CN113077049B - Intelligent regulation and control method for mobile terminal - Google Patents
Intelligent regulation and control method for mobile terminal Download PDFInfo
- Publication number
- CN113077049B CN113077049B CN202110628287.2A CN202110628287A CN113077049B CN 113077049 B CN113077049 B CN 113077049B CN 202110628287 A CN202110628287 A CN 202110628287A CN 113077049 B CN113077049 B CN 113077049B
- Authority
- CN
- China
- Prior art keywords
- server
- mobile terminal
- field
- neural network
- user
- 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
- 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
-
- 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
Abstract
The invention relates to an intelligent regulation and control method for a mobile terminal, which solves the problems in the prior art and has the technical scheme that: the intelligent fault prompting method is matched with a server and a mobile terminal for use, the server is in communication connection with the mobile terminal after being subjected to safety scanning and login verification, and the server is in communication connection with the mobile terminal in an external network communication mode and comprises an initialization step, a fault data self-calculation step of collecting field data, a neural network training step of using a fault data self-calculation result and a fault intelligent prompting step of using a trained neural network. The invention adopts an ingenious mode to realize the source of big data, namely, the opportunity of personal ability test is fully utilized, the ability is trained and the parameters for neural network training are determined in a manual determination mode, and a large amount of available relatively accurate data is obtained by the method to provide a fault solution.
Description
Technical Field
The invention relates to an intelligent regulation and control method suitable for a power grid system, in particular to an intelligent regulation and control method suitable for a mobile terminal of the power grid system.
Background
In recent years, technologies such as artificial intelligence and cloud computing are mature day by day, and the combination of artificial intelligence, cloud computing and power grid regulation and control business is a necessary trend of power grid regulation and control business development, has an important meaning for promoting regulation and control business intellectualization, and lays a solid foundation for intelligent scheduling. Along with the gradual expansion of the scale of a power grid, the accumulation of massive power grid operation data and the incapability of visually displaying the power grid operation data, the integrated display of full models of equipment governed by all levels of scheduling mechanisms is needed in space, the grid network structure change and the power grid historical section operation data are inverted on a time scale, the power grid development perspective planning is carried out, the intelligent monitoring is carried out on the power grid equipment information, the intelligent analysis and reminding are carried out on equipment abnormal information, fault information and the like, the working efficiency is improved, the mobile APP technology is introduced, the service operation cooperativity is enhanced, the key information is obtained more timely, the examination and approval duration of important processes is reduced, and the learning, control and application of the technology are assisted to be improved.
In view of the above technical problems, the existing solutions are as follows: for example, the chinese patent application No. 201610613748.8 discloses a method for determining faults inside and outside a transmission line area and selecting a phase of the faults based on a convolutional neural network; a method for judging faults inside and outside a transmission line area and selecting a phase of the faults based on a convolutional neural network comprises the steps of firstly, building a simulation model according to a schematic diagram of a double-end power supply system to obtain input of a training sample, using the fault type and the faults inside and outside the area as output of the training sample, and similarly generating a test sample. Secondly, listing the network structure, and obtaining the optimal network structure by testing the error rate of the sample. And finally, after each fault, acquiring fault current, inputting the fault current into a trained network, and judging whether the fault occurs in the area or not and performing fault phase selection without retraining. The method improves the output of the convolutional neural network, solves the two types of non-independent classification problems of fault judgment and fault phase selection inside and outside the region simultaneously by using the same network, and realizes weight sharing of the two types of non-independent classification problems. The method has low requirement on the sampling rate, does not need to calculate various setting values, is not influenced by factors such as system frequency, fault positions, load current, transition resistance and the like, and has accurate and reliable results. The method and the system for identifying the fault type of the power transmission line are characterized in that the fault time sequence data of each type of a target line are generated by using a power transmission line fault simulation model, and are processed to generate target domain data samples facing a Convolutional Neural Network (CNN); carrying out convolution kernel migration training on the pre-training model by using the target domain data sample to form a target domain model; and identifying the fault type of the target line by adopting the target domain model. The method can realize the migration of the deep model by using a small amount of data, generate the deep learning model suitable for the target line, avoid the condition that the deep learning model is trained independently aiming at each line, and improve the generalization capability of the deep learning model. At present, intelligent AI fault judgment and intelligent scheduling represented by a convolutional neural network have been applied more maturely, but in the running process of the neural network, a problem needs to be solved, namely, under the condition that power is required to be supplied to a circuit configuration by variable power supply, the acquisition source of big data is less, and the error of a final result is larger due to the lack of training data, so that if a mobile terminal intelligent regulation and control method which can obtain a stable and accurate big data source and is suitable for a power grid system exists, the effectiveness of intelligent regulation and control can be greatly improved.
Disclosure of Invention
Aiming at the problems that in the technical scheme, under the condition that power is required to be supplied to a circuit configuration in a variable mode, the number of big data acquisition sources is small, and the error of the final result is large due to the lack of training data, the invention provides the mobile terminal intelligent regulation and control method which can obtain stable and accurate big data sources and is suitable for the power grid system.
