CN112365024A - High-voltage direct-current converter station energy efficiency prediction method and system based on deep learning - Google Patents
High-voltage direct-current converter station energy efficiency prediction method and system based on deep learning Download PDFInfo
- Publication number
- CN112365024A CN112365024A CN202011079041.6A CN202011079041A CN112365024A CN 112365024 A CN112365024 A CN 112365024A CN 202011079041 A CN202011079041 A CN 202011079041A CN 112365024 A CN112365024 A CN 112365024A
- Authority
- CN
- China
- Prior art keywords
- converter station
- neural network
- sample data
- network model
- energy efficiency
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 68
- 238000013135 deep learning Methods 0.000 title claims abstract description 28
- 238000003062 neural network model Methods 0.000 claims abstract description 147
- 238000009826 distribution Methods 0.000 claims abstract description 71
- 238000012549 training Methods 0.000 claims abstract description 46
- 238000013528 artificial neural network Methods 0.000 claims abstract description 39
- 238000007781 pre-processing Methods 0.000 claims abstract description 15
- 230000006870 function Effects 0.000 claims description 45
- 230000004913 activation Effects 0.000 claims description 34
- 238000009499 grossing Methods 0.000 claims description 22
- 230000008569 process Effects 0.000 claims description 17
- 230000005540 biological transmission Effects 0.000 claims description 14
- 238000004422 calculation algorithm Methods 0.000 claims description 11
- 238000012795 verification Methods 0.000 claims description 11
- 238000013459 approach Methods 0.000 claims description 7
- 238000010791 quenching Methods 0.000 claims description 7
- 238000010586 diagram Methods 0.000 description 12
- 238000012545 processing Methods 0.000 description 8
- 238000004590 computer program Methods 0.000 description 7
- 238000005516 engineering process Methods 0.000 description 5
- 238000005457 optimization Methods 0.000 description 5
- 238000012360 testing method Methods 0.000 description 5
- 238000011156 evaluation Methods 0.000 description 4
- 238000005259 measurement Methods 0.000 description 4
- 238000011160 research Methods 0.000 description 4
- 238000003860 storage Methods 0.000 description 3
- 230000008033 biological extinction Effects 0.000 description 2
- 210000004556 brain Anatomy 0.000 description 2
- 238000010276 construction Methods 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 238000007430 reference method Methods 0.000 description 2
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000002790 cross-validation Methods 0.000 description 1
- 238000005265 energy consumption Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000008450 motivation Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000008520 organization Effects 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 230000000171 quenching effect Effects 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/06—Electricity, gas or water supply
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
- Y02E40/00—Technologies for an efficient electrical power generation, transmission or distribution
- Y02E40/70—Smart grids as climate change mitigation technology in the energy generation sector
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Abstract
The invention discloses a deep learning-based energy efficiency prediction method and system for a high-voltage direct current converter station, which comprises the following steps: acquiring an energy efficiency distribution sample data set of the high-voltage direct-current converter station, and preprocessing the energy efficiency distribution sample data set to acquire a preprocessed energy efficiency distribution sample data set; respectively constructing a converter station total loss neural network model and/or a converter station loss ratio neural network model; respectively training and optimizing the converter station total loss neural network model and/or the converter station loss ratio neural network model based on the preprocessed energy efficiency distribution sample data set so as to determine a converter station total loss neural network optimal model and a converter station loss ratio neural network optimal model; and predicting the energy efficiency of the high-voltage direct-current converter station to be tested by utilizing the converter station total loss neural network optimal model and/or the converter station loss ratio neural network optimal model according to the operation data of the high-voltage direct-current converter station to be tested so as to obtain an energy efficiency prediction result.
Description
Technical Field
The invention relates to the technical field of electric energy metering, in particular to a deep learning-based energy efficiency prediction method and system for a high-voltage direct-current converter station.
Background
In a high voltage direct current transmission system, due to the non-linear load characteristic of a converter, it is generally difficult to directly measure the loss of main components of a converter station. Instead the total loss of the converter station is obtained by theoretical calculation of the losses of the individual components in the station separately and adding them. This method is highly practical, but its accuracy is sometimes not satisfactory.
Deep learning is a new field in machine learning research, and its motivation is to create and simulate a neural network for human brain to analyze and learn, which simulates the mechanism of human brain to interpret data such as images, sounds and texts. Deep learning is one type of unsupervised learning. The concept of deep learning stems from the study of artificial neural networks. A multi-layer perceptron with multiple hidden layers is a deep learning structure. Deep learning forms more abstract high-level representation attribute classes or features by combining low-level features to find a distributed feature representation of data that utilizes spatial relative relationships to reduce the number of parameters to improve training performance.
Therefore, a method for predicting the energy efficiency of the high-voltage direct-current converter station based on deep learning is needed.
Disclosure of Invention
The invention provides a deep learning-based energy efficiency prediction method and system for a high-voltage direct-current converter station, and aims to solve the problem of how to efficiently and accurately predict the energy efficiency of a high-voltage branch converter station.
In order to solve the above problem, according to an aspect of the present invention, there is provided a deep learning-based energy efficiency prediction method for a high voltage direct current converter station, the method including:
the method comprises the steps of obtaining an energy efficiency distribution sample data set of the high-voltage direct-current converter station, and preprocessing the energy efficiency distribution sample data set to obtain a preprocessed energy efficiency distribution sample data set;
respectively constructing a converter station total loss neural network model and/or a converter station loss ratio neural network model;
training and optimizing the converter station total loss neural network model and/or the converter station loss ratio neural network model respectively based on the preprocessed energy efficiency distribution sample data set so as to determine a converter station total loss neural network optimal model and a converter station loss ratio neural network optimal model;
and predicting the energy efficiency of the high-voltage direct-current converter station to be tested by utilizing the converter station total loss neural network optimal model and/or the converter station loss ratio neural network optimal model according to the operation data of the high-voltage direct-current converter station to be tested so as to obtain an energy efficiency prediction result.
