-2页 * |
| 张睿智.基于BP-Adaboost强分类器的声音环境识别.《电子设计工程》.2021,第29卷(第9期),第146-150页. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115563543A (en) | 2023-01-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107394773B (en) | Distribution network information physical system reliability assessment method considering fault processing overall process | |
| CN107359617B (en) | Active power distribution network source-network-load coordination control system comprising micro-grid group | |
| CN110988596B (en) | Novel active power distribution network fault section positioning method | |
| CN111697566B (en) | Reliability assessment method for active power distribution network information physical system considering information failure | |
| CN111952972A (en) | Main distribution integrated load transfer method for high-quality power supply service | |
| CN100367620C (en) | Power network topology error identification method based on mixed state estimation | |
| CN104463351A (en) | Communication bandwidth prediction method and device based on power business requirements | |
| CN104934968A (en) | Multi-agent based distribution network disaster responding recovery coordinate control method and multi-agent based distribution network disaster responding recovery coordinate control device | |
| CN105467971A (en) | Electric power secondary equipment monitoring system and method | |
| CN105425768A (en) | Electric power secondary equipment monitoring device and method | |
| CN105207910B (en) | A kind of power telecom network routing optimization method based on particle group optimizing | |
| CN113219328B (en) | Intelligent fault diagnosis method for circuit breaker operating mechanism based on information fusion | |
| CN113162226A (en) | Primary side fault injection FA system level test system and method based on HIL | |
| CN111882125B (en) | Electric power information physical system cascading failure prediction method based on risk element theory | |
| CN112327098A (en) | Power distribution network fault section positioning method based on low-voltage distribution network comprehensive monitoring unit | |
| CN113328437A (en) | CPS (control performance Standard) topology construction method and fault recovery method for intelligent power distribution network | |
| CN109687432B (en) | Multi-stage cooperation-based power distribution terminal site selection and sizing method and system | |
| CN115563543B (en) | Channel quality multi-terminal electric quantity protection principle selection method, device and system | |
| CN115663749B (en) | Real-time selection method, device and system for multi-terminal electrical quantity protection wireless channel | |
| CN109638817B (en) | Power distribution network vulnerability analysis and prevention control method considering user side model | |
| CN115588961A (en) | Setting value self-adaptive setting method based on power distribution network full-model protection | |
| CN106329714A (en) | Method for dynamically judging switch type of power grid | |
| CN112560322B (en) | Transformer substation protection room temperature prediction method based on BP neural network | |
| CN114595543A (en) | Power distribution network topology identification method based on voltage correlation coefficient | |
| CN111025135B (en) | Cloud platform-based test method for on-site FTU coordination control capability |
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 |