The technical scheme adopted by the invention for solving the technical problems is as follows: an intelligent regulation and control method of a mobile terminal is matched with a server and the mobile terminal for use, wherein the server is in communication connection with the mobile terminal in an external network communication mode after passing through a safety scanning and login verification step, and comprises an initialization step, a fault data self-calculation step of collecting field data, a neural network training step of using a fault data self-calculation result and an intelligent fault prompting step of adopting a trained neural network;
the initialization step is as follows:
the method comprises the steps that a simulation model is built by a server according to field equipment, the server imports field equipment parameters according to setting and sets correlation among the field equipment, the server imports a plurality of manually set normal running states, power system load flow calculation is respectively executed according to the set normal running states, and the server obtains normal running state data of the field equipment through the load flow calculation; manually setting a plurality of faults and corresponding fault processing step constraint rules as default data for the field equipment in the server, communicating an initialization result to the mobile terminal by the server, displaying a field equipment mapping icon obtained synchronously by the mobile terminal, and associating corresponding operation options for the field equipment mapping icon;
the fault data self-calculation step:
step S1, the server sets field input parameters and field device faults randomly, then obtains simulated field device fault data through a power flow algorithm, obtains all feasible lines through a traversal method according to set constraint rules of fault processing steps, obtains a plurality of alternative lines through an evaluation algorithm, selects an alternative optimal line according to the sequence, and associates the field input parameters and the field device faults, the plurality of alternative lines and the alternative optimal line to form a detection set;
step S2, the server sends a detection set to the mobile terminal, and the user of the mobile terminal operates the field device mapping icon of the mobile terminal to form a user set circuit;
step S3, the server compares the user set circuit uploaded by the mobile terminal with a plurality of alternative circuits and alternative best circuits in the detection set, if the user set circuit belongs to the alternative circuit and is the same as the alternative best circuit, the alternative best circuit is set as the best circuit, and the current field input parameter and field equipment fault are associated with the best circuit to form an operation set and stored, if the user set circuit belongs to the alternative circuit and is different from the alternative best circuit, the alternative best circuit is issued for the user to refer to, the user selects or adjusts the circuit as the best circuit, and the best circuit selected or adjusted by the user is associated with the field input parameter and the current field equipment fault to form an operation set and stored, if the user set circuit does not belong to the alternative circuit, the simulation input parameter set by the user is uploaded to the server, the server re-simulates and sends the prompt to the mobile terminal, the user of the mobile terminal adjusts the line, and the step S3 is re-executed;
the neural network training step:
g1, establishing a multi-input multi-output neural network, selecting a plurality of optimal circuits with the most occurrence in all operation sets by the server, taking the optimal circuits as neural network training results, and taking field input parameters and field equipment faults corresponding to the optimal circuits as input variables of the neural network for neural network training;
g2, when the error of the neural network after each round of training is calculated, randomly set field input parameters and field equipment faults are generated by the established simulation model, if the optimal line obtained through calculation of the neural network is feasible in the simulation model and the evaluation score obtained through an evaluation algorithm is larger than a preset value, the current neural network is judged to correctly calculate the randomly set field input parameters and the field equipment faults, otherwise, the current neural network is judged to wrongly calculate the randomly set field input parameters and the field equipment faults; repeating the step G1 until the trained neural network meets the set requirement;
the intelligent fault prompting step comprises:
and T1, the server and the mobile terminal acquire field input parameters and equipment faults, the server obtains an optimal line by utilizing the calculation of the neural network according to the field input parameters and the equipment faults, the optimal line is confirmed through the load flow calculation of the simulation model, if the optimal line is confirmed, the optimal line is used as a fault intelligent prompt optimal line to be issued, the issued optimal line and the load flow calculation result of the simulation model are associated and issued to the mobile terminal, and the user determines the optimal line and the load flow calculation result to be used.
The invention adopts an ingenious mode to realize the source of big data, namely, the opportunity of personal ability test is fully utilized, the ability is trained and the parameters for neural network training are determined in a manual determination mode, and a large amount of available and relatively accurate data are obtained by the method, so that the mobile terminal intelligent regulation and control method which obtains stable and accurate big data source and is suitable for a power grid system is provided. In the invention, four steps are mainly adopted, the first step is an initialization step, a simulation model is established, the simulation model can generate a plurality of theoretically feasible regulation and control schemes, the mobile terminal obtains corresponding equipment icons which can be used for regulation and control according to the simulation model, then, a self-calculation step of fault data is carried out, namely, a plurality of field equipment fault conditions and field input parameters including power utilization environment, power supply object economy and the like can be generated, a feasible regulation and control scheme is theoretically realized through simulation model generation, the current fault and power utilization environment is issued, the feasible regulation and control scheme is collected and compared with the existing regulation and control scheme uploaded manually, the control scheme in the step is mainly embodied in a set which contains the optimal line and the operation of the user for forming the optimal line; the neural network training step is relatively simple, and only the optimal line is used as a neural network training result, the field input parameter and the field equipment fault corresponding to the optimal line are used as input variables of the neural network to carry out neural network training, and the feasibility of theoretical calculation is kept; when the neural network can meet the requirements after being trained, fault intelligent prompting can be carried out in a neural network assisted mode, and optional steps are provided for an actual regulator to confirm or refer. In the invention, load flow calculation and evaluation calculation both need large calculation power, so calculation needs to be carried out in a computer mode, the time spent for judging by a user according to specifications, experiences and the like is short after calculation, but the verification of the judgment result can be directly verified with a transmitted detection set, and the user can consciously judge whether the selection made by the user under the environmental condition is suitable or not in the comparison process of verification. If the test set is the same, the user's selection may be verified as correct. If the deviation occurs, the user can further correct according to the judgment of the detection set, so that the effect of obtaining the correct line can be achieved. The correct route here may be only one of the alternate routes because the route operability of the alternate route is deterministic, and if the correct route is not one of the alternate routes because there is a potential for inoperability, then more verification is required. In a word, the load flow calculation and the evaluation calculation performed by the traversal method need larger calculation force and longer calculation time, and part of the content is defined by a computer, so that a user can make faster selection by more depending on the conditions such as experience and the like; the correctness of the selection can be used for assisting the acquisition of the big data of the computer and evaluating and testing the dispatching level of the user. The data of the neural network training are checked and compared in the two aspects of artificial rechecking and computer traversal calculation, so that the accuracy can be ensured, meanwhile, the randomness is sufficient, the sufficient data volume is generated, and the neural network can be assisted to generate an accurate prediction auxiliary line.
Preferably, in step S1, when obtaining a plurality of candidate routes and an optimal candidate route by using an evaluation algorithm after obtaining all feasible routes by a traversal method according to a set constraint rule of the fault handling step, the function establishment of the evaluation algorithm includes the following steps,
the method comprises the following steps that firstly, a plurality of feasible lines and evaluation parameters corresponding to the feasible lines are manually selected, wherein the evaluation parameters of the feasible lines comprise line parameters obtained through load flow calculation simulation and field input parameters set manually;
the evaluation algorithm step two, the importance degree of the evaluation parameters relative to the optimal line selection is compared, and a comparison scale matrix of the importance degree between the evaluation parameters is established;
evaluating the algorithm step three, carrying out forward processing and standardization processing on the comparison scale matrix, and determining a weighting decision matrix used in evaluation;
and step four of evaluating the algorithm, performing weighted calculation on the weighted decision matrix and the evaluation parameters related to the feasible lines, obtaining the optimal ideal solution and the worst ideal solution of the feasible lines according to the result of the weighted calculation, recording the optimal ideal solution and the corresponding evaluation parameters, and recording the worst ideal solution and the corresponding evaluation parameters.
