CN113219282B - Novel microcomputer protection constant value checking method - Google Patents
Novel microcomputer protection constant value checking method Download PDFInfo
- Publication number
- CN113219282B CN113219282B CN202110478593.2A CN202110478593A CN113219282B CN 113219282 B CN113219282 B CN 113219282B CN 202110478593 A CN202110478593 A CN 202110478593A CN 113219282 B CN113219282 B CN 113219282B
- Authority
- CN
- China
- Prior art keywords
- model
- constant value
- checking
- value
- input
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- 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
-
- 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
Abstract
The invention discloses a novel microcomputer protection constant value checking method, which comprises the steps of collecting a constant value of a microcomputer protection device, and carrying out standardized processing on the constant value to obtain model sample data; establishing a constant value check model based on a neural network, inputting model sample data to an input layer of the constant value check model, and obtaining a model weight; adjusting the model weight by designing an error function, stopping adjusting until the total error measure of the model sample data meets the requirement, and finishing the training of the constant value check model; checking the internal constant value of the microcomputer protection device by using the trained constant value checking model; the invention realizes the quick and accurate check of the microcomputer definite value by combining the neural network, on one hand, the labor cost is reduced, and on the other hand, the huge safety economic risk caused by the negligence of checking the definite value by personnel is avoided.
Description
Technical Field
The invention relates to the technical field of electric power, in particular to a novel microcomputer protection constant value checking method.
Background
The fixed value check is an important link of daily and maintenance work of the microcomputer relay protection device. After each time of finishing protection check or modifying protection fix planting, the fixed value of the latest version needs to be checked repeatedly. If the fixed value input by the microcomputer protection device deviates from the standard fixed value on the fixed value sheet, potential safety hazards are caused.
At present, the microcomputer protection setting value check mostly adopts a manual check method, and one person checks and then rechecks by a second person. Usually, a 6kV bus carries more than 20 loads, each load has hundreds of protection constants, and a generator transformer bank protection device has hundreds of protection constants. The workload of manually checking the fixed value is huge, the process is complicated, and once negligence happens, protection refusal or misoperation can be caused, so that the safe and stable operation of the unit is seriously threatened.
Disclosure of Invention
This section is for the purpose of summarizing some aspects of embodiments of the invention and to briefly introduce some preferred embodiments. In this section, as well as in the abstract and title of the application, simplifications or omissions may be made to avoid obscuring the purpose of the section, the abstract and the title, and such simplifications or omissions are not intended to limit the scope of the invention.
The present invention has been made in view of the above-mentioned conventional problems.
Therefore, the invention provides a novel microcomputer protection constant value checking method which can clearly and quickly find problem items in a plurality of constant values and finish constant value checking and modifying work.
In order to solve the technical problems, the invention provides the following technical scheme: the method comprises the steps of collecting a fixed value of a microcomputer protection device, and carrying out standardization processing on the fixed value to obtain model sample data; establishing a constant value check model based on a neural network, inputting model sample data to an input layer of the constant value check model, and obtaining a model weight; adjusting the model weight by designing an error function, stopping adjusting until the total error measure of the model sample data meets the requirement, and finishing the training of the constant value check model; and checking the internal constant value of the microcomputer protection device by using the trained constant value checking model.
As a preferable embodiment of the novel microcomputer protection constant value checking method according to the present invention, wherein: the fixed values include a motor protection fixed value, a transformer differential protection fixed value, a capacitor protection fixed value and a generator differential protection fixed value.
As a preferable embodiment of the novel microcomputer protection constant value checking method according to the present invention, wherein: the normalization process includes the steps of,
wherein i and j are sample numbers, n is the number of samples, y i For the model sample data, x i The constant value is obtained.
As a preferable embodiment of the novel microcomputer protection constant value checking method according to the present invention, wherein: the fixed value checking model is constructed by an input layer node Q, an output layer node U and a hidden layer node H, and the function expression of the fixed value checking model is as follows:
wherein Y is an output vector of the constant value check model, x is an input vector of the constant value check model,
is the weight coefficient, ω, of the ith input level node to the ith hidden level node H HY The weight coefficient from the hidden layer node to the output layer node U, and a is the applied bias.
As a preferable embodiment of the novel microcomputer protection constant value checking method according to the present invention, wherein: the model weight values include, for example,
where ω (i + 1) is the model weight at the next moment, ω (i) is the initial weight of the model, δ (i) is the learning rate, and λ is the momentum factor.
