CN112964961A - Electric-gas coupling comprehensive energy system fault positioning method and system - Google Patents
Electric-gas coupling comprehensive energy system fault positioning method and system Download PDFInfo
- Publication number
- CN112964961A CN112964961A CN202110155560.4A CN202110155560A CN112964961A CN 112964961 A CN112964961 A CN 112964961A CN 202110155560 A CN202110155560 A CN 202110155560A CN 112964961 A CN112964961 A CN 112964961A
- Authority
- CN
- China
- Prior art keywords
- node
- gas
- fault
- power
- sensitivity
- 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.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/08—Locating faults in cables, transmission lines, or networks
- G01R31/081—Locating faults in cables, transmission lines, or networks according to type of conductors
- G01R31/086—Locating faults in cables, transmission lines, or networks according to type of conductors in power transmission or distribution networks, i.e. with interconnected conductors
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/08—Locating faults in cables, transmission lines, or networks
- G01R31/088—Aspects of digital computing
-
- 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
- Y04S10/52—Outage or fault management, e.g. fault detection or location
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Theoretical Computer Science (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
Compared with the traditional fault positioning method, the fault characteristics which cannot be expressed simply by using a mathematical formula can be captured by adopting the convolutional neural network, and the node voltage data of the power grid is selected as the input of the convolutional neural network, and the node air pressure data of the air grid node with high sensitivity is simultaneously selected as the input, so that the fault characteristics can be extracted by the convolutional neural network more favorably, the fault positioning accuracy is improved, the intelligent level of the electric-gas coupling comprehensive energy system is improved, and the technical problem of low fault positioning accuracy of the existing fault positioning method of the electric-gas coupling comprehensive energy system is solved.
Description
Technical Field
The application relates to the technical field of comprehensive energy, in particular to a fault positioning method and system for an electric-gas coupling comprehensive energy system.
Background
Integrated Energy Systems (IES) are known as one of the main ways to realize comprehensive utilization of multiple Energy sources, and scientific management and optimized scheduling between different grades of Energy sources such as cold/heat/electricity/gas in the system can improve the efficient utilization efficiency of the Energy sources and realize the full consumption of renewable Energy sources. However, the IES achieves the above-mentioned objective and brings an overall risk of safe operation of the system, and in the face of tight coupling of multiple energy sources, "pulling one energy to move the whole body" between different energy supply systems, which mutually affect each other, and if the power grid fails, the state of the coupling unit will be affected, and further the state of the natural gas grid will be changed, so that when the power grid fails, a cascading failure may be caused to affect the natural gas grid subsystem, and finally the economic and reliable operation of the integrated energy system is threatened. Therefore, the energy system is quickly and accurately integrated to carry out fault location, and the premise of avoiding cascading faults and recovering power supply is provided.
After a fault occurs, for a power grid, node voltage and injection power change instantly, a random power grid line protection device acts rapidly to cause power failure in a partial area, power grid load fluctuates, air pressure of an air grid node is influenced through a coupling link, and the influence degree is related to the position of the fault.
Disclosure of Invention
The application provides a fault positioning method and system for an electric-gas coupling comprehensive energy system, which are used for solving the technical problem that the fault positioning accuracy is low in the existing fault positioning method for the electric-gas coupling comprehensive energy system.
In view of the above, a first aspect of the present application provides an electric-gas coupling integrated energy system fault location method, including:
analyzing the sensitivity of the node air pressure of the air network node to the power grid node injection power, and acquiring the air network node with the higher sensitivity by 30% as an air pressure monitoring node;
performing optimal configuration on a synchronous vector measurement unit of a power grid by taking the complete observability of a power grid system as a target to obtain a voltage monitoring node corresponding to the air pressure monitoring node;
simulating a power grid line fault, and acquiring fault data of the air pressure monitoring node and the voltage monitoring node;
inputting the fault data into a convolutional neural network which takes the difference value of the node air pressure before and after the fault and the difference value and the phase angle of the node voltage before and after the fault as input and takes a fault section as output to train so as to obtain a convolutional neural network model for fault diagnosis;
and positioning the fault of the electric-gas coupling integrated energy system by using the convolutional neural network model for fault diagnosis.
Optionally, the analyzing the sensitivity of the node air pressure of the air network node to the power grid node injection power, and acquiring the air network node with a height of 30% before the sensitivity as an air pressure monitoring node includes:
forming a node-branch incidence matrix of the gas network according to the read state information of the gas network, and forming a power grid node admittance matrix according to the read power grid network information;
performing energy flow calculation on the electricity-gas coupling comprehensive energy system, solving an energy flow equation set, and obtaining a current operating point and a Jacobian matrix of the current operating point;
calculating gas pressure-gas load sensitivity and gas boiler output-node injection power sensitivity according to the Jacobi matrix;
calculating gas pressure-node injection power sensitivity according to the gas pressure-gas load sensitivity and the gas boiler output-node injection power sensitivity;
and sequencing the gas pressure-node injection power sensitivity in a descending order, and taking the gas network node 30% before the gas pressure-node injection power sensitivity as a gas pressure monitoring node.