The evaluation parameters in the invention are related to the importance of the power supply target, the power supply line loss and the like, and the number of alternative regulation and control schemes generated after the simulation model is traversed can be reduced through the establishment of the evaluation system.
Preferably, the field input parameters comprise daily power generation adjustment, power supply scheduling plan, power limitation, power plant power adjustment, generator set starting or stopping instruction and load economic index; the field device faults include power grid frequency or voltage out-of-range, power transmission and transformation device load out-of-range, and trunk line power values as a percentage of specified stability limits.
Preferably, in step S1, the feasible lines to be evaluated and the evaluation parameters are read, the euclidean distance meter for calculating the evaluation parameters and recording the evaluation parameters of the optimal ideal solution is made dsg, the euclidean distance meter for calculating the evaluation parameters and recording the evaluation parameters of the worst ideal solution is made dsb, and the evaluation value ds:
ds=dsg/(dsg+dsb),0<ds；
and sequencing all the feasible lines in sequence from large to small according to the corresponding evaluation values ds, selecting a plurality of previous feasible lines from all the feasible lines according to the sequencing as a plurality of alternative lines, and selecting the feasible line with the highest ds value as the optimal alternative line.
Preferably, in the step G1, a mimo neural network is established, the server selects the best lines with the highest occurrence in all operation sets, the occurrence number of each best line selected by the server is greater than a manually set threshold, and the sum of the occurrence numbers of all best lines selected by the server is also greater than a manually set threshold. By this setting, an excessively small number of line selections can be excluded, and at the same time, since the occurrence frequency of all the optimal lines is also larger than a manually set threshold, it can be ensured that the finally obtained data amount is necessary for training the neural network.
Preferably, the generation of the field device fault is limited to a number of field devices that are manually set. I.e. manually set faults including field device faults in the grid-enabled faulty operating state.
Preferably, in the step S3, when the user adjusts the route as the optimal route, the adjusted route must belong to the candidate routes, and if the route adjusted by the user does not belong to the candidate routes, the simulation input parameters set by the user are uploaded to the server, the server re-simulates the route, gives a prompt, and sends the prompt to the mobile terminal, the user of the mobile terminal performs the route adjustment, and the step S3 is re-executed. In the invention, the user is required to manually confirm and simulate the circuit, so that the data sent into the neural network is correct and reasonable.
Preferably, in step S3, the route adjusted by the user does not belong to the candidate route and the user regards the set simulation input parameter as the optimal route, the server records the optimal route as the manual candidate route, and then repeatedly sends the corresponding detection set to be selected at this time to the plurality of mobile terminals, and if the number of times that the user-set route uploaded by the mobile terminals is the same as the manual candidate route is greater than the set value, the manual candidate route is set as the optimal route. Furthermore, different authorities are set in the mobile terminal, whether the manual best line to be selected is available or not is determined by adopting a weighting algorithm, and when the authority of a user is enough, the manual best line to be selected can be directly set as the best line, namely a leader approval mode.
Preferably, the method further comprises a device maintenance prompting step, wherein the server sets a convolution function according to the maintenance time and the service life function of the field device, calculates the probability of faults occurring when the field devices are idle in maintenance, and adjusts the field devices with the faults occurring at random according to the probability of the faults of the field devices when the field input parameters and the field devices are set at random. According to the maintenance frequency of the field equipment, the probability of random failure is further challenged, namely the probability of the equipment failure is relatively increased when the next maintenance date is approached, and the corresponding purpose can be achieved by directly selecting a proper convolution function in the calculation.
Preferably, a plurality of simulation models are stored in the server, and after the mobile terminal acquires the field device mapping icon of one simulation model in the server, the server issues a corresponding detection set for the mobile terminal.
Preferably, a plurality of simulation models are stored in the server, and after the mobile terminal acquires the field device mapping icon of one simulation model in the server, the server issues a corresponding detection set for the mobile terminal. The setting of the invention can issue the data of a plurality of simulation models, namely, various forms of power supply circuits and power supply environments are issued as the forms of test subjects, so that personnel participating in determining the data can select different lines, on one hand, the amount of a database is enlarged, on the other hand, the amount of users participating in determining the lines is enlarged, and therefore, enough and accurate data amount is formed for the neural network to use.
The substantial effects of the invention are as follows: the mobile terminal intelligent regulation and control method provided by the invention fully utilizes the opportunity of personal ability test, and determines parameters for neural network training while training the ability in a manual determination mode. In the invention, all the feasible lines are subjected to parameter evaluation instead of any feasible line directly selected, the accuracy and the practicability of the line are higher, the data trained by the neural network are checked and compared in two aspects of artificial recheck and computer traversal calculation, the accuracy can be ensured, and meanwhile, the randomness is enough, enough data volume is generated, and the neural network can be assisted to generate an accurate prediction auxiliary line.
Drawings
FIG. 1 is a schematic diagram of an initial relationship comparison in the present invention;
FIG. 2 is a schematic flow chart of the self-calculation steps of the fault data in the present invention;
FIG. 3 is a schematic flow chart of neural network training in the present invention;
fig. 4 is a schematic flow chart of the fault intelligent prompting step in the invention.
Detailed Description
The technical solution of the present invention will be further specifically described below by way of specific examples.
Example 1:
the intelligent regulation and control method of the mobile terminal is matched with a server and the mobile terminal for use, the server is in communication connection with the mobile terminal after being subjected to safety scanning and login verification, and the server is in communication connection with the mobile terminal in an external network communication mode.