Preferably, the energy efficiency distribution sample dataset comprises: an input sample data subset and an output sample data subset of the neural network model;
the variables in the subset of output sample data include: the total loss of the converter station and the loss of the converter, the converter transformer and the smoothing reactor account for the ratio;
when the type of the converter station is a rectifier station, the variables in the subset of input sample data include: the method comprises the following steps of (1) alternating-current end voltage, alternating-current end current, alternating-current end power factor, direct-current end voltage, direct-current end current, direct-current transmission power, a trigger angle and a commutation overlap angle of a rectifier;
when the type of the converter station is an inverter station, the variables in the input sample data subset include: ac terminal voltage, ac terminal current, ac terminal power factor, dc terminal voltage, dc terminal current, dc delivered power, inverter lead angle, arc-quenching angle, and commutation overlap angle.
Preferably, the preprocessing the energy efficiency distribution sample data set to obtain a preprocessed energy efficiency distribution sample data set includes:
calculating the average value and the standard deviation of each input variable according to the input sample data in the energy efficiency distribution sample data set;
for any input sample data, calculating the difference value between each input variable in the input sample data and the corresponding average value, calculating the ratio of the difference value to the corresponding standard deviation, and taking the ratio as the processed input sample data corresponding to the input sample data;
and determining the preprocessed energy efficiency distribution sample data set according to the processed data of each input sample and the output sample data corresponding to the processed data.
Preferably, the converter station total loss neural network model is a sequence structure and comprises 5 hidden layers, each layer comprises 48 units, and an activation function is arranged between the layers; the number of units is determined by the input layer according to the type of the converter station; the output layer is 1 unit, an activation function is not used, and the output value is the total loss of the converter station;
the converter station loss ratio neural network model is of a sequence structure and comprises 7 hidden layers, 64 units are arranged on each layer, and an activation function is arranged between the layers; the number of units is determined by the input layer according to the type of the converter station; the output layer has 3 units, and does not use an activation function, and the output values are loss ratios of the current converter, the converter transformer and the smoothing reactor respectively.
Preferably, wherein the method further comprises:
in the training process of the converter station total loss neural network model and/or the converter station loss ratio neural network model, adjusting weight parameters of internal units of the neural network model by using a gradient descent algorithm so as to enable the output result of the neural network model to approach to a target value;
and after the converter station total loss neural network model and/or the converter station loss ratio neural network model are trained, verifying the accuracy of the neural network model prediction by using a cross-folding cross verification method.
According to another aspect of the invention, a deep learning-based energy efficiency prediction system for a high-voltage direct current converter station is provided, and the system comprises:
the sample data preprocessing unit is used for acquiring an energy efficiency distribution sample data set of the high-voltage direct-current converter station and preprocessing the energy efficiency distribution sample data set to acquire a preprocessed energy efficiency distribution sample data set;
the neural network model building unit is used for respectively building a converter station total loss neural network model and/or a converter station loss ratio neural network model;
the neural network model training unit is used for respectively training and optimizing the converter station total loss neural network model and/or the converter station loss ratio neural network model based on the preprocessed energy efficiency distribution sample data set so as to determine a converter station total loss neural network optimal model and a converter station loss ratio neural network optimal model;
and the energy efficiency prediction unit is used for predicting the energy efficiency of the high-voltage direct-current converter station to be tested by utilizing the converter station total loss neural network optimal model and/or the converter station loss ratio neural network optimal model according to the operation data of the high-voltage direct-current converter station to be tested so as to obtain an energy efficiency prediction result.
Preferably, the energy efficiency distribution sample dataset comprises: an input sample data subset and an output sample data subset of the neural network model;
the variables in the subset of output sample data include: the total loss of the converter station and the loss of the converter, the converter transformer and the smoothing reactor account for the ratio;
when the type of the converter station is a rectifier station, the variables in the subset of input sample data include: the method comprises the following steps of (1) alternating-current end voltage, alternating-current end current, alternating-current end power factor, direct-current end voltage, direct-current end current, direct-current transmission power, a trigger angle and a commutation overlap angle of a rectifier;
when the type of the converter station is an inverter station, the variables in the input sample data subset include: ac terminal voltage, ac terminal current, ac terminal power factor, dc terminal voltage, dc terminal current, dc delivered power, inverter lead angle, arc-quenching angle, and commutation overlap angle.
Preferably, the sample data preprocessing unit is configured to preprocess the energy efficiency distribution sample data set to obtain a preprocessed energy efficiency distribution sample data set, and includes:
calculating the average value and the standard deviation of each input variable according to the input sample data in the energy efficiency distribution sample data set;
for any input sample data, calculating the difference value between each input variable in the input sample data and the corresponding average value, calculating the ratio of the difference value to the corresponding standard deviation, and taking the ratio as the processed input sample data corresponding to the input sample data;
and determining the preprocessed energy efficiency distribution sample data set according to the processed data of each input sample and the output sample data corresponding to the processed data.
Preferably, the converter station total loss neural network model is a sequence structure and comprises 5 hidden layers, each layer comprises 48 units, and an activation function is arranged between the layers; the number of units is determined by the input layer according to the type of the converter station; the output layer is 1 unit, an activation function is not used, and the output value is the total loss of the converter station;
the converter station loss ratio neural network model is of a sequence structure and comprises 7 hidden layers, 64 units are arranged on each layer, and an activation function is arranged between the layers; the number of units is determined by the input layer according to the type of the converter station; the output layer has 3 units, and does not use an activation function, and the output values are loss ratios of the current converter, the converter transformer and the smoothing reactor respectively.