Optionally, the system of fluence equations is:
wherein, a is a node-branch incidence matrix of the gas network, Y is a network node admittance matrix, and x ═ θ, V, p]For the state variable of the electric-gas coupling comprehensive energy system, theta is the phase angle of each node of the power distribution network except for the balance node, V is the voltage of each electric load node of each node of the power distribution network except for the balance node, P is the gas pressure of each gas load node of each node of the power distribution network except for the balance node, and u is [ P ═ P [ [ P ]sp,Qsp,Lsp]For a control variable, P, of an electro-pneumatic coupled integrated energy systemspInjecting active power, Q, for a given nodespFor a given reactive power, LspFor a given gas load, f is the steady-state flow column vector of the pipeline, PMTFor the output electric power of a gas boiler, LMTFor gas consumption of gas boilers, cgeThe conversion efficiency of the gas boiler.
Optionally, the formula for calculating the gas pressure-node injection power sensitivity is as follows:
optionally, the convolutional neural network comprises an input layer, a convolutional layer, a pooling layer, a full-link layer and a softmax layer, the activation function is a ReLU, and a Dropout algorithm is adopted in the convolutional layer.
Optionally, the locating the fault of the electro-pneumatic coupling integrated energy system by using the convolutional neural network model for fault diagnosis further comprises:
and verifying the accuracy of the convolutional neural network model for fault diagnosis through untrained fault data, and if the fault positioning accuracy of the convolutional neural network model for fault diagnosis is greater than preset accuracy, the verification is passed.
The present application provides in a second aspect an electric-electric coupling integrated energy system fault location system, comprising:
the air pressure monitoring node acquisition unit is used for analyzing the sensitivity of the node air pressure of the air network node to the power grid node injection power and acquiring the air network node with the height of 30% of the sensitivity as the air pressure monitoring node;
the voltage monitoring node acquisition unit is used for carrying out optimized configuration on a synchronous vector measurement unit of the power grid by taking the complete observability of the power grid system as a target to obtain a voltage monitoring node corresponding to the air pressure monitoring node;
the fault data acquisition unit is used for simulating a power grid line fault and acquiring fault data of the air pressure monitoring node and the voltage monitoring node;
the model training unit is used for inputting the fault data into a convolutional neural network which takes the difference value of the node air pressure before and after the fault and the difference value and the phase angle of the node voltage before and after the fault as input and takes a fault section as output to train so as to obtain a convolutional neural network model for fault diagnosis;
and the fault positioning unit is used for positioning the fault of the electric-gas coupling comprehensive energy system by using the convolutional neural network model for fault diagnosis.
Optionally, the air pressure monitoring node is specifically configured to:
forming a node-branch incidence matrix of the gas network according to the read state information of the gas network, and forming a power grid node admittance matrix according to the read power grid network information;
performing energy flow calculation on the electricity-gas coupling comprehensive energy system, solving an energy flow equation set, and obtaining a current operating point and a Jacobian matrix of the current operating point;
calculating gas pressure-gas load sensitivity and gas boiler output-node injection power sensitivity according to the Jacobi matrix;
calculating gas pressure-node injection power sensitivity according to the gas pressure-gas load sensitivity and the gas boiler output-node injection power sensitivity;
and sequencing the gas pressure-node injection power sensitivity in a descending order, and taking the gas network node 30% before the gas pressure-node injection power sensitivity as a gas pressure monitoring node.
Optionally, the system of fluence equations is:
wherein, a is a node-branch incidence matrix of the gas network, Y is a network node admittance matrix, and x ═ θ, V, p]For the state variable of the electric-gas coupling comprehensive energy system, theta is the phase angle of each node of the power distribution network except for the balance node, V is the voltage of each electric load node of each node of the power distribution network except for the balance node, P is the gas pressure of each gas load node of each node of the power distribution network except for the balance node, and u is [ P ═ P [ [ P ]sp,Qsp,Lsp]For a control variable, P, of an electro-pneumatic coupled integrated energy systemspInjecting active power, Q, for a given nodespFor a given reactive power, LspFor a given gas load, f is the steady-state flow column vector of the pipeline, PMTFor the output electric power of a gas boiler, LMTFor gas consumption of gas boilers, cgeThe conversion efficiency of the gas boiler.
Optionally, the method further comprises:
and the model verification unit is used for verifying the accuracy of the convolutional neural network model for fault diagnosis through untrained fault data, and if the fault positioning accuracy of the convolutional neural network model for fault diagnosis is greater than preset accuracy, the verification is passed.
According to the technical scheme, the embodiment of the application has the following advantages:
the application provides an electric-gas coupling comprehensive energy system fault positioning method, which comprises the following steps: analyzing the sensitivity of the node air pressure of the air network node to the power grid node injection power, and acquiring the air network node with the higher sensitivity by 30% as an air pressure monitoring node; the method comprises the steps that optimal configuration is carried out on a synchronous vector measurement unit of a power grid by taking the complete observability of a power grid system as a target, and voltage monitoring nodes corresponding to air pressure monitoring nodes are obtained; simulating a power grid line fault, and acquiring fault data of an air pressure monitoring node and a voltage monitoring node; inputting fault data into a convolutional neural network which takes the difference value of node air pressure before and after the fault and the difference value and phase angle of node voltage before and after the fault as input and takes a fault section as output to train so as to obtain a convolutional neural network model for fault diagnosis; and positioning the fault of the electric-gas coupling integrated energy system by using a convolutional neural network model for fault diagnosis.