The initialization step (see fig. 1):
the method comprises the steps that a simulation model is built by a server according to field equipment, the server imports field equipment parameters according to setting and sets correlation among the field equipment, the server imports a plurality of manually set normal running states, power system load flow calculation is respectively executed according to the set normal running states, and the server obtains normal running state data of the field equipment through the load flow calculation; manually setting a plurality of faults and corresponding fault processing step constraint rules as default data for the field equipment in the server, enabling the server to pass an initialization result to the mobile terminal, displaying a field equipment mapping icon obtained synchronously by the mobile terminal, and associating corresponding operation options for the field equipment mapping icon; the constraint rule of the fault processing step comprises a set fixed operation sequence, namely, only after the associated equipment executes the corresponding operation option, the operation option corresponding to the associated icon is opened. It is further possible, therefore, to treat several interrelated field devices as one unified large device, even if the so-called unified large device is completely separate in the field. Meanwhile, the operation options in this embodiment include replacing or repairing the equipment, and the set value is accumulated when the equipment is replaced or repaired or when the abnormal time in the line is evaluated as a parameter in the degree of importance of the subsequent evaluation parameter with respect to the optimal line selection.
The establishment of the simulation model is the prior art of the department, the normal operation state data of each field device obtained through the power system load flow calculation is mainly used for providing default data for reference for subsequent set faults, meanwhile, the reference is also established for fault processing rules, the server passes an initialization result to the mobile terminal, and the mobile terminal displays the field device mapping icon obtained synchronously. For example, if a common switch appears, the operable options include open and close, so that a connected state and an open state are formed, and the selection switch has more operation options and can select which connected path is, so that the field device mapping icon associated with the corresponding operation option should be manually set according to default data provided by the simulation model; the generation of field device faults is limited to a number of field devices that are set manually. That is, manually set faults include field device faults in the event of a fault-tolerant operating state of the power grid, i.e., the type of fault that typically allows remote control.
The failure data self-calculation step (see fig. 2):
step S1, the server sets field input parameters and field device faults randomly, then the load flow calculation method calculates the simulated field device fault data, all feasible lines are obtained through a traversal method according to the set constraint rules of the fault processing steps, a plurality of alternative lines and alternative optimal lines are obtained through an evaluation algorithm, and the field input parameters, the field device faults, the alternative lines and the alternative optimal lines are associated to form a detection set; furthermore, the several candidate lines and the best candidate line are recorded in time series with respect to the specific operations performed on the field device, and therefore, the display of the candidate lines and the best candidate line in the detection set must be restored and displayed by the initialization result of the mobile terminal.
Furthermore, in step S1, when obtaining a plurality of candidate routes and an optimal candidate route by using an evaluation algorithm after obtaining all feasible routes by a traversal method according to the set constraint rule of the fault handling step, the function establishment of the evaluation algorithm includes the following steps,
the method comprises the following steps that firstly, a plurality of feasible lines and evaluation parameters corresponding to the feasible lines are manually selected, wherein the evaluation parameters of the feasible lines comprise line parameters obtained through load flow calculation simulation and field input parameters set manually;
the evaluation algorithm step two, the importance degree of the evaluation parameters relative to the optimal line selection is compared, and a comparison scale matrix of the importance degree between the evaluation parameters is established;
evaluating the algorithm step three, carrying out forward processing and standardization processing on the comparison scale matrix, and determining a weighting decision matrix used in evaluation; firstly, performing forward processing on the matrix, wherein the data after the forward processing is larger the closer the numerical value is to the optimal value, otherwise, the data after the forward processing is smaller, the evaluation data can be of a data extremely large type, a data extremely small type and a data intermediate type, and the extremely small index is converted into an extremely large index: maximizing the data with max-x;
the intermediate index is converted into an extremely large index:
is a set of intermediate indexes, the optimal value of the set is
Then, the forward formulation is:
,
；
the interval index is converted into a very large index:
is a set of intermediate indexes, the optimal value interval of the set is
Then, the forward formulation is:
preferably, this embodiment recommends that the form of the contrast scale matrix is established by comparing the evaluation parameters two by two, and 1 represents that two factors have the same degree of importance compared with each other, and 9 represents that one factor is extremely important compared with the other. A has the same importance degree compared with B, A is 1 to B, B is 1 to A; the factor A is extremely important over the other factor C, and thus, the ratio of A to C is 9, and the ratio of C to A is 1/9.
And step four of evaluating the algorithm, performing weighted calculation on the weighted decision matrix and the evaluation parameters related to the feasible lines, obtaining the optimal ideal solution and the worst ideal solution of the feasible lines according to the result of the weighted calculation, recording the optimal ideal solution and the corresponding evaluation parameters, and recording the worst ideal solution and the corresponding evaluation parameters. In the weighted calculation of the parameters related to the feasible lines in the step, firstly, the parameters are classified and evaluated, for example, the power factors are classified according to the advantages and disadvantages of 90% or less, 90% -91%, 91% -92% and the like, the classified parameters form a metering value between 0 and 1 according to the advantages and disadvantages, 1 represents the optimal value, and 0 represents the worst value. And performing weighted calculation according to the metering value and the weighted decision matrix, performing weighted calculation by using the highest value appearing in all feasible lines when calculating the optimal solution, and performing weighted calculation by using the lowest value appearing in all feasible lines when calculating the worst solution, namely, firstly selecting the optimal metering value and the worst metering value in all parameters in actual operation, and respectively weighting the optimal metering value and the worst metering value to obtain the optimal and worst solution. The evaluation parameters in the embodiment are related to the importance of the power supply target, the power supply line loss and the like, and the number of the alternative regulation and control schemes generated after the simulation model is traversed can be reduced through the establishment of the evaluation system. The field input parameters comprise daily power generation adjustment, power supply scheduling plan, power limitation, power plant power adjustment, generator set starting or stopping instruction and load economic index; the field equipment faults comprise power grid frequency or voltage range exceeding, power transmission and transformation equipment load value exceeding range main line power value, specified stability limit percentage and the like.
More specifically, in step S1, the feasible line and the evaluation parameter to be evaluated are read, the evaluation parameter and the euclidean distance meter recording the evaluation parameter of the optimal ideal solution are calculated as dsg, the evaluation parameter and the euclidean distance meter recording the evaluation parameter of the worst ideal solution are calculated as dsb, and the evaluation value ds:
ds=dsg/(dsg+dsb),0<ds；
and sequencing all the feasible lines in sequence from large to small according to the corresponding evaluation values ds, selecting a plurality of previous feasible lines from all the feasible lines according to the sequencing as a plurality of alternative lines, and selecting the feasible line with the highest ds value as the optimal alternative line. In general, the measured value of the evaluation parameter may be selected for the euclidean distance calculation, or a specific numerical value may be selected for the euclidean distance calculation, but the present invention is not limited thereto, but in the present embodiment, it is recommended that the measured value of the evaluation parameter is used for the euclidean distance calculation, and for the index of the power factor, there is a significant difference between the euclidean distance represented by the index calculations of 90% and 95% of the power factors and the euclidean distance represented by the evaluations of the extreme evaluation 0.1 and the good evaluation 0.7.