Preferably, the neural network model training unit further includes:
in the training process of the converter station total loss neural network model and/or the converter station loss ratio neural network model, adjusting weight parameters of internal units of the neural network model by using a gradient descent algorithm so as to enable the output result of the neural network model to approach to a target value;
and after the converter station total loss neural network model and/or the converter station loss ratio neural network model are trained, verifying the accuracy of the neural network model prediction by using a cross-folding cross verification method.
The invention provides a high-voltage direct-current converter station energy efficiency prediction method and system based on deep learning, wherein a converter station total loss neural network model and/or a converter station loss ratio neural network model are established and trained according to obtained converter station energy efficiency data, the loss condition of a converter station under given operation parameters is predicted by utilizing the trained deep neural network model, the total loss of the converter station and the loss ratios of a converter, a converter transformer and a smoothing reactor are predicted, the universality is strong, and after the deep neural network model is trained, an energy efficiency distribution prediction task can be quickly completed; the energy efficiency distribution data reference method can provide energy efficiency distribution data reference for management and operation of the converter station, provide reference for research of the energy efficiency measurement technology of the high-voltage direct-current power transmission system and construction of an energy efficiency measurement monitoring system in future, and provide technical basis for aspects such as type selection of high-voltage direct-current power transmission system equipment, optimization operation mode of a power grid enterprise and the like.
Drawings
A more complete understanding of exemplary embodiments of the present invention may be had by reference to the following drawings in which:
fig. 1 is a flowchart of a deep learning-based energy efficiency prediction method 100 for a high-voltage direct current converter station according to an embodiment of the invention;
FIG. 2 is a schematic structural diagram of a converter station total loss neural network model according to an embodiment of the invention;
FIG. 3 is a schematic structural diagram of a converter station loss-to-fraction neural network model according to an embodiment of the invention;
fig. 4 is a schematic structural diagram of a deep learning-based energy efficiency prediction system 400 for a high-voltage direct current converter station according to an embodiment of the invention.
Detailed Description
The exemplary embodiments of the present invention will now be described with reference to the accompanying drawings, however, the present invention may be embodied in many different forms and is not limited to the embodiments described herein, which are provided for complete and complete disclosure of the present invention and to fully convey the scope of the present invention to those skilled in the art. The terminology used in the exemplary embodiments illustrated in the accompanying drawings is not intended to be limiting of the invention. In the drawings, the same units/elements are denoted by the same reference numerals.
Unless otherwise defined, terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. Further, it will be understood that terms, such as those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the relevant art and will not be interpreted in an idealized or overly formal sense.
Fig. 1 is a flowchart of a deep learning-based energy efficiency prediction method 100 for a high-voltage direct current converter station according to an embodiment of the present invention. As shown in fig. 1, according to the energy efficiency prediction method for a high-voltage direct-current converter station based on deep learning provided by the embodiment of the invention, a total loss neural network model and/or a loss proportion neural network model of the converter station are established and trained according to the obtained energy efficiency data of the converter station, the loss condition of the converter station under a given operation parameter is predicted by using the trained deep neural network model, the total loss of the converter station and the loss proportions of a converter, a converter transformer and a smoothing reactor are predicted, the universality is strong, and after the training of the deep neural network model is completed, an energy efficiency distribution prediction task can be quickly completed; the energy efficiency distribution data reference method can provide energy efficiency distribution data reference for management and operation of the converter station, provide reference for research of the energy efficiency measurement technology of the high-voltage direct-current power transmission system and construction of an energy efficiency measurement monitoring system in future, and provide technical basis for aspects such as type selection of high-voltage direct-current power transmission system equipment, optimization operation mode of a power grid enterprise and the like. According to the energy efficiency prediction method 100 for the high-voltage direct current converter station based on deep learning, which is provided by the embodiment of the invention, from step 101, an energy efficiency distribution sample data set of the high-voltage direct current converter station is obtained in step 101, and the energy efficiency distribution sample data set is preprocessed to obtain the preprocessed energy efficiency distribution sample data set.
Preferably, the energy efficiency distribution sample dataset comprises: an input sample data subset and an output sample data subset of the neural network model;
the variables in the subset of output sample data include: the total loss of the converter station and the loss of the converter, the converter transformer and the smoothing reactor account for the ratio;
when the type of the converter station is a rectifier station, the variables in the subset of input sample data include: the method comprises the following steps of (1) alternating-current end voltage, alternating-current end current, alternating-current end power factor, direct-current end voltage, direct-current end current, direct-current transmission power, a trigger angle and a commutation overlap angle of a rectifier;
when the type of the converter station is an inverter station, the variables in the input sample data subset include: ac terminal voltage, ac terminal current, ac terminal power factor, dc terminal voltage, dc terminal current, dc delivered power, inverter lead angle, arc-quenching angle, and commutation overlap angle.
Preferably, the preprocessing the energy efficiency distribution sample data set to obtain a preprocessed energy efficiency distribution sample data set includes:
calculating the average value and the standard deviation of each input variable according to the input sample data in the energy efficiency distribution sample data set;
for any input sample data, calculating the difference value between each input variable in the input sample data and the corresponding average value, calculating the ratio of the difference value to the corresponding standard deviation, and taking the ratio as the processed input sample data corresponding to the input sample data;
and determining the preprocessed energy efficiency distribution sample data set according to the processed data of each input sample and the output sample data corresponding to the processed data.
The energy efficiency prediction method is based on a deep learning technology, and a neural network model is obtained by training sample data of converter station energy efficiency distribution. In the training stage of the neural network, a large amount of computer resources and time are consumed for training the neural network, but after the training is finished, in the process of predicting by using the neural network, the neural network can quickly obtain an output result according to input data, and the method has good real-time performance.