As a preferable embodiment of the novel microcomputer protection constant value checking method according to the present invention, wherein: the error function may include, for example,
where s is the input sample of the model, P is the error function, U(s) is the model output, and T is the transposed symbol.
As a preferable embodiment of the novel microcomputer protection constant value checking method according to the present invention, wherein: the total degree of error comprises the following steps,
setting the weight value adjusting conditions as follows:
E<0.05；
wherein E is the total error measure, and k is the number of model learning times.
As a preferable embodiment of the novel microcomputer protection constant value checking method according to the present invention, wherein: the input vector of the constant value checking model comprises,
wherein x is i (s) is the input vector of the s-th input, β (-) is the activation function of the hidden node, n Q Is the number of nodes of the input layer, n H For the number of hidden layer nodes, omega QH As weights, ω, of the input layer to the hidden layer HH L is the s-th input, which is the weight from hidden layer to hidden layer.
The invention has the beneficial effects that: the invention realizes the quick and accurate check of the microcomputer definite value by combining the neural network, on one hand, the labor cost is reduced, and on the other hand, the huge safety economic risk caused by the negligence of checking the definite value by personnel is avoided.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the description of the embodiments will be briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without inventive exercise. Wherein:
FIG. 1 is a schematic flow chart illustrating a method for checking a protection setting value of a microcomputer according to a first embodiment of the present invention;
FIG. 2 is a schematic diagram of a calibration value checking model of a novel microcomputer protection calibration value checking method according to a first embodiment of the present invention;
FIG. 3 is a diagram illustrating the verification result of a novel microcomputer protection constant value verification method according to a second embodiment of the present invention;
fig. 4 is a schematic diagram illustrating a verification result of a novel microcomputer protection constant value verification method according to a second embodiment of the present invention.
Detailed Description
In order to make the aforementioned objects, features and advantages of the present invention comprehensible, specific embodiments accompanied with figures are described in detail below, and it is apparent that the described embodiments are a part of the embodiments of the present invention, not all of the embodiments. All other embodiments, which can be obtained by a person skilled in the art without making any creative effort based on the embodiments in the present invention, shall fall within the protection scope of the present invention.
In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present invention, but the present invention may be practiced in other ways than those specifically described and will be readily apparent to those of ordinary skill in the art without departing from the spirit of the present invention, and therefore the present invention is not limited to the specific embodiments disclosed below.
Furthermore, reference herein to "one embodiment" or "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one implementation of the invention. The appearances of the phrase "in one embodiment" in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments.
The present invention will be described in detail with reference to the drawings, wherein the cross-sectional views illustrating the structure of the device are not enlarged partially in general scale for convenience of illustration, and the drawings are only exemplary and should not be construed as limiting the scope of the present invention. In addition, the three-dimensional dimensions of length, width and depth should be included in the actual fabrication.
Meanwhile, in the description of the present invention, it should be noted that the terms "upper, lower, inner and outer" and the like indicate orientations or positional relationships based on the orientations or positional relationships shown in the drawings, and are only for convenience of describing the present invention and simplifying the description, but do not indicate or imply that the referred device or element must have a specific orientation, be constructed in a specific orientation and operate, and thus, cannot be construed as limiting the present invention. Furthermore, the terms first, second, or third are used for descriptive purposes only and are not to be construed as indicating or implying relative importance.
The terms "mounted, connected and connected" in the present invention are to be understood broadly, unless otherwise explicitly specified or limited, for example: can be fixedly connected, detachably connected or integrally connected; they may be mechanically, electrically, or directly connected, or indirectly connected through intervening media, or may be interconnected between two elements. The specific meanings of the above terms in the present invention can be understood in specific cases to those skilled in the art.
Example 1
Referring to fig. 1 to 2, a first embodiment of the present invention provides a novel microcomputer protection fixed value checking method, including:
s1: and collecting a fixed value of the microcomputer protection device, and carrying out standardization processing on the fixed value to obtain model sample data.
The collection constant value types of the embodiment include a motor protection constant value, a transformer differential protection constant value, a capacitor protection constant value and a generator differential protection constant value.
Further, the collected fixed values are subjected to standardization processing to eliminate error data; the normalization process is performed according to the following formula:
wherein i and j are sample numbers, n is the number of samples, y i For model sample data, x i Is a constant value.