Compared with the traditional fault positioning method, the fault characteristics which cannot be expressed simply by using a mathematical formula can be captured by adopting the convolutional neural network, and the node voltage data of the power grid is selected as the input of the convolutional neural network, and the node air pressure data of the air grid node with high sensitivity is simultaneously selected as the input, so that the fault characteristics can be extracted by the convolutional neural network more favorably, the fault positioning accuracy is improved, the intelligent level of the electric-gas coupling comprehensive energy system is improved, and the technical problem of low fault positioning accuracy of the existing fault positioning method of the electric-gas coupling comprehensive energy system is solved.
Meanwhile, when the electric-gas coupling comprehensive energy system fault positioning method is used for configuring the synchronous vector measuring unit, the synchronous vector measuring unit of the power grid is optimally configured under the condition of complete observability of the system, so that the redundancy of sampled data is reduced, and the economy is improved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other related drawings can be obtained according to the drawings without creative efforts.
Fig. 1 is a schematic flow chart of a fault location method of an electrical-electrical coupling integrated energy system provided in an embodiment of the present application;
FIG. 2 is a schematic structural diagram of a convolutional neural network model provided in an embodiment of the present application;
fig. 3 is a schematic diagram of an electric-gas coupling integrated energy system model provided in an application example of the present application.
Detailed Description
In order to make the technical solutions of the present application better understood, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only a part of the embodiments of the present application, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
Example 1
For easy understanding, referring to fig. 1 to 2, an embodiment of a fault location method for an electrical-electrical coupling integrated energy system provided by the present application includes:
It should be noted that the sensitivity required to be obtained in the embodiment of the present application is the sensitivity of the gas pressure of the gas network node to the injection power of the power network node, and is intended to select a node with higher sensitivity as a gas pressure monitoring node.
And 102, optimally configuring a synchronous vector measurement unit of the power grid by taking the complete observability of the power grid system as a target to obtain a voltage monitoring node corresponding to the air pressure monitoring node.
It should be noted that, in consideration of the economy of the power distribution network, in the embodiment of the present application, a method for configuring a Phasor Measurement Unit (PMU) targeting complete observability of the system is adopted, and the minimum number n of PMUs is configuredpAnd obtaining the optimal configuration result of the PMU by the most appropriate position and the maximum network structure observability and the larger data redundancy, namely obtaining the corresponding voltage monitoring node.
And 103, simulating a power grid line fault, and acquiring fault data of the air pressure monitoring node and the voltage monitoring node.
It should be noted that, after the air pressure monitoring node and the voltage monitoring node are obtained, the line fault of the power grid is simulated through simulation software, and information of the corresponding monitoring node is recorded.
And 104, inputting fault data into a convolutional neural network which takes the difference value of the node air pressure before and after the fault and the difference value of the node voltage before and after the fault as input and takes a fault section as output to train so as to obtain a convolutional neural network model for fault diagnosis.
The difference value of the node air pressure before and after the fault and the amplitude and the phase angle of the difference value of the node voltage before and after the fault are used as the input of the convolutional neural network, and the corresponding fault section is used as the output of the convolutional neural network, so that the convolutional neural network is trained. The fault data can be divided into a training set and a testing set according to the proportion of (7:3), and the fault diagnosis can be carried out through the trained convolutional neural network model.
And 105, positioning the fault of the electric-gas coupling comprehensive energy system by using the convolutional neural network model for fault diagnosis.
It should be noted that after the convolutional neural network model for fault diagnosis is obtained, the convolutional neural network model for fault diagnosis may be applied to the electric-pneumatic coupling integrated energy system, and the difference between the node air pressures before and after the fault and the amplitude and phase angle of the difference between the node voltages before and after the fault are input to the convolutional neural network model for fault diagnosis, so as to obtain the fault location result output by the convolutional neural network model for fault diagnosis.
Compared with the traditional fault positioning method, the fault characteristics which cannot be expressed simply by using a mathematical formula can be captured by adopting the convolutional neural network, and the node voltage data of the power grid is selected as the input of the convolutional neural network, and the node air pressure data of the air grid node with high sensitivity is simultaneously selected as the input, so that the fault characteristics can be extracted by the convolutional neural network more favorably, the fault positioning accuracy is improved, the intelligent level of the electric-gas coupling comprehensive energy system is improved, and the technical problem that the fault positioning accuracy is not high in the conventional fault positioning method of the electric-gas coupling comprehensive energy system is solved.
Meanwhile, when the electric-gas coupling comprehensive energy system fault positioning method is used for configuring the synchronous vector measuring unit, the synchronous vector measuring unit of the power grid is optimally configured under the condition of complete observability of the system, so that the redundancy of sampled data is reduced, and the economy is improved.
Example 2
Another embodiment of a method for locating a fault in an electrical-to-electrical coupled integrated energy system, comprising:
step 201, according to the read state information of the gas network, a node-branch correlation matrix of the gas network is formed, and according to the read network information of the power network, a node admittance matrix of the power network is formed.
Step 202, performing energy flow calculation on the electricity-gas coupling comprehensive energy system, solving an energy flow equation set, and obtaining a current operating point and a jacobian matrix of the current operating point.
And step 203, calculating the gas pressure-gas load sensitivity and the gas boiler output-node injection power sensitivity according to the Jacobian matrix.
And step 204, calculating the gas pressure-node injection power sensitivity according to the gas pressure-gas load sensitivity and the gas boiler output-node injection power sensitivity.
And step 205, sequencing the gas pressure-node injection power sensitivity in a descending order, and taking the gas network node 30% before the gas pressure-node injection power sensitivity as a gas pressure monitoring node.