For a vector a consisting of variable data completely describing the operating states of the converter station (a ═ a)1,a2,a3,..) corresponding to the unique vector B (B) consisting of the energy consumption distribution data of the converter stations1,b2,b3,..), that there is a non-linear multidimensional function: and f (a) ═ B, which describes the relationship between vector a and vector B. The method for calculating the loss of the converter station through theoretical analysis is an approximation of the multidimensional function, the deep learning technology uses a deep neural network model containing multiple processing layers, self parameters are continuously optimized from a large amount of training sample data by means of a back propagation algorithm, and intrinsic rules contained in the sample data are learned, so that the function can be accurately simulated by using the neural network model.
In the invention, an energy efficiency distribution sample data set is collected, and comprises an input sample data subset and an output sample data subset which are used for a neural network model. Depending on the type of converter station (rectifier station or inverter station), the input variables in the collected subset of input sample data are slightly different, but the output variables in the subset of output sample data are the same. Wherein, for the rectifier station, inputting variables in the subset of input sample data comprises: ac terminal voltage, current and power factor, dc terminal voltage and current, dc delivered power, trigger angle and commutation overlap angle of the rectifier. For an inverter station, inputting variables in the subset of input sample data includes: ac terminal voltage, current and power factor, dc terminal voltage and current, dc delivered power, lead angle, extinction angle and commutation overlap angle of the inverter. And the output variables in the output sample data subset comprise the total loss of the converter station and the loss ratio of the converter, the converter transformer and the smoothing reactor.
In the embodiment of the invention, when the energy efficiency distribution sample data set axis is obtained, the input sample data subset needs to be normalized, so that the input sample data subset can be processed by the deep neural network. The following operations are performed on each sample data in the input sample data subset by using the following method to realize normalization, including:
and calculating the average value of each input variable according to the sample data in the input sample data subset. For example, for the input variable ac terminal voltage, the ac terminal voltage data in all the input data samples are taken out individually, that is, the same number of ac terminal voltage data as the number of samples are obtained, and then the data are averaged.
And calculating the standard deviation of each input variable according to the sample data in the input sample data subset. The data processing method refers to the above step.
And for each data in the input sample data subset, subtracting the corresponding average value from the data according to the type of the input variable, and dividing the data by the corresponding standard deviation to obtain the normalized input sample processing data corresponding to each input sample data.
And recombining each input sample processing data according to the organization form of the original input sample data and the output sample data subset to obtain the processed energy efficiency distribution sample data set.
In step 102, a converter station total loss neural network model and/or a converter station loss ratio neural network model are/is constructed respectively.
Preferably, the converter station total loss neural network model is a sequence structure and comprises 5 hidden layers, each layer comprises 48 units, and an activation function is arranged between the layers; the number of units is determined by the input layer according to the type of the converter station; the output layer is 1 unit, an activation function is not used, and the output value is the total loss of the converter station;
the converter station loss ratio neural network model is of a sequence structure and comprises 7 hidden layers, 64 units are arranged on each layer, and an activation function is arranged between the layers; the number of units is determined by the input layer according to the type of the converter station; the output layer has 3 units, and does not use an activation function, and the output values are loss ratios of the current converter, the converter transformer and the smoothing reactor respectively.
In an embodiment of the invention, two different neural network models are constructed for prediction respectively. One neural network model is used for predicting the total loss of the converter station, and the other neural network model is used for predicting the loss ratio of the converter, the converter transformer and the smoothing reactor. Although the two neural network models have different structures, the training methods are basically the same.
And constructing a converter station total loss neural network model for predicting the converter station total loss. The structure of the neural network model is shown in fig. 2, and is a sequence structure, which comprises 5 hidden layers, each layer comprises 48 units, and an appropriate activation function is selected. The input layer has 8 or 9 units depending on the converter station type (rectifying station or inverting station). The output layer has only 1 unit and does not use the activation function, and the output value is the total loss of the converter station. For the training process of the neural network model, a proper loss function and an evaluation index should be selected. And constructing a converter station loss ratio neural network model for predicting the loss ratio of the converter, the converter transformer and the smoothing reactor. The structure of the neural network model is shown in fig. 3, and is a sequence structure, which comprises 7 hidden layers, each layer comprises 64 units, and an appropriate activation function is selected. The input layer has 8 or 9 units depending on the converter station type (rectifying station or inverting station). The output layer has 3 units, an activation function is not used, and the output value corresponds to the loss ratio of a current converter, a converter transformer and a smoothing reactor. For the training process of the neural network model, a proper loss function and an evaluation index should be selected.
In step 103, training and optimizing the converter station total loss neural network model and/or the converter station loss ratio neural network model respectively based on the preprocessed energy efficiency distribution sample data set, so as to determine a converter station total loss neural network optimal model and a converter station loss ratio neural network optimal model.
Preferably, wherein the method further comprises:
in the training process of the converter station total loss neural network model and/or the converter station loss ratio neural network model, adjusting weight parameters of internal units of the neural network model by using a gradient descent algorithm so as to enable the output result of the neural network model to approach to a target value;
and after the converter station total loss neural network model and/or the converter station loss ratio neural network model are trained, verifying the accuracy of the neural network model prediction by using a cross-folding cross verification method.
In an embodiment of the invention, the converter station total loss neural network model and/or the converter station loss ratio neural network model are/is trained and optimized respectively based on the preprocessed energy efficiency distribution sample data set, so as to determine a converter station total loss neural network optimal model and a converter station loss ratio neural network optimal model. In the training process, the training times need to be adjusted reasonably, and the prediction performance of the neural network model is greatly influenced by too few or too many training times. The training times should be smaller, and then gradually increased until the overfitting phenomenon is found, so that the proper training times can be obtained.