S2: and constructing a constant value check model based on the neural network, inputting model sample data to an input layer of the constant value check model, and obtaining a model weight.
A constant value check model is constructed based on a Back Propagation (BP) neural network, and referring to fig. 2, the constant value check model is composed of an input layer node Q, an output layer node U and a hidden layer node H, and through experimental verification, the learning effect of the model is best when the input layer node Q is set to be 8, the output layer node U is set to be 3 and the hidden layer node H is set to be 7;
the function of the constant value collation model is expressed as follows:
wherein Y is an output vector of the constant value check model, x is an input vector of the constant value check model,
is the weight coefficient, ω, of the ith input level node to the ith hidden level node H HY The weight coefficient from the hidden layer node to the output layer node U is shown, and a is the external deviation.
Specifically, the input vector of the constant value check model is as follows:
wherein x is i (s) is the input vector of the s-th input, β (-) is the activation function of the hidden node, n Q Is the number of nodes of the input layer, n H For the number of hidden layer nodes, omega QH As weights, ω, of the input layer to the hidden layer HH L is the s-th input, which is the weight from hidden layer to hidden layer.
The model weight is as follows:
wherein ω (i + 1) is the model weight at the next moment, ω (i) is the initial weight of the model, δ (i) is the learning rate, and λ is the momentum factor.
S3: and adjusting the model weight by designing an error function, and stopping adjusting until the total error measure of the model sample data meets the requirement, thereby finishing the training of the constant value check model.
Because feedback connection exists in the constant value check model, when training aiming at data check is carried out on the constant value check model, the calculation of an error gradient is more complex, and thus the weight estimation of the constant value check model is also complex; therefore, the difficulty of weight adjustment is reduced by designing an error function.
Designing an error function based on an output vector of the constant value check model:
where s is the input sample of the model, P is the error function, U(s) is the model output, and T is the transposed symbol.
Setting the weight value adjusting conditions as follows:
E<0.05；
and when the total error measure is less than 0.05, stopping adjusting the weight value.
Wherein the function expression of the error total measure is as follows:
wherein E is the total error measure, and k is the model learning times.
S4: and checking the internal constant value of the microcomputer protection device by using the trained constant value checking model.
Inputting the collected internal fixed value (the type of the internal fixed value is also the motor protection fixed value, the transformer differential protection fixed value, the capacitor protection fixed value and the generator differential protection fixed value) of the microcomputer protection device to an input layer of a trained fixed value check model, and obtaining a check result by an output layer.
Example 2
In order to verify and explain the technical effects adopted in the method, the embodiment selects the traditional technical scheme and adopts the method to perform comparison test, and compares the test results by means of scientific demonstration to verify the real effect of the method.
The traditional technical scheme generally adopts a manual checking mode and a software checking mode, the workload of manually checking a fixed value is huge, the process is complicated, and once negligence happens, protection refusal or misoperation can be caused, so that the safe and stable operation of a unit is seriously threatened; before each protection verification, the software verification guides the internal fixed value of each protection device into a computer through the supporting software of the protection device respectively, and the internal fixed value is filed into a txt or xls format and is consistent with the fixed value on the fixed value list; after the protection verification is finished, the fixed value of the protection device is imported into the computer in the same format; the method compares the whole contents of the two constant value files by using Beyond company software, has higher checking speed and accuracy than manual checking, but has complicated steps and unsatisfied checking effect.
In order to verify that the method has higher checking efficiency and accuracy compared with the traditional technical scheme, the traditional technical scheme and the method are adopted to respectively check and compare 1000 microcomputer setting values in real time in the embodiment.
And (3) testing environment: an Intercore i7-6500U,12G memory and a CPU main frequency of 2.5GHz; the method adopts Python as an operation platform.
Respectively checking 1000 No. 1 total variation 6kV 1B definite values (250 motor protection definite values, 250 transformer differential protection definite values, 250 capacitor protection definite values and 250 generator differential protection definite values) by adopting a traditional technical scheme (manual checking and software checking), wherein the manual checking is performed by one person and then is rechecked by a second person; the method inputs 1000 microcomputer constant values into a constant value checking model through Python to obtain results, and partial checking effects of the method are shown in the following table and are shown in figures 3 and 4.
Table 1: and (5) checking a result comparison table.