Step S2, the server sends the primary detection set to the mobile terminal, the user of the mobile terminal operates the field device mapping icon of the mobile terminal to form a user set circuit,
step S3, the server compares the user set circuit uploaded by the mobile terminal with a plurality of alternative circuits and alternative best circuits in the detection set, if the user set circuit belongs to the alternative circuit and is the same as the alternative best circuit, the alternative best circuit is set as the best circuit, and the current field input parameter and field equipment fault are associated with the best circuit to form an operation set and stored, if the user set circuit belongs to the alternative circuit and is different from the alternative best circuit, the alternative best circuit is issued for the user to refer to, the circuit is selected or adjusted by the user as the best circuit, and the best circuit selected or adjusted by the user is associated with the field input parameter and the current field equipment fault to form an operation set and stored, if the user set circuit does not belong to the alternative circuit, the simulation input parameter set by the user is uploaded to the server, giving a prompt after re-simulation by the server, sending the prompt to the mobile terminal, performing line adjustment by a user of the mobile terminal, and re-executing the step S3;
in step S3, when the user adjusts the line as the best line, the adjusted line must belong to the alternative line, and if the line adjusted by the user does not belong to the alternative line, the simulation input parameter set by the user is uploaded to the server, and the server re-simulates the line and then sends the simulation input parameter to the mobile terminal, and the user of the mobile terminal performs the line adjustment and re-executes step S3. In this embodiment, the user is required to perform manual validation and simulation validation on the line, so as to ensure that the data sent to the neural network is correct and reasonable.
The neural network training step (see fig. 3):
g1, establishing a multi-input multi-output neural network, selecting a plurality of optimal circuits with the most occurrence in all operation sets by the server, taking the optimal circuits as neural network training results, and taking field input parameters and field equipment faults corresponding to the optimal circuits as input variables of the neural network for neural network training; in the step G1, a mimo neural network is established, the server selects a plurality of best lines with the highest occurrence in all operation sets, the occurrence number of each best line selected by the server is greater than a manually set threshold, and the sum of the occurrence numbers of all best lines selected by the server is also greater than a manually set threshold. By this setting, an excessively small number of line selections can be excluded, and at the same time, since the occurrence frequency of all the optimal lines is also larger than a manually set threshold, it can be ensured that the finally obtained data amount is necessary for training the neural network. G1, establishing a multi-input multi-output neural network, selecting a plurality of optimal circuits with the most occurrence in all operation sets by the server, taking the optimal circuits as neural network training results, and taking field input parameters and field equipment faults corresponding to the optimal circuits as input variables of the neural network for neural network training; in the step G1, a mimo neural network is established, the server selects a plurality of best lines with the highest occurrence in all operation sets, the occurrence number of each best line selected by the server is greater than a manually set threshold, and the sum of the occurrence numbers of all best lines selected by the server is also greater than a manually set threshold. By this setting, an excessively small number of line selections can be excluded, and at the same time, since the occurrence frequency of all the optimal lines is also larger than a manually set threshold, it can be ensured that the finally obtained data amount is necessary for training the neural network. In this embodiment, the mimo neural network is a common neural network and may have a plurality of neural network layers, in the mimo neural network in this embodiment, the input quantities are field input parameters and field device faults, and the output result is an optimal line, the neural network in this embodiment may be considered as a multi-input classifier, for example, an input field input parameter (set) of a1 and a field device fault (set) of a1 corresponds to an optimal line a, an input field input parameter (set) of a2 and a field device fault (set) of a3 still correspond to an optimal line a, an input field input parameter (set) of B4 and a field device fault (set) of B5 correspond to an optimal line B, and so on. The goal is to achieve a known or unknown input field input parameter (set) xn and a known or unknown field device fault (set) xn, to obtain an optimal line, which is one of the known optimal lines. The establishment method and method of the mimo neural network are not important, and for example, how to determine and quickly adjust the weighted data belongs to the prior art, and details are not described in this application.
G2, when the error of the neural network after each round of training is calculated, randomly set field input parameters and field equipment faults are generated by the established simulation model, if the optimal line obtained through calculation of the neural network is feasible in the simulation model and the evaluation score obtained through an evaluation algorithm is larger than a preset value, the current neural network is judged to correctly calculate the randomly set field input parameters and the field equipment faults, otherwise, the current neural network is judged to wrongly calculate the randomly set field input parameters and the field equipment faults; repeating the step G1 until the trained neural network meets the set requirement;
the intelligent fault prompting step (see the attached figure 4):
and T1, the server and the mobile terminal acquire field input parameters and equipment faults, the server obtains an optimal line by utilizing the calculation of the neural network according to the field input parameters and the equipment faults, the optimal line is confirmed through the load flow calculation of the simulation model, if the optimal line is confirmed, the optimal line is used as a fault intelligent prompt optimal line to be issued, the issued optimal line and the load flow calculation result of the simulation model are associated and issued to the mobile terminal, and the user determines the optimal line and the load flow calculation result to be used. The determined use is not limited to the scheme that the user must adopt, the user can still adjust or select the supplied scheme, and the adjusted and selected scheme is not necessarily directly applicable and can be continuously applicable after being checked or approved by a higher-authority user.
In this embodiment, the optimal route is confirmed through load flow calculation of the simulation model, which is a simple review step, that is, the field input parameters and the field device faults are input to operate corresponding to the selected optimal route, and the operated data must be allowed and permitted by the specification. The rechecking method is to obtain the circuit operation parameter by load flow calculation, and the operation parameter is compared with the rated data and belongs to the rated data, so that the operation parameter can be determined. In general, the possibility of a problem in this step is not high due to the rationality and redundancy of the best line itself, but a large deviation is still possible due to the large variation of the input parameters, and therefore, it is still necessary to set this review step.