Before the training of the neural network model is started, the training times are selected to be 20 times preliminarily, and a proper optimization algorithm is used as an optimizer to carry out steepest descent optimization on the neural network model. In the process of model training, the gradient descent algorithm adjusts the weight parameters of the internal units of the neural network model, and finally the output result approaches to a target value.
After the deep neural network model is trained, the prediction accuracy is subjected to actual verification by using a verification method such as ten-fold cross verification, and if the model does not have an under-fitting phenomenon caused by too few training times or an over-fitting phenomenon caused by too many training times after the verification is finished, a final prediction model can be obtained, and the model can complete better prediction on new input data. The ten-fold cross validation method is as follows: dividing the data sample set into ten parts equally, then using each part as a test sample set and the rest nine parts as training sample sets in the ten training processes, testing the obtained neural network model by using the test sample set after each training, comparing the ten finally obtained test results, and if the high-accuracy prediction can be stably carried out in the ten test results according to the parameter configuration of the neural network model, showing that the parameter configuration of the neural network model is reasonable without the problem of overfitting, otherwise, adjusting the training times to avoid the phenomena of overfitting and underfitting.
And 104, predicting the energy efficiency of the high-voltage direct-current converter station to be tested by using the optimal model of the total loss neural network of the converter station and/or the optimal model of the loss ratio neural network of the converter station according to the operation data of the high-voltage direct-current converter station to be tested so as to obtain an energy efficiency prediction result.
According to the method, two different neural network models are used according to data such as alternating-current terminal voltage, current and power factor, direct-current terminal voltage and current, direct-current transmission power, trigger angle, lead angle, arc quenching angle and commutation overlap angle of a converter and the like in the converter station, so that the total loss of the converter station and the loss proportion of the converter, a converter transformer and a smoothing reactor can be well predicted, energy efficiency distribution data reference is provided for management and operation of the converter station, and a solid foundation can be provided for research of energy efficiency metering technology of a high-voltage direct-current transmission system in the future.
The following specifically exemplifies embodiments of the present invention
The method provided by the embodiment of the invention can predict the ratio of the total loss of the direct current converter station to the loss of main elements in the converter station. The energy efficiency prediction method of the high-voltage direct current converter station based on deep learning comprises the following steps:
step 1: selecting variables such as alternating current end voltage, current and power factor, direct current end voltage and current, direct current transmission power, trigger angle, lead angle, extinction angle and commutation overlap angle of a converter from historical operation data of the converter station to form an input vector, selecting total loss of the converter station to form an output vector P, selecting variables such as loss ratio of the converter, a converter transformer and a smoothing reactor to form an output vector R, and collecting a plurality of corresponding input and output variable data to form a sample data set. For one sample, the input vector of the neural network model of the rectifier station has 8 elements, the input vector of the neural network model of the inverter station has 9 elements, the output vector P of the neural network model for predicting the total loss of the converter station has 1 element, and the output vector R of the neural network model for predicting the loss ratio of the converter, the converter transformer and the smoothing reactor has 3 elements.
Step 2: uniformly normalizing the input data in the sample data set obtained in the step 1, wherein the specific implementation process is as follows:
step 2-1: the input variables of each type in the sample data set are averaged.
Step 2-2: the average value found in step 2-1 is subtracted from each type of input variable in the sample data set.
Step 2-3: the standard deviation is calculated for each type of input variable in the sample data set.
Step 2-4: and (4) subtracting the standard deviation obtained in the step 2-3 from each type of input variable in the sample data set.
And step 3: training a neural network model by using the sample data set obtained after the processing in the steps 1 and 2, wherein the specific implementation process is as follows:
step 3-1: a neural network model of a sequence structure is built, the neural network model comprises 5 hidden layers, each layer comprises 48 units, and the hidden layers are dense layers which use modified linear units (ReLU) as activation functions. The input layer has 8 or 9 units depending on the converter station type (rectifying station or inverting station) and is a dense layer using modified linear units (relus) as activation functions. The output layer has only 1 cell and does not use the activation function. In the neural network model process, Mean Square Error (MSE) is used as a loss function, and Mean Absolute Error (MAE) is used as an evaluation index. The model is trained using the output vector and the output vector P for predicting the total loss of the dc converter station.
Step 3-2: another sequence structured neural network model was constructed, which contained 7 hidden layers, each layer containing 64 cells, and the hidden layers were all dense layers using modified linear cells (relus) as activation functions. The input layer has 8 or 9 units depending on the converter station type (rectifying station or inverting station) and is a dense layer using modified linear units (relus) as activation functions. The output layer has 3 cells and does not use the activation function. In the neural network model process, Mean Square Error (MSE) is used as a loss function, and Mean Absolute Error (MAE) is used as an evaluation index. The model is trained by using an output vector and an output vector R and is used for predicting the loss ratio of a current converter, a converter transformer and a smoothing reactor.
Step 3-3: and (3) as for the neural network models built in the steps 3-1 and 3-2, training by using an Adam algorithm as an optimization algorithm, firstly training for 20 times by using a sample data set, and observing the prediction performance result of the models.
And 4, step 4: and (4) verifying the neural network model trained in the step (3-3) by using a ten-fold cross verification method, and obtaining two final neural network models which are respectively used for different prediction tasks.
And 5: and predicting the energy efficiency of the high-voltage direct-current converter station to be tested by utilizing a converter station total loss neural network optimal model and/or a converter station loss ratio neural network optimal model according to the operation data of the high-voltage direct-current converter station to be tested so as to obtain an energy efficiency prediction result.