It should be noted that the sensitivity required to be obtained in the embodiment of the present application is the sensitivity of the gas pressure of the gas network node to the injection power of the power network node, and is intended to select a node with higher sensitivity as a gas pressure monitoring node. Firstly, state information of a natural gas network (namely, a gas network) is read, wherein the state information comprises natural gas pipeline parameters and pipeline topology information, a node-branch incidence matrix A of the gas network is formed, and network information of a power distribution network (namely, a power grid) is read, and a power grid node admittance matrix Y is formed. And then, performing energy flow calculation on the electricity-gas coupling comprehensive energy system, and solving an energy flow equation set to obtain a current operating point and a Jacobian matrix J of the current operating point. The energy flow equation in the embodiment of the application is as follows:
wherein, a is a node-branch incidence matrix of the gas network, Y is a network node admittance matrix, and x ═ θ, V, p]For the state variable of the electric-gas coupling comprehensive energy system, theta is the phase angle of each node of the power distribution network except for the balance node, V is the voltage of each electric load node of each node of the power distribution network except for the balance node, P is the gas pressure of each gas load node of each node of the power distribution network except for the balance node, and u is [ P ═ P [ [ P ]sp,Qsp,Lsp]For a control variable, P, of an electro-pneumatic coupled integrated energy systemspInjecting active power, Q, for a given nodespFor a given reactive power, LspFor a given gas load, f is the steady-state flow column vector of the pipeline, PMTFor the output electric power of a gas boiler, LMTFor gas consumption of gas boilers, cgeThe conversion efficiency of the gas boiler.
Iterative solution is carried out on the energy flow equation set by a Newton-Raphson method, and the specific iterative equation is as follows:
wherein, Δ F is the deviation value of the energy flow equation set, J is the Jacobian matrix, and x(k)And Δ x(k)Respectively are the state variable and the deviation value of the state variable in the kth iteration, and the superscript k represents the iteration number.
The jacobian matrix J is specifically represented as:
in the formula, the diagonal blocks are respectively the relations between the self tidal currents and the self state variables of the independent electric and gas systems, and the non-diagonal blocks are respectively the coupling relations between different energy sources.
According to the Jacobian matrix J, calculating the gas pressure-gas load sensitivity SggOutput-node injection power sensitivity S of gas boileree. The specific calculation method comprises the following steps:
for SggThere are the following calculation formulas:
for SeeAssume that node i injects power of
And the network loss of the power distribution network is PlossThen S is solved by the following formulaee：
Sensitivity S according to gas pressure-gas loadggOutput-node injection power sensitivity S of gas boilereeCalculating gas pressure-node injection power sensitivity SgeThe calculation method is as follows:
and (4) sequencing sensitivity of injection power of the gas pressure-node, and taking the gas network node with the sensitivity of the first 30% as a gas pressure monitoring node.
And step 206, taking the complete observability of the power grid system as a target, optimally configuring the synchronous vector measurement unit of the power grid, and obtaining a voltage monitoring node corresponding to the air pressure monitoring node.
It should be noted that, in consideration of the economy of the power distribution network, in the embodiment of the present application, a method for configuring a Phasor Measurement Unit (PMU) targeting complete observability of the system is adopted, and the minimum number n of PMUs is configuredpAnd the most suitable position and the maximum network structure observability and large data redundancy are achieved. The optimization problem can be expressed as:
wherein n isp,minIs the network structure to be observed satisfying the maximumMinimum number of PMU configurations, S (n)p) Is the corresponding PMU configuration location set. The method is a discontinuous high-dimensional nonlinear combined optimization problem, generally has a large number of local extreme points, and solving a global optimal solution by using a conventional optimization method is generally impossible.
And step 207, simulating a power grid line fault, and acquiring fault data of the air pressure monitoring node and the voltage monitoring node.
It should be noted that after the air pressure and voltage monitoring nodes are obtained, line faults of the power grid in different sections are simulated through simulation software, and information of the corresponding monitoring nodes is recorded.
And step 208, inputting the fault data into a convolutional neural network which takes the difference value of the node air pressure before and after the fault and the difference value of the node voltage before and after the fault as input and takes a fault section as output for training to obtain a convolutional neural network model for fault diagnosis.
The convolutional neural network adopted in the embodiment of the application includes an input layer, a convolutional layer, a pooling layer, a full link layer, and a softmax layer, the activation function adopts a modified Linear Unit (ReLU), and a Dropout method is adopted in the convolutional layer to prevent overfitting, and the model is shown in fig. 2.
An input layer:
and according to the data before and after the fault, which are uploaded by the air pressure and voltage monitoring nodes, taking the difference value of the node air pressure before and after the fault and the amplitude and phase angle of the variable quantity of the node voltage as the input of the first layer convolution layer.
And (3) rolling layers:
each convolution layer is composed of several convolution kernels WkThe matrix R is the size of the convolution kernel of the convolution neural network, the weight of the convolution kernel of each layer is shared, the complexity of the model is effectively reduced, overfitting is avoided, the generalization capability of the model can be improved, and the output of each layer is used as the convolution layer of the next layerThe input is convolved with the layer of convolution kernel as follows.