The server stores a plurality of simulation models, and after the mobile terminal acquires a field device mapping icon of one simulation model in the server, the server issues a corresponding detection set for the mobile terminal. The server stores a plurality of simulation models, and after the mobile terminal acquires a field device mapping icon of one simulation model in the server, the server issues a corresponding detection set for the mobile terminal. The setting of this embodiment can be used to issue data of multiple simulation models, that is, various forms of power supply circuits and power supply environments are issued as the form of test questions, so that the personnel participating in determining data can select from various lines, thereby expanding the amount of the database on one hand, and expanding the amount of users participating in determining lines on the other hand, and forming a sufficient and accurate data amount for the neural network to use. Namely, the problem that the dispatcher A at the A can do includes not only the problem of the self line at the A, but also the problem of the line at B, C, D, E, and the corresponding problem of the self line at the A is not only familiar to the dispatcher A at the A, and can also be issued to the dispatcher B at the B, the dispatcher C and the like, and certainly, in the process of collecting big data, the dispatchers A, B, C and the like are in the same authority, and in the process of actual fault handling regulation and control, the dispatcher A at the A can only regulate the line at the A, and the setting is limited by the identity authority of the mobile terminal.
As a further optimization, the method also comprises a device maintenance prompting step, wherein the server sets a convolution function according to the maintenance time and the service life function of the field device, calculates the probability of faults occurring when the field devices are not maintained, and adjusts the field devices which randomly have faults according to the probability of the faults of the field devices when field input parameters and the field devices are randomly set. In this embodiment, the probability of random failure is further challenged according to the maintenance frequency of the field device, that is, the probability of failure of the field device is relatively increased as the maintenance date is closer to the next time, and the calculation directly selects a proper convolution function to achieve the corresponding purpose.
The embodiment fully utilizes the opportunity of personal ability test, trains the ability and determines the parameters for neural network training at the same time in a manual determination mode, and a large amount of available relatively accurate data is obtained by the method, so that the mobile terminal intelligent regulation and control method suitable for the power grid system and capable of obtaining stable and accurate large data sources is provided. In the embodiment, there are mainly four steps, firstly, the initialization step is performed to establish a simulation model, the simulation model can generate a plurality of theoretically feasible regulation and control schemes, the mobile terminal obtains a corresponding device icon for regulation and control according to the simulation model, the self-calculation step of fault data is then executed to generate a plurality of field device fault conditions and field input parameters including power utilization environment, power supply object economy and the like, a feasible regulation and control scheme is theoretically realized through a simulation model, the current fault and power utilization environment is issued, the feasible regulation and control scheme is collected and compared with the existing regulation and control scheme uploaded manually, the control scheme in the step is mainly embodied in a set which contains the optimal line and the operation of the user for forming the optimal line; the neural network training step is relatively simple, and only the optimal line is used as a neural network training result, the field input parameter and the field equipment fault corresponding to the optimal line are used as input variables of the neural network to carry out neural network training, and the feasibility of theoretical calculation is kept; when the neural network can meet the requirements after being trained, fault intelligent prompting can be carried out in a neural network assisted mode, and optional steps are provided for an actual regulating person to confirm or refer.
Example 2:
the embodiment is basically the same as embodiment 1, except that in this embodiment, the mobile terminal may further include a security verification deployment unit and an application unit, the security verification deployment unit includes a security scanning module, a login verification module, an extranet service management module, and an account management module, the application unit includes a fault monitoring module, a voice recognition understanding module, and a learning training module of a target field device, and the learning training module includes a power grid technology training module, a mobile examination module, a special training module, a learning material management module, and a training library.
The safety verification deployment unit performs data transmission interaction through a regulation cloud platform and a communication safety transmission platform, the safety scanning module is used for scanning a login environment of a user, the login verification module is used for verifying login information of the user, the extranet service management module is used for managing a network of the safety verification deployment unit, the account management module is used for setting identities of management personnel and operators, and the application issuing configuration module is used for setting application configuration parameters and synchronizing with a server side.
The fault monitoring module is mainly used for receiving fault information to compile, look up and process, and by receiving the information, a user clicks the operation option of the equipment to execute the remote operation of the equipment according to the field equipment mapping icon synchronously obtained by the display of the mobile terminal.
The power grid technology training module is used for automatically setting out questions, answering questions and marking papers according to the work requirements of monitoring staff and giving comprehensive judgment, wherein in the method, a plurality of alternative lines and an alternative optimal line are obtained by using an evaluation algorithm, and a field input parameter, a field equipment fault, a plurality of alternative lines and an alternative optimal line are associated to form a detection set, namely the detection set is issued to users with corresponding authorities in an automatic question setting mode. Further, in step S3, if the line adjusted by the user does not belong to the alternative line and the user considers that the set simulation input parameter should be the optimal line, the server records the optimal line to be manually selected, and then repeatedly sends the corresponding detection set to be manually selected to the plurality of mobile terminals, and if the number of times that the user-set line uploaded by the mobile terminals is the same as the number of times that the optimal line to be manually selected is greater than the set value, the optimal line to be manually selected is set as the optimal line. Furthermore, different authorities are set in the mobile terminal, whether the manual best line to be selected is available or not is determined by adopting a weighting algorithm, and when the authority of a user is enough, the manual best line to be selected can be directly set as the best line, namely a leadership approval mode. In summary, the specific method in this embodiment is hidden in the function of providing the simulated examination and the special training according to the role requirement, and the simulation model at the server side, which is also called as a power grid accident simulation module, is used to perform simulation preview on an accident that may occur in a power grid, perform simulation training on a user, evaluate according to the simulation training, predict the existing risk potential of the power grid for the historical operation condition of the equipment, the existing defect condition, the load trend condition and the wiring mode of the power grid framework, propose a monitoring key point, and perform regular pushing training. As another way for forming a detection set to be issued, the special training module is used for a monitor to simulate actual operation, has a monitoring personnel system using method training function, and can record a dispatcher training record and check historical questions.
In the above embodiment, the data of the neural network training is subjected to check comparison in two aspects of artificial rechecking and computer traversal calculation, so that the accuracy can be ensured, and meanwhile, the sufficient randomness is provided, the sufficient data volume is generated, and the neural network can be assisted to generate an accurate prediction auxiliary line.