Fig. 4 is a schematic structural diagram of a deep learning-based energy efficiency prediction system 400 for a high-voltage direct current converter station according to an embodiment of the invention. As shown in fig. 4, a deep learning based energy efficiency prediction system 400 for a high voltage direct current converter station according to an embodiment of the present invention includes: the energy efficiency prediction method comprises a sample data preprocessing unit 401, a neural network model building unit 402, a neural network model training unit 403 and an energy efficiency prediction unit 404.
Preferably, the sample data preprocessing unit 401 is configured to acquire an energy efficiency distribution sample data set of the high-voltage direct-current converter station, and preprocess the energy efficiency distribution sample data set to acquire the preprocessed energy efficiency distribution sample data set.
Preferably, the energy efficiency distribution sample dataset comprises: an input sample data subset and an output sample data subset of the neural network model;
the variables in the subset of output sample data include: the total loss of the converter station and the loss of the converter, the converter transformer and the smoothing reactor account for the ratio;
when the type of the converter station is a rectifier station, the variables in the subset of input sample data include: the method comprises the following steps of (1) alternating-current end voltage, alternating-current end current, alternating-current end power factor, direct-current end voltage, direct-current end current, direct-current transmission power, a trigger angle and a commutation overlap angle of a rectifier;
when the type of the converter station is an inverter station, the variables in the input sample data subset include: ac terminal voltage, ac terminal current, ac terminal power factor, dc terminal voltage, dc terminal current, dc delivered power, inverter lead angle, arc-quenching angle, and commutation overlap angle.
Preferably, the sample data preprocessing unit 401 preprocesses the energy efficiency distribution sample data set to obtain a preprocessed energy efficiency distribution sample data set, including:
calculating the average value and the standard deviation of each input variable according to the input sample data in the energy efficiency distribution sample data set;
for any input sample data, calculating the difference value between each input variable in the input sample data and the corresponding average value, calculating the ratio of the difference value to the corresponding standard deviation, and taking the ratio as the processed input sample data corresponding to the input sample data;
and determining the preprocessed energy efficiency distribution sample data set according to the processed data of each input sample and the output sample data corresponding to the processed data.
Preferably, the neural network model building unit 402 is configured to respectively build a converter station total loss neural network model and/or a converter station loss ratio neural network model.
Preferably, the converter station total loss neural network model is a sequence structure and comprises 5 hidden layers, each layer comprises 48 units, and an activation function is arranged between the layers; the number of units is determined by the input layer according to the type of the converter station; the output layer is 1 unit, an activation function is not used, and the output value is the total loss of the converter station;
the converter station loss ratio neural network model is of a sequence structure and comprises 7 hidden layers, 64 units are arranged on each layer, and an activation function is arranged between the layers; the number of units is determined by the input layer according to the type of the converter station; the output layer has 3 units, and does not use an activation function, and the output values are loss ratios of the current converter, the converter transformer and the smoothing reactor respectively.
Preferably, the neural network model training unit 403 is configured to train and optimize the converter station total loss neural network model and/or the converter station loss occupation ratio neural network model based on the preprocessed energy efficiency distribution sample data set, so as to determine a converter station total loss neural network optimal model and a converter station loss occupation ratio neural network optimal model.
Preferably, the neural network model training unit 403 further includes:
in the training process of the converter station total loss neural network model and/or the converter station loss ratio neural network model, adjusting weight parameters of internal units of the neural network model by using a gradient descent algorithm so as to enable the output result of the neural network model to approach to a target value;
and after the converter station total loss neural network model and/or the converter station loss ratio neural network model are trained, verifying the accuracy of the neural network model prediction by using a cross-folding cross verification method.
Preferably, the energy efficiency predicting unit 404 is configured to predict the energy efficiency of the to-be-tested high-voltage direct-current converter station by using the converter station total loss neural network optimal model and/or the converter station loss ratio neural network optimal model according to the operation data of the to-be-tested high-voltage direct-current converter station, so as to obtain an energy efficiency prediction result.
The deep learning-based energy efficiency prediction system 400 of the high-voltage direct current converter station according to the embodiment of the present invention corresponds to the deep learning-based energy efficiency prediction method 100 of the high-voltage direct current converter station according to another embodiment of the present invention, and details thereof are not repeated herein.
The invention has been described with reference to a few embodiments. However, other embodiments of the invention than the one disclosed above are equally possible within the scope of the invention, as would be apparent to a person skilled in the art from the appended patent claims.
Generally, all terms used in the claims are to be interpreted according to their ordinary meaning in the technical field, unless explicitly defined otherwise herein. All references to "a/an/the [ device, component, etc ]" are to be interpreted openly as referring to at least one instance of said device, component, etc., unless explicitly stated otherwise. The steps of any method disclosed herein do not have to be performed in the exact order disclosed, unless explicitly stated.
As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
Finally, it should be noted that: the above embodiments are only for illustrating the technical solutions of the present invention and not for limiting the same, and although the present invention is described in detail with reference to the above embodiments, those of ordinary skill in the art should understand that: modifications and equivalents may be made to the embodiments of the invention without departing from the spirit and scope of the invention, which is to be covered by the claims.
Claims (10)
1. The method for predicting the energy efficiency of the high-voltage direct-current converter station based on deep learning is characterized by comprising the following steps:
the method comprises the steps of obtaining an energy efficiency distribution sample data set of the high-voltage direct-current converter station, and preprocessing the energy efficiency distribution sample data set to obtain a preprocessed energy efficiency distribution sample data set;
respectively constructing a converter station total loss neural network model and/or a converter station loss ratio neural network model;
training and optimizing the converter station total loss neural network model and/or the converter station loss ratio neural network model respectively based on the preprocessed energy efficiency distribution sample data set so as to determine a converter station total loss neural network optimal model and a converter station loss ratio neural network optimal model;
and predicting the energy efficiency of the high-voltage direct-current converter station to be tested by utilizing the converter station total loss neural network optimal model and/or the converter station loss ratio neural network optimal model according to the operation data of the high-voltage direct-current converter station to be tested so as to obtain an energy efficiency prediction result.