Wherein, CkE R is the output of the k-th convolutional layer, XkIs the input of the k-th convolutional layer, WkIs the convolution kernel of the kth convolutional layer, bkIs the amount of the offset that is,
is a convolution operator. For better mining the relevant features of data, the convolutional layer needs a saturation nonlinear function as an activation function, and the embodiment of the present application uses a ReLU function, whose calculation formula is as follows:
f(x)＝Max(x,0)
therefore, the output of the convolutional layer can be obtained according to the following formula:
Rk＝f(Ck)＝Max(Ck,0)
in order to further avoid over-training fitting, the Dropout method is adopted in the convolutional layer to enable the output value of the neuron to become 0 with a certain probability, the neuron does not participate in the forward propagation and backward propagation processes of the CNN, and due to the randomness of the Dropout method, the robustness of a training model can be improved.
A pooling layer:
the convolution layer is a pooling layer which plays a role in secondary feature extraction, and the generalization capability of the model is improved by reducing the number of neurons. The commonly used pooling method includes maximum pooling (maxpoloring), i.e., taking the point with the maximum value in the local domain, Mean pooling (Mean pooling), i.e., averaging and randomly pooling (stationary pooling) all the values in the local domain, and the maximum pooling (maxpoloring), i.e., performing the following operation on the output of the convolutional layer, is used in the embodiment of the present application.
Pk＝Maxpooling(Rk)
Full connection layer:
roll to roll by the Flatten operationThe output of the lamination layer is transited to a full-connection layer after being subjected to one-dimensional operation, each neuron of the full-connection layer is connected with all neurons of the previous layer, and similar to the lamination layer, a ReLU function is also selected as an activation function of the full-connection layer. The output of the last full connection layer is classified through a Softmax layer, and a classification result, namely the serial number Y of the fault line, is finally obtainedj。
And 209, verifying the accuracy of the convolutional neural network model for fault diagnosis through untrained fault data, and if the fault positioning accuracy of the convolutional neural network model for fault diagnosis is greater than preset accuracy, the verification is passed.
And step 210, positioning the fault of the electric-gas coupling comprehensive energy system by using a convolutional neural network model for fault diagnosis.
Compared with the traditional fault positioning method, the fault characteristics which cannot be expressed simply by using a mathematical formula can be captured by adopting the convolutional neural network, and the node voltage data of the power grid is selected as the input of the convolutional neural network, and the node air pressure data of the air grid node with high sensitivity is simultaneously selected as the input, so that the fault characteristics can be extracted by the convolutional neural network more favorably, the fault positioning accuracy is improved, the intelligent level of the electric-gas coupling comprehensive energy system is improved, and the technical problem that the fault positioning accuracy is not high in the conventional fault positioning method of the electric-gas coupling comprehensive energy system is solved.
Meanwhile, when the electric-gas coupling comprehensive energy system fault positioning method is used for configuring the synchronous vector measuring unit, the synchronous vector measuring unit of the power grid is optimally configured under the condition of complete observability of the system, so that the redundancy of sampled data is reduced, and the economy is improved.
Example 3
For easy understanding, please refer to fig. 3, an application example of the fault location method of the electrical-electrical coupling integrated energy system is provided in the present application, which is illustrated by taking an IEGS formed by coupling an IEEE33 node power distribution system and an 11 node gas network through an MT as an example, as shown in fig. 3.
According to the analysis method of the node sensitivity, the sensitivity of each node of the air network is calculated, the nodes with the sensitivity of the first 30% comprise nodes 8, 9 and 11, the three nodes are used as node air pressure monitoring nodes of the air network, and the node air pressure values before and after the fault are recorded.
According to the PMU optimal configuration method, PMU configuration of the power distribution network is optimized, the optimization result is displayed on nodes 1, 5, 10, 12, 19, 23 and 27, the nodes are used as node voltage monitoring nodes of the power distribution network, and node voltage values before and after a fault are recorded.
And carrying out simulation experiments, setting short-circuit faults including single-phase grounding short circuit, two-phase grounding short circuit, three-phase grounding short circuit, two-phase short circuit and three-phase short circuit in different sections of the power grid, inputting fault data obtained by simulation into the convolutional neural network provided by the invention for deep learning, and storing the trained neural network model.
And testing the trained neural network model, and after fault data which is not learned by the neural network is input into the neural network, displaying that the fault positioning accuracy can reach more than 95% as a result.
Example 4
The application provides an application example of an electricity-gas coupling comprehensive energy system fault positioning system, which comprises:
the air pressure monitoring node acquisition unit is used for analyzing the sensitivity of the node air pressure of the air network node to the power grid node injection power and acquiring the air network node with the height of 30% of the sensitivity as the air pressure monitoring node;
the voltage monitoring node acquisition unit is used for carrying out optimized configuration on a synchronous vector measurement unit of the power grid by taking the complete observability of the power grid system as a target to obtain a voltage monitoring node corresponding to the air pressure monitoring node;
the fault data acquisition unit is used for simulating a power grid line fault and acquiring fault data of the air pressure monitoring node and the voltage monitoring node;
the model training unit is used for inputting the fault data into a convolutional neural network which takes the difference value of the node air pressure before and after the fault and the difference value and the phase angle of the node voltage before and after the fault as input and takes a fault section as output to train so as to obtain a convolutional neural network model for fault diagnosis;
and the fault positioning unit is used for positioning the fault of the electric-gas coupling comprehensive energy system by using the convolutional neural network model for fault diagnosis.