The above-described embodiment is only a preferred embodiment of the present invention, and is not intended to limit the present invention in any way, and other variations and modifications may be made without departing from the spirit and scope of the invention as set forth in the appended claims.
Claims (10)
1. An intelligent regulation and control method of a mobile terminal is characterized in that the intelligent regulation and control method is matched with a server and the mobile terminal for use, the server is in communication connection with the mobile terminal through an external network communication mode after passing through safety scanning and login verification steps, and comprises an initialization step, a fault data self-calculation step carried out by collecting field data, a neural network training step carried out by using a fault data self-calculation result and a fault intelligent prompting step carried out by adopting a trained neural network;
the initialization step is as follows:
the method comprises the steps that a simulation model is built by a server according to field equipment, the server imports field equipment parameters according to setting and sets correlation among the field equipment, the server imports a plurality of manually set normal running states, power system load flow calculation is respectively executed according to the set normal running states, and the server obtains normal running state data of the field equipment through the load flow calculation; manually setting a plurality of faults and corresponding fault processing step constraint rules as default data for the field equipment in the server, communicating an initialization result to the mobile terminal by the server, displaying a field equipment mapping icon obtained synchronously by the mobile terminal, and associating corresponding operation options for the field equipment mapping icon;
the fault data self-calculation step:
step S1, the server sets field input parameters and field device faults randomly, then obtains simulated field device fault data through a power flow algorithm, obtains all feasible lines through a traversal method according to set constraint rules of fault processing steps, obtains a plurality of alternative lines through an evaluation algorithm, selects an alternative optimal line according to the sequence, and associates the field input parameters and the field device faults, the plurality of alternative lines and the alternative optimal line to form a detection set;
step S2, the server sends a detection set to the mobile terminal, and the user of the mobile terminal operates the field device mapping icon of the mobile terminal to form a user set circuit;
step S3, the server compares the user set circuit uploaded by the mobile terminal with a plurality of alternative circuits and alternative best circuits in the detection set, if the user set circuit belongs to the alternative circuit and is the same as the alternative best circuit, the alternative best circuit is set as the best circuit, and the current field input parameter and field equipment fault are associated with the best circuit to form an operation set and stored, if the user set circuit belongs to the alternative circuit and is different from the alternative best circuit, the alternative best circuit is issued for the user to refer to, the user selects or adjusts the circuit as the best circuit, and the best circuit selected or adjusted by the user is associated with the field input parameter and the current field equipment fault to form an operation set and stored, if the user set circuit does not belong to the alternative circuit, the simulation input parameter set by the user is uploaded to the server, the server re-simulates and sends the prompt to the mobile terminal, the user of the mobile terminal adjusts the line, and the step S3 is re-executed;
the neural network training step:
g1, establishing a multi-input multi-output neural network, selecting a plurality of optimal circuits with the most occurrence in all operation sets by the server, taking the optimal circuits as neural network training results, and taking field input parameters and field equipment faults corresponding to the optimal circuits as input variables of the neural network for neural network training;
g2, when the error of the neural network after each round of training is calculated, randomly set field input parameters and field equipment faults are generated by the established simulation model, if the optimal line obtained through calculation of the neural network is feasible in the simulation model and the evaluation score obtained through an evaluation algorithm is larger than a preset value, the current neural network is judged to correctly calculate the randomly set field input parameters and the field equipment faults, otherwise, the current neural network is judged to wrongly calculate the randomly set field input parameters and the field equipment faults; repeating the step G1 until the trained neural network meets the set requirement;
the intelligent fault prompting step comprises:
and T1, the server and the mobile terminal acquire field input parameters and equipment faults, the server obtains an optimal line by utilizing the calculation of the neural network according to the field input parameters and the equipment faults, the optimal line is confirmed through the load flow calculation of the simulation model, if the optimal line is confirmed, the optimal line is used as a fault intelligent prompt optimal line to be issued, the issued optimal line and the load flow calculation result of the simulation model are associated and issued to the mobile terminal, and the user determines the optimal line and the load flow calculation result to be used.
2. The intelligent regulation and control method of a mobile terminal according to claim 1, wherein in the step S1, after all feasible routes are obtained through a traversal method according to the set constraint rule of the fault handling step, when a plurality of candidate routes and the candidate optimal route are obtained by using an evaluation algorithm, the function establishment of the evaluation algorithm comprises the following steps,
the method comprises the following steps that firstly, a plurality of feasible lines and evaluation parameters corresponding to the feasible lines are manually selected, wherein the evaluation parameters of the feasible lines comprise line parameters obtained through load flow calculation simulation and field input parameters set manually;
the evaluation algorithm step two, the importance degree of the evaluation parameters relative to the optimal line selection is compared, and a comparison scale matrix of the importance degree between the evaluation parameters is established;
evaluating the algorithm step three, carrying out forward processing and standardization processing on the comparison scale matrix, and determining a weighting decision matrix used in evaluation;
and step four of evaluating the algorithm, performing weighted calculation on the weighted decision matrix and the evaluation parameters related to the feasible lines, obtaining the optimal ideal solution and the worst ideal solution of the feasible lines according to the result of the weighted calculation, recording the optimal ideal solution and the corresponding evaluation parameters, and recording the worst ideal solution and the corresponding evaluation parameters.
3. The intelligent regulation and control method of a mobile terminal according to claim 2, wherein in step S1, the feasible lines to be evaluated and the evaluation parameters are read, the euclidean distance meter for calculating the evaluation parameters and recording the evaluation parameters of the optimal ideal solution is made dsg, the euclidean distance meter for calculating the evaluation parameters and recording the evaluation parameters of the worst ideal solution is made dsb, and the evaluation values ds:
ds=dsg/(dsg+dsb),0<ds；
and sequencing all the feasible lines in sequence from large to small according to the corresponding evaluation values ds, selecting a plurality of previous feasible lines from all the feasible lines according to the sequencing as a plurality of alternative lines, and selecting the feasible line with the highest ds value as the optimal alternative line.