2. The method of claim 1, wherein the energy efficiency distribution sample dataset comprises: an input sample data subset and an output sample data subset of the neural network model;
the variables in the subset of output sample data include: the total loss of the converter station and the loss of the converter, the converter transformer and the smoothing reactor account for the ratio;
when the type of the converter station is a rectifier station, the variables in the subset of input sample data include: the method comprises the following steps of (1) alternating-current end voltage, alternating-current end current, alternating-current end power factor, direct-current end voltage, direct-current end current, direct-current transmission power, a trigger angle and a commutation overlap angle of a rectifier;
when the type of the converter station is an inverter station, the variables in the input sample data subset include: ac terminal voltage, ac terminal current, ac terminal power factor, dc terminal voltage, dc terminal current, dc delivered power, inverter lead angle, arc-quenching angle, and commutation overlap angle.
3. The method according to claim 1, wherein the preprocessing the energy efficiency distribution sample dataset to obtain a preprocessed energy efficiency distribution sample dataset includes:
calculating the average value and the standard deviation of each input variable according to the input sample data in the energy efficiency distribution sample data set;
for any input sample data, calculating the difference value between each input variable in the input sample data and the corresponding average value, calculating the ratio of the difference value to the corresponding standard deviation, and taking the ratio as the processed input sample data corresponding to the input sample data;
and determining the preprocessed energy efficiency distribution sample data set according to the processed data of each input sample and the output sample data corresponding to the processed data.
4. The method according to claim 1, wherein the converter station total loss neural network model is a sequence structure and comprises 5 hidden layers, each layer comprises 48 units, and activation functions are arranged among the layers; the number of units is determined by the input layer according to the type of the converter station; the output layer is 1 unit, an activation function is not used, and the output value is the total loss of the converter station;
the converter station loss ratio neural network model is of a sequence structure and comprises 7 hidden layers, 64 units are arranged on each layer, and an activation function is arranged between the layers; the number of units is determined by the input layer according to the type of the converter station; the output layer has 3 units, and does not use an activation function, and the output values are loss ratios of the current converter, the converter transformer and the smoothing reactor respectively.
5. The method of claim 1, further comprising:
in the training process of the converter station total loss neural network model and/or the converter station loss ratio neural network model, adjusting weight parameters of internal units of the neural network model by using a gradient descent algorithm so as to enable the output result of the neural network model to approach to a target value;
and after the converter station total loss neural network model and/or the converter station loss ratio neural network model are trained, verifying the accuracy of the neural network model prediction by using a cross-folding cross verification method.
6. A high-voltage direct current converter station energy efficiency prediction system based on deep learning is characterized by comprising the following components:
the sample data preprocessing unit is used for acquiring an energy efficiency distribution sample data set of the high-voltage direct-current converter station and preprocessing the energy efficiency distribution sample data set to acquire a preprocessed energy efficiency distribution sample data set;
the neural network model building unit is used for respectively building a converter station total loss neural network model and/or a converter station loss ratio neural network model;
the neural network model training unit is used for respectively training and optimizing the converter station total loss neural network model and/or the converter station loss ratio neural network model based on the preprocessed energy efficiency distribution sample data set so as to determine a converter station total loss neural network optimal model and a converter station loss ratio neural network optimal model;
and the energy efficiency prediction unit is used for predicting the energy efficiency of the high-voltage direct-current converter station to be tested by utilizing the converter station total loss neural network optimal model and/or the converter station loss ratio neural network optimal model according to the operation data of the high-voltage direct-current converter station to be tested so as to obtain an energy efficiency prediction result.
7. The system of claim 6, wherein the energy efficiency distribution sample dataset comprises: an input sample data subset and an output sample data subset of the neural network model;
the variables in the subset of output sample data include: the total loss of the converter station and the loss of the converter, the converter transformer and the smoothing reactor account for the ratio;
when the type of the converter station is a rectifier station, the variables in the subset of input sample data include: the method comprises the following steps of (1) alternating-current end voltage, alternating-current end current, alternating-current end power factor, direct-current end voltage, direct-current end current, direct-current transmission power, a trigger angle and a commutation overlap angle of a rectifier;
when the type of the converter station is an inverter station, the variables in the input sample data subset include: ac terminal voltage, ac terminal current, ac terminal power factor, dc terminal voltage, dc terminal current, dc delivered power, inverter lead angle, arc-quenching angle, and commutation overlap angle.
8. The system according to claim 6, wherein the sample data preprocessing unit preprocesses the energy efficiency distribution sample dataset to obtain a preprocessed energy efficiency distribution sample dataset, and includes:
calculating the average value and the standard deviation of each input variable according to the input sample data in the energy efficiency distribution sample data set;
for any input sample data, calculating the difference value between each input variable in the input sample data and the corresponding average value, calculating the ratio of the difference value to the corresponding standard deviation, and taking the ratio as the processed input sample data corresponding to the input sample data;
and determining the preprocessed energy efficiency distribution sample data set according to the processed data of each input sample and the output sample data corresponding to the processed data.
9. The system according to claim 6, wherein the converter station total loss neural network model is a sequence structure and comprises 5 hidden layers, each layer comprises 48 units, and activation functions are arranged among the layers; the number of units is determined by the input layer according to the type of the converter station; the output layer is 1 unit, an activation function is not used, and the output value is the total loss of the converter station;
the converter station loss ratio neural network model is of a sequence structure and comprises 7 hidden layers, 64 units are arranged on each layer, and an activation function is arranged between the layers; the number of units is determined by the input layer according to the type of the converter station; the output layer has 3 units, and does not use an activation function, and the output values are loss ratios of the current converter, the converter transformer and the smoothing reactor respectively.