In one embodiment, the air pressure monitoring node is specifically configured to:
forming a node-branch incidence matrix of the gas network according to the read state information of the gas network, and forming a power grid node admittance matrix according to the read power grid network information;
performing energy flow calculation on the electricity-gas coupling comprehensive energy system, solving an energy flow equation set, and obtaining a current operating point and a Jacobian matrix of the current operating point;
calculating gas pressure-gas load sensitivity and gas boiler output-node injection power sensitivity according to the Jacobi matrix;
calculating gas pressure-node injection power sensitivity according to the gas pressure-gas load sensitivity and the gas boiler output-node injection power sensitivity;
and sequencing the gas pressure-node injection power sensitivity in a descending order, and taking the gas network node 30% before the gas pressure-node injection power sensitivity as a gas pressure monitoring node.
In one embodiment, further comprising:
and the model verification unit is used for verifying the accuracy of the convolutional neural network model for fault diagnosis through untrained fault data, and if the fault positioning accuracy of the convolutional neural network model for fault diagnosis is greater than preset accuracy, the verification is passed.
In one embodiment, the energy flow equation is:
wherein, a is a node-branch incidence matrix of the gas network, Y is a network node admittance matrix, and x ═ θ, V, p]For the state variable of the electric-gas coupling comprehensive energy system, theta is the phase angle of each node of the power distribution network except for the balance node, V is the voltage of each electric load node of each node of the power distribution network except for the balance node, P is the gas pressure of each gas load node of each node of the power distribution network except for the balance node, and u is [ P ═ P [ [ P ]sp,Qsp,Lsp]For a control variable, P, of an electro-pneumatic coupled integrated energy systemspInjecting active power, Q, for a given nodespFor a given reactive power, LspFor a given gas load, f is the steady-state flow column vector of the pipeline, PMTFor the output electric power of a gas boiler, LMTFor gas consumption of gas boilers, cgeThe conversion efficiency of the gas boiler.
The formula for calculating the gas pressure-node injection power sensitivity is as follows:
compared with the traditional fault positioning method, the fault characteristic which cannot be expressed simply by using a mathematical formula can be captured by adopting the convolutional neural network, and the node voltage data of the power grid is selected as the input of the convolutional neural network, and the node air pressure data of the air grid node with high sensitivity is simultaneously selected as the input, so that the fault characteristic extraction of the convolutional neural network is facilitated, the fault positioning accuracy is improved, the intelligent level of the electric-gas coupling comprehensive energy system is improved, and the technical problem that the fault positioning accuracy is not high in the existing fault positioning method of the electric-gas coupling comprehensive energy system is solved.
Meanwhile, when the electric-gas coupling comprehensive energy system fault positioning method is used for configuring the synchronous vector measuring unit, the synchronous vector measuring unit of the power grid is optimally configured under the condition of complete observability of the system, so that the redundancy of sampled data is reduced, and the economy is improved.
The above embodiments are only used for illustrating the technical solutions of the present application, and not for limiting the same; although the present application has been described in detail with reference to the foregoing embodiments, it should be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions in the embodiments of the present application.
Claims (10)
1. An electric-gas coupling comprehensive energy system fault positioning method is characterized by comprising the following steps:
analyzing the sensitivity of the node air pressure of the air network node to the power grid node injection power, and acquiring the air network node with the higher sensitivity by 30% as an air pressure monitoring node;
performing optimal configuration on a synchronous vector measurement unit of a power grid by taking the complete observability of a power grid system as a target to obtain a voltage monitoring node corresponding to the air pressure monitoring node;
simulating a power grid line fault, and acquiring fault data of the air pressure monitoring node and the voltage monitoring node;
inputting the fault data into a convolutional neural network which takes the difference value of the node air pressure before and after the fault and the difference value and the phase angle of the node voltage before and after the fault as input and takes a fault section as output to train so as to obtain a convolutional neural network model for fault diagnosis;
and positioning the fault of the electric-gas coupling integrated energy system by using the convolutional neural network model for fault diagnosis.
2. The method for locating the fault of the electric-electric coupling comprehensive energy system according to claim 1, wherein the step of analyzing the sensitivity of the node air pressure of the air network node to the injected power of the power network node to obtain the air network node with the sensitivity higher than 30% as an air pressure monitoring node comprises the following steps:
forming a node-branch incidence matrix of the gas network according to the read state information of the gas network, and forming a power grid node admittance matrix according to the read power grid network information;
performing energy flow calculation on the electricity-gas coupling comprehensive energy system, solving an energy flow equation set, and obtaining a current operating point and a Jacobian matrix of the current operating point;
calculating gas pressure-gas load sensitivity and gas boiler output-node injection power sensitivity according to the Jacobi matrix;
calculating gas pressure-node injection power sensitivity according to the gas pressure-gas load sensitivity and the gas boiler output-node injection power sensitivity;
and sequencing the gas pressure-node injection power sensitivity in a descending order, and taking the gas network node 30% before the gas pressure-node injection power sensitivity as a gas pressure monitoring node.
3. The method of fault location in an electric-to-gas coupled integrated energy system of claim 2, wherein the set of energy flow equations is:
wherein, a is a node-branch incidence matrix of the gas network, Y is a network node admittance matrix, and x ═ θ, V, p]For the state variable of the electric-gas coupling comprehensive energy system, theta is the phase angle of each node of the power distribution network except for the balance node, V is the voltage of each electric load node of each node of the power distribution network except for the balance node, P is the gas pressure of each gas load node of each node of the power distribution network except for the balance node, and u is [ P ═ P [ [ P ]sp,Qsp,Lsp]For a control variable, P, of an electro-pneumatic coupled integrated energy systemspInjecting active power, Q, for a given nodespFor a given reactive power, LspFor a given gas load, f is the steady-state flow column vector of the pipeline, PMTFor the output electric power of a gas boiler, LMTFor gas consumption of gas boilers, cgeThe conversion efficiency of the gas boiler.