4. The intelligent regulation and control method for the mobile terminal according to claim 3, wherein the field input parameters comprise daily power generation regulation, power supply scheduling plan, power limitation, power plant power regulation, generator set starting or stopping instruction and load economic index; the field device faults include power grid frequency or voltage out-of-range, power transmission and transformation device load out-of-range, and trunk line power values as a percentage of specified stability limits.
5. The intelligent regulation and control method of a mobile terminal as claimed in claim 1, 2 or 3, wherein in the step G1, a multi-input multi-output neural network is established, the server selects the best lines with the highest occurrence in all operation sets, the occurrence number of each best line selected by the server is greater than a manually set threshold, and the sum of the occurrence numbers of all best lines selected by the server is also greater than a manually set threshold.
6. The intelligent regulation and control method for the mobile terminal according to claim 5, wherein the generation of the field device fault is limited to a plurality of manually set field devices.
7. The method as claimed in claim 1, wherein in step S3, when the user adjusts the route as the best route, the adjusted route must belong to the alternative route, if the route adjusted by the user does not belong to the alternative route, the simulation input parameters set by the user are uploaded to the server, the server re-simulates the simulation input parameters and sends the simulation input parameters to the mobile terminal, the user of the mobile terminal performs the route adjustment, and step S3 is re-executed.
8. The method as claimed in claim 3, wherein in step S3, the route adjusted by the user does not belong to the alternative route and the user considers that the set simulation input parameter should be the best route, the server records the route as the best route to be manually selected, and then repeatedly sends the corresponding detection set to be manually selected to a plurality of mobile terminals, and if the number of times that the route set by the user uploaded by the mobile terminals is the same as the best route to be manually selected is greater than a set value, the best route to be manually selected is set as the best route.
9. The intelligent regulation and control method of the mobile terminal according to claim 8, further comprising a device maintenance prompting step, wherein the server sets a convolution function according to the maintenance time and the life function of the field device, calculates the probability of failure occurring when each field device is idle in maintenance, and adjusts the field device which has a failure at random according to the probability of the failure of the field device when the field input parameters and the failure of the field device are set at random.
10. The intelligent regulation and control method of the mobile terminal according to claim 8 or 9, wherein a plurality of simulation models are stored in the server, and after the mobile terminal obtains the field device mapping icon of one simulation model in the server, the server issues a corresponding detection set for the mobile terminal.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110628287.2A CN113077049B (en) | 2021-06-04 | 2021-06-04 | Intelligent regulation and control method for mobile terminal |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110628287.2A CN113077049B (en) | 2021-06-04 | 2021-06-04 | Intelligent regulation and control method for mobile terminal |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113077049A CN113077049A (en) | 2021-07-06 |
| CN113077049B true CN113077049B (en) | 2021-08-03 |
Family
ID=76617151
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110628287.2A Active CN113077049B (en) | 2021-06-04 | 2021-06-04 | Intelligent regulation and control method for mobile terminal |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113077049B (en) |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US11330004B1 (en) * | 2018-03-07 | 2022-05-10 | Amdocs Development Limited | System, method, and computer program for mitigating falsified log data provided to an AI-learning system |
| CN104269934B (en) * | 2014-10-23 | 2016-04-20 | 国家电网公司 | Based on the supervisory control of substation information adjustment method of moving analog scheduling station |
| CN109713794B (en) * | 2018-12-28 | 2020-12-29 | 广东电网有限责任公司 | Distributed intelligent self-recovery system and method |
| CN110021929B (en) * | 2018-12-29 | 2023-01-13 | 国网内蒙古东部电力有限公司经济技术研究院 | Electromagnetic transient time domain simulation modeling method for fast switch type fault current limiter |
| CN110334740A (en) * | 2019-06-05 | 2019-10-15 | 武汉大学 | The electrical equipment fault of artificial intelligence reasoning fusion detects localization method |
| CN111046581B (en) * | 2019-12-27 | 2022-10-04 | 国网江苏省电力有限公司电力科学研究院 | Power transmission line fault type identification method and system |
-
2021
- 2021-06-04 CN CN202110628287.2A patent/CN113077049B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN113077049A (en) | 2021-07-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107358366B (en) | Distribution transformer fault risk monitoring method and system | |
| Huang et al. | Dissolved gas analysis of mineral oil for power transformer fault diagnosis using fuzzy logic | |
| CN110929918B (en) | 10kV feeder fault prediction method based on CNN and LightGBM | |
| US10915835B2 (en) | Systems and methods for maximizing expected utility of signal injection test patterns in utility grids | |
| Sundaramurthi et al. | Human reliability modeling for the next generation system code | |
| WO2016144587A1 (en) | Cascaded identification in building automation | |
| CN105930861A (en) | Adaboost algorithm based transformer fault diagnosis method | |
| CN102928720A (en) | Defect rate detecting method of oil immersed type main transformer | |
| CN109800995A (en) | A kind of grid equipment fault recognition method and system | |
| CN113189451A (en) | Power distribution network fault positioning studying and judging method, system, computer equipment and storage medium | |
| CN112906764B (en) | Communication safety equipment intelligent diagnosis method and system based on improved BP neural network | |
| CN106779066A (en) | A kind of radar circuit plate method for diagnosing faults | |
| CN110826228A (en) | Regional power grid operation quality limit evaluation method | |
| CN105842607A (en) | Test point quantitative selection method and device in testing design | |
| Wang et al. | Design and implementation of early warning system based on educational big data | |
| Eltyshev et al. | Intelligent decision support in the electrical equipment diagnostics | |
| WO2019140553A1 (en) | Method and device for determining health index of power distribution system and computer storage medium | |
| CN110880055A (en) | Building intelligent ammeter system | |
| CN109255389A (en) | A kind of equipment evaluation method, device, equipment and readable storage medium storing program for executing | |
| CN113077049B (en) | Intelligent regulation and control method for mobile terminal | |
| Gu et al. | Investigation on the quality assurance procedure and evaluation methodology of machine learning building energy model systems | |
| CN111553581B (en) | Equipment maintainability evaluation model based on entropy value | |
| CN117394529A (en) | SCADA-based auxiliary decision method and system for main distribution network loop-closing reverse power supply control conditions | |
| CN112286088A (en) | Method and application system for online application of power equipment fault prediction model | |
| Aizpurua et al. | Determining appropriate data analytics for transformer health monitoring |
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 |