As can be seen from table 1, fig. 3 and fig. 4, compared with the conventional technical method, the method has the advantages that the fixed value checking effect is greatly improved, more than 98% of time is saved, on one hand, the labor cost is reduced, and on the other hand, the huge safety economic risk caused by negligence of personnel checking the fixed value is avoided.
It should be noted that the above-mentioned embodiments are only for illustrating the technical solutions of the present invention and not for limiting, and although the present invention has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that modifications or equivalent substitutions may be made on the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, which should be covered by the claims of the present invention.
Claims (1)
1. A novel microcomputer protection constant value checking method is characterized in that: comprises the steps of (a) preparing a mixture of a plurality of raw materials,
collecting a fixed value of a microcomputer protection device, and carrying out standardization processing on the fixed value to obtain model sample data;
constructing a constant value check model based on a neural network, inputting model sample data to an input layer of the constant value check model, and obtaining a model weight;
adjusting the model weight by designing an error function, stopping adjusting until the total error measure of the model sample data meets the requirement, and finishing the training of the constant value check model;
checking the internal constant value of the microcomputer protection device by using the trained constant value checking model;
the fixed value comprises that the fixed value comprises,
the protection constant value of the motor, the differential protection constant value of the transformer, the protection constant value of the capacitor and the differential protection constant value of the generator;
the normalization process includes the steps of,
wherein i and j are sample numbers, n is the number of samples, y i For the model sample data, x i Is the constant value;
constructing the constant-value collation model includes,
the definite value checking model is composed of an input layer node Q, an output layer node U and a hidden layer node H, and the function expression of the definite value checking model is as follows:
wherein Y is an output vector of the constant value check model, x is an input vector of the constant value check model,
is the weight coefficient, ω, of the ith input level node to the ith hidden level node H HY The weight coefficient from the hidden layer node to the output layer node U is shown, and a is the external deviation;
the model weight values include, for example,
wherein, ω (i + 1) is the model weight at the next moment, ω (i) is the initial weight of the model, δ (i) is the learning rate, and λ is the momentum factor;
the error function may include one or more of,
wherein s is an input sample of the model, P is the error function, U(s) is a model output, and T is a transposed symbol;
the total degree of error comprises the following steps,
setting the weight value adjusting conditions as follows:
E<0.05；
wherein E is the total error measure, and k is the model learning times;
the input vector of the constant value checking model comprises,
wherein x is i (s) is the input vector of the s-th input, β (-) is the activation function of the hidden node, n Q For the number of nodes of the input layer, n H For the number of hidden layer nodes, omega QH As weights, ω, of the input layer to the hidden layer HH L is the s-th input, which is the weight from hidden layer to hidden layer.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110478593.2A CN113219282B (en) | 2021-04-30 | 2021-04-30 | Novel microcomputer protection constant value checking method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110478593.2A CN113219282B (en) | 2021-04-30 | 2021-04-30 | Novel microcomputer protection constant value checking method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113219282A CN113219282A (en) | 2021-08-06 |
| CN113219282B true CN113219282B (en) | 2022-12-27 |
Family
ID=77090213
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110478593.2A Active CN113219282B (en) | 2021-04-30 | 2021-04-30 | Novel microcomputer protection constant value checking method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113219282B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113690860A (en) * | 2021-08-09 | 2021-11-23 | 国网江苏省电力有限公司 | Fixed value checking method for server |
| CN116187248B (en) * | 2023-03-13 | 2023-08-25 | 华能新能源股份有限公司河北分公司 | Relay protection fixed value analysis and verification method and system based on big data |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH0744512A (en) * | 1993-08-04 | 1995-02-14 | Meidensha Corp | Learning system of recurrent neural network of series coupling type |