4. The method for fault location of an electric-gas coupling integrated energy system according to claim 3, wherein the formula for calculating the gas pressure-node injection power sensitivity is as follows:
5. the method according to claim 4, wherein the convolutional neural network comprises an input layer, a convolutional layer, a pooling layer, a full connection layer and a softmax layer, the activation function is ReLU, and a Dropout algorithm is adopted in the convolutional layer.
6. The method according to claim 1, wherein the using the convolutional neural network model for fault diagnosis to locate the fault of the electric-pneumatic coupled integrated energy system further comprises:
and verifying the accuracy of the convolutional neural network model for fault diagnosis through untrained fault data, and if the fault positioning accuracy of the convolutional neural network model for fault diagnosis is greater than preset accuracy, the verification is passed.
7. An electric-gas coupling integrated energy system fault location system, comprising:
the air pressure monitoring node acquisition unit is used for analyzing the sensitivity of the node air pressure of the air network node to the power grid node injection power and acquiring the air network node with the height of 30% of the sensitivity as the air pressure monitoring node;
the voltage monitoring node acquisition unit is used for carrying out optimized configuration on a synchronous vector measurement unit of the power grid by taking the complete observability of the power grid system as a target to obtain a voltage monitoring node corresponding to the air pressure monitoring node;
the fault data acquisition unit is used for simulating a power grid line fault and acquiring fault data of the air pressure monitoring node and the voltage monitoring node;
the model training unit is used for inputting the fault data into a convolutional neural network which takes the difference value of the node air pressure before and after the fault and the difference value and the phase angle of the node voltage before and after the fault as input and takes a fault section as output to train so as to obtain a convolutional neural network model for fault diagnosis;
and the fault positioning unit is used for positioning the fault of the electric-gas coupling comprehensive energy system by using the convolutional neural network model for fault diagnosis.
8. The electrical-to-electrical coupling integrated energy system fault location system of claim 7, wherein the air pressure monitoring node is specifically configured to:
forming a node-branch incidence matrix of the gas network according to the read state information of the gas network, and forming a power grid node admittance matrix according to the read power grid network information;
performing energy flow calculation on the electricity-gas coupling comprehensive energy system, solving an energy flow equation set, and obtaining a current operating point and a Jacobian matrix of the current operating point;
calculating gas pressure-gas load sensitivity and gas boiler output-node injection power sensitivity according to the Jacobi matrix;
calculating gas pressure-node injection power sensitivity according to the gas pressure-gas load sensitivity and the gas boiler output-node injection power sensitivity;
and sequencing the gas pressure-node injection power sensitivity in a descending order, and taking the gas network node 30% before the gas pressure-node injection power sensitivity as a gas pressure monitoring node.
9. The system according to claim 8, wherein the set of power flow equations is:
wherein, a is a node-branch incidence matrix of the gas network, Y is a network node admittance matrix, and x ═ θ, V, p]For the state variable of the electric-gas coupling comprehensive energy system, theta is the phase angle of each node of the power distribution network except for the balance node, V is the voltage of each electric load node of each node of the power distribution network except for the balance node, P is the gas pressure of each gas load node of each node of the power distribution network except for the balance node, and u is [ P ═ P [ [ P ]sp,Qsp,Lsp]For a control variable, P, of an electro-pneumatic coupled integrated energy systemspInjecting active power, Q, for a given nodespFor a given reactive power, LspFor a given gas load, f is the steady-state flow column vector of the pipeline, PMTFor the output electric power of a gas boiler, LMTFor gas consumption of gas boilers, cgeThe conversion efficiency of the gas boiler.