10. The system of claim 6, wherein the neural network model training unit further comprises:
in the training process of the converter station total loss neural network model and/or the converter station loss ratio neural network model, adjusting weight parameters of internal units of the neural network model by using a gradient descent algorithm so as to enable the output result of the neural network model to approach to a target value;
and after the converter station total loss neural network model and/or the converter station loss ratio neural network model are trained, verifying the accuracy of the neural network model prediction by using a cross-folding cross verification method.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011079041.6A CN112365024A (en) | 2020-10-10 | 2020-10-10 | High-voltage direct-current converter station energy efficiency prediction method and system based on deep learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011079041.6A CN112365024A (en) | 2020-10-10 | 2020-10-10 | High-voltage direct-current converter station energy efficiency prediction method and system based on deep learning |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112365024A true CN112365024A (en) | 2021-02-12 |
Family
ID=74507923
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011079041.6A Pending CN112365024A (en) | 2020-10-10 | 2020-10-10 | High-voltage direct-current converter station energy efficiency prediction method and system based on deep learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112365024A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116345447A (en) * | 2023-04-12 | 2023-06-27 | 刘泓利 | Power generation electric energy transmission loss evaluation system |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103995981A (en) * | 2014-06-06 | 2014-08-20 | 中国能源建设集团广东省电力设计研究院 | Method for assessing loss of MMC current converter in flexible direct-current transmission system |
| CN104408562A (en) * | 2014-11-25 | 2015-03-11 | 许继集团有限公司 | Photovoltaic system generating efficiency comprehensive evaluation method based on BP (back propagation) neural network |
| CN105116233A (en) * | 2015-08-18 | 2015-12-02 | 中国南方电网有限责任公司超高压输电公司检修试验中心 | Method for predicting loss of DC system |
-
2020
- 2020-10-10 CN CN202011079041.6A patent/CN112365024A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103995981A (en) * | 2014-06-06 | 2014-08-20 | 中国能源建设集团广东省电力设计研究院 | Method for assessing loss of MMC current converter in flexible direct-current transmission system |
| CN104408562A (en) * | 2014-11-25 | 2015-03-11 | 许继集团有限公司 | Photovoltaic system generating efficiency comprehensive evaluation method based on BP (back propagation) neural network |
| CN105116233A (en) * | 2015-08-18 | 2015-12-02 | 中国南方电网有限责任公司超高压输电公司检修试验中心 | Method for predicting loss of DC system |
Non-Patent Citations (2)
| Title |
|---|
| 史丹等: ""高压直流换流站损耗计算软件的开发和应用"", 《电力系统自动化》, 31 October 2006 (2006-10-31), pages 81 - 83 * |
| 赵向阳等: ""基于神经网络的变压器损耗计算方法"", 《电力科学与工程》, 31 January 2015 (2015-01-31), pages 44 - 48 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116345447A (en) * | 2023-04-12 | 2023-06-27 | 刘泓利 | Power generation electric energy transmission loss evaluation system |
| CN116345447B (en) * | 2023-04-12 | 2023-10-13 | 山东奥德智慧新能源有限公司 | Power generation electric energy transmission loss evaluation system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Chen et al. | Applications of multi-objective dimension-based firefly algorithm to optimize the power losses, emission, and cost in power systems | |
| US5625751A (en) | Neural network for contingency ranking dynamic security indices for use under fault conditions in a power distribution system | |
| CN113805064B (en) | Lithium ion battery pack health state prediction method based on deep learning | |
| CN112147530B (en) | Battery state evaluation method and device | |
| CN102880908B (en) | Method for calculating remanufactured part environmental loss based on back propagation (BP) neural network | |
| Li et al. | Application of fuzzy analytic hierarchy process and neural network in power transformer risk assessment | |
| CN109784692A (en) | A kind of fast and safely constraint economic load dispatching method based on deep learning | |
| CN111563682A (en) | Test evaluation method for distribution automation equipment | |
| CN110309979A (en) | Methods of electric load forecasting, device and equipment based on echo state network | |
| CN108090615A (en) | Low-limit frequency Forecasting Methodology after electric power system fault based on cross entropy integrated study | |
| CN112101736A (en) | Micro-grid operation mode evaluation method and device, storage medium and electronic equipment | |
| CN113743674A (en) | Energy storage output prediction method, system, equipment and medium based on deep learning | |
| CN112241836A (en) | Virtual load dominant parameter identification method based on incremental learning | |
| CN108647838A (en) | A kind of power grid Situation Awareness method based on Random Matrices Theory and online limit of sequence learning machine | |
| CN111582630A (en) | Method and system for determining low-voltage transformer area line loss rate evaluation value | |
| Uzun Araz et al. | A reactive scheduling approach based on fuzzy inference for hybrid flowshop systems | |
| CN114897331A (en) | Power transformer risk assessment method based on three-parameter interval gray number decision | |
| CN114757392A (en) | Power grid system comprehensive evaluation index system optimization method based on game weight method | |
| CN112365024A (en) | High-voltage direct-current converter station energy efficiency prediction method and system based on deep learning | |
| CN112101673A (en) | Power grid development trend prediction method and system based on hidden Markov model | |
| CN113435677A (en) | Power grid operation risk active analysis system and method | |
| Guddanti et al. | Graph neural networks for voltage stability margins with topology flexibilities | |
| CN116823008A (en) | Park energy utilization efficiency evaluation method, system, equipment and storage medium | |
| CN114091243A (en) | Product concept design reliability prediction method and system | |
| Cao et al. | A novel pca-bp fuzzy neural network model for software defect prediction |
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 |