| CN103714382A (en) * | 2013-12-31 | 2014-04-09 | 北京交通大学 | Multi-index comprehensive evaluation method for reliability of urban rail train security detection sensor network |
| WO2019037700A1 (en) * | 2017-08-22 | 2019-02-28 | 腾讯科技（深圳）有限公司 | Speech emotion detection method and apparatus, computer device, and storage medium |
| CN111527380A (en) * | 2018-03-28 | 2020-08-11 | 英特尔公司 | Safety enhanced computer-aided driving method and device |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101387632B (en) * | 2008-10-17 | 2011-11-16 | 北京工业大学 | Soft measurement method for biochemical oxygen demand BOD in process of sewage disposal |
| CN102854296B (en) * | 2012-08-30 | 2015-03-11 | 北京工业大学 | Sewage-disposal soft measurement method on basis of integrated neural network |
| CN108491404B (en) * | 2018-01-22 | 2022-07-01 | 国电南瑞科技股份有限公司 | State estimation bad data identification method based on BP neural network |
| CN111564817A (en) * | 2020-06-15 | 2020-08-21 | 国网山东省电力公司电力科学研究院 | Relay protection constant value comparison method and system based on file comparison tool |
-
2021
- 2021-04-30 CN CN202110478593.2A patent/CN113219282B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH0744512A (en) * | 1993-08-04 | 1995-02-14 | Meidensha Corp | Learning system of recurrent neural network of series coupling type |
| CN103714382A (en) * | 2013-12-31 | 2014-04-09 | 北京交通大学 | Multi-index comprehensive evaluation method for reliability of urban rail train security detection sensor network |
| WO2019037700A1 (en) * | 2017-08-22 | 2019-02-28 | 腾讯科技（深圳）有限公司 | Speech emotion detection method and apparatus, computer device, and storage medium |
| CN111527380A (en) * | 2018-03-28 | 2020-08-11 | 英特尔公司 | Safety enhanced computer-aided driving method and device |
Non-Patent Citations (3)
| Title |
|---|
| 《A modified particle swarm optimization-based dynamic recurrent neural network for identifying and controlling nonlinear systems》;Hong-Wei Ge等;《Computers & Structures》;20071130;全文 * |
| 《一种新型的动态递归-直觉模糊神经网络及其学习算法》;殷宏燕等;《微电子学与计算机》;20090630;全文 * |
| 《基于动态递归反馈型神经网络的永磁同步电机转矩观测器设计》;闫俞佰等;《集成技术》;20200930;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113219282A (en) | 2021-08-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113219282B (en) | Novel microcomputer protection constant value checking method | |
| White et al. | Monte Carlo calculation of dynamical properties of the two-dimensional Hubbard model | |
| Benincasa et al. | Sound waves in strongly coupled non-conformal gauge theory plasma | |
| Tesauro et al. | Climbing of dislocations in nonequilibrium patterns | |
| Vrard et al. | Amplitude and lifetime of radial modes in red giant star spectra observed by Kepler | |
| Horoi et al. | Chaotic wave functions and exponential convergence of low-lying energy eigenvalues | |
| CN107356978B (en) | Boat magnetic compensation method based on principal component analysis | |
| CN106850254A (en) | Key node recognition methods in a kind of power telecom network | |
| Racz et al. | Numerical investigation of the late-time Kerr tails | |
| CN112115648A (en) | Transformer top layer oil temperature prediction method based on improved deep learning method | |
| Shestopalov et al. | On a rigorous proof of the existence of complex waves in a dielectric waveguide of circular cross section | |
| CN110458616A (en) | A kind of finance product recommended method based on GAMxNN model | |
| Li et al. | Comparative study of partial least squares and neural network models of near-infrared spectroscopy for aging condition assessment of insulating paper | |
| CN109918720A (en) | Diagnosis Method of Transformer Faults based on krill group's Support Vector Machines Optimized | |
| CN113408200A (en) | Aviation equipment vibration environment analysis and prediction method | |
| Duncan et al. | Multiway contingency analysis with a scaled response or factor | |
| CN114583767B (en) | Data-driven wind power plant frequency modulation response characteristic modeling method and system | |
| Baltianski et al. | Analysis of impedance spectroscopy data—Finding the best system function | |
| Bartel et al. | Creep and creep recovery response of load cells tested according to US and international evaluation procedures | |
| Kurdyukov et al. | Homotopy method for solving anisotropy-based stochastic H∞-optimization problem with uncertainty | |
| CN112884312A (en) | Chemical device risk assessment method based on OWA operator empowerment | |
| Soleimanian et al. | Theoretical, numerical, and experimental analyses of free vibrations of glass fiber reinforced polymer plates with central cutouts and free boundaries | |
| CN113704990B (en) | Electromagnetic transient model single-parameter online correction method for direct-current measurement data | |
| Zhu et al. | EM-aided fast posterior covariance computation in Bayesian FFT method | |
| Carlucci et al. | Addressing load sensitivity of rational macromodels |
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 |