10. The system according to claim 7, further comprising:
and the model verification unit is used for verifying the accuracy of the convolutional neural network model for fault diagnosis through untrained fault data, and if the fault positioning accuracy of the convolutional neural network model for fault diagnosis is greater than preset accuracy, the verification is passed.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110155560.4A CN112964961B (en) | 2021-02-04 | 2021-02-04 | Electric-gas coupling comprehensive energy system fault positioning method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110155560.4A CN112964961B (en) | 2021-02-04 | 2021-02-04 | Electric-gas coupling comprehensive energy system fault positioning method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112964961A true CN112964961A (en) | 2021-06-15 |
| CN112964961B CN112964961B (en) | 2022-12-06 |
Family
ID=76273780
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110155560.4A Active CN112964961B (en) | 2021-02-04 | 2021-02-04 | Electric-gas coupling comprehensive energy system fault positioning method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112964961B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113938938A (en) * | 2021-09-27 | 2022-01-14 | 新疆天富能源股份有限公司 | Fault positioning method and device for comprehensive energy distribution network |
| CN114188944A (en) * | 2022-02-15 | 2022-03-15 | 国网江西省电力有限公司电力科学研究院 | Method and device for calculating short-circuit current of power grid with distributed power supply |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1360199A (en) * | 2002-01-24 | 2002-07-24 | 天津大学 | In-line detection system for failure in gas pipeline network |
| CN108107324A (en) * | 2017-12-22 | 2018-06-01 | 北京映翰通网络技术股份有限公司 | A kind of electrical power distribution network fault location method based on depth convolutional neural networks |
| CN108416507A (en) * | 2018-02-07 | 2018-08-17 | 天津大学 | The static sensitivity analysis method of integrated energy system is coupled towards electric-gas |
| CN108564112A (en) * | 2018-03-27 | 2018-09-21 | 湖南大学 | A kind of failure of integrated energy system unifies localization method |
| CN111398714A (en) * | 2020-04-01 | 2020-07-10 | 深圳市中电电力技术股份有限公司 | Multi-data fusion power quality fault diagnosis method and system |
| CN111769603A (en) * | 2020-07-13 | 2020-10-13 | 国网天津市电力公司 | Unit optimal scheduling method based on safety margin of electric-gas coupling system |
-
2021
- 2021-02-04 CN CN202110155560.4A patent/CN112964961B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1360199A (en) * | 2002-01-24 | 2002-07-24 | 天津大学 | In-line detection system for failure in gas pipeline network |
| CN108107324A (en) * | 2017-12-22 | 2018-06-01 | 北京映翰通网络技术股份有限公司 | A kind of electrical power distribution network fault location method based on depth convolutional neural networks |
| CN108416507A (en) * | 2018-02-07 | 2018-08-17 | 天津大学 | The static sensitivity analysis method of integrated energy system is coupled towards electric-gas |
| CN108564112A (en) * | 2018-03-27 | 2018-09-21 | 湖南大学 | A kind of failure of integrated energy system unifies localization method |
| CN111398714A (en) * | 2020-04-01 | 2020-07-10 | 深圳市中电电力技术股份有限公司 | Multi-data fusion power quality fault diagnosis method and system |
| CN111769603A (en) * | 2020-07-13 | 2020-10-13 | 国网天津市电力公司 | Unit optimal scheduling method based on safety margin of electric-gas coupling system |
Non-Patent Citations (2)
| Title |
|---|
| 林嘉伟: ""基于行波理论与卷积神经网络的配电网故障测距技术"", 《中国优秀博硕士学位论文全文数据库(硕士) 工程科技Ⅱ辑》 * |
| 骆柏锋等: ""基于统一潮流模型的电_气耦合综合能源系统静态灵敏度分析"", 《电力系统自动化》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113938938A (en) * | 2021-09-27 | 2022-01-14 | 新疆天富能源股份有限公司 | Fault positioning method and device for comprehensive energy distribution network |
| CN114188944A (en) * | 2022-02-15 | 2022-03-15 | 国网江西省电力有限公司电力科学研究院 | Method and device for calculating short-circuit current of power grid with distributed power supply |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112964961B (en) | 2022-12-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Kim et al. | State-of-health diagnosis based on hamming neural network using output voltage pattern recognition for a PEM fuel cell | |
| CN112964961B (en) | Electric-gas coupling comprehensive energy system fault positioning method and system | |
| CN108053149A (en) | A kind of photovoltaic electric station grid connection methods of risk assessment for considering meteorologic factor | |
| CN105938578A (en) | Large-scale photovoltaic power station equivalent modeling method based on clustering analysis | |
| CN109239527A (en) | Distribution network failure recognition methods based on depth confidence network | |
| Yan et al. | Deep learning based total transfer capability calculation model | |
| CN107462810A (en) | A kind of fault section location method suitable for active power distribution network | |
| CN107179688B (en) | Power system reliability analysis method considering Monte Carlo state sampling truncation | |
| Zhang et al. | A data transfer method based on one dimensional convolutional neural network for cross-building load prediction | |
| CN115469184A (en) | New energy transmission line fault identification method based on convolutional network | |
| Shi et al. | A fault location method for distribution system based on one-dimensional convolutional neural network | |
| Li et al. | Visualization of CNN transient voltage classification based on feature recognition and enhancement | |
| CN104037756B (en) | A kind of Power System Stability Assessment method containing Complex Power device model | |
| CN111179110B (en) | Virtual power plant variable order aggregation equivalent robust dynamic model modeling method and device | |
| CN114548762A (en) | Real-time power system cascading failure risk assessment method and system based on space-time diagram neural network | |
| CN113536650A (en) | Method for solving multi-target multi-energy power supply planning model through particle swarm algorithm | |
| Zhao et al. | Application of deep neural networks for fault diagnosis in a hybrid AC/DC power grid | |
| Tang et al. | Power System Inertia Estimation Based on 1-Dimensional Deep Convolution Neural Network | |
| Chen et al. | Research on Real-time Carbon Emission Factor Prediction Method for Load Nodes of Power System based on Gated Recurrent Cell Network | |
| Wang et al. | Calculation of Line Loss with Machine Learning in Distribution Network Considering Power Supply Curve and Load Distribution Characteristics | |
| Ruirong et al. | Research on Fault Location Technology of Intelligent Distribution Network based on Neural Network | |
| Bhuyan et al. | Fault Classification in a DG Connected Power System using Artificial Neural Network | |
| Li et al. | Fault Diagnosis of Microgrids Using Branch Convolution Neural Network and Majority Voting | |
| CN115935711B (en) | Multi-energy complementary active distribution network system reliability assessment method based on graph theory | |
| Glazunova et al. | Active power measurement verification for electric power systems with battery energy storage |
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 |