CN117891160B - Intelligent control system and method for switch cabinet - Google Patents
Intelligent control system and method for switch cabinet Download PDFInfo
- Publication number
- CN117891160B CN117891160B CN202410285435.9A CN202410285435A CN117891160B CN 117891160 B CN117891160 B CN 117891160B CN 202410285435 A CN202410285435 A CN 202410285435A CN 117891160 B CN117891160 B CN 117891160B
- Authority
- CN
- China
- Prior art keywords
- sliding window
- operation data
- data
- center
- noise
- 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
- 238000000034 method Methods 0.000 title claims abstract description 64
- 238000003062 neural network model Methods 0.000 claims abstract description 41
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 27
- 230000008569 process Effects 0.000 claims description 22
- 238000004364 calculation method Methods 0.000 claims description 15
- 238000012937 correction Methods 0.000 claims description 12
- 230000006870 function Effects 0.000 claims description 8
- 238000004590 computer program Methods 0.000 claims description 5
- 230000004913 activation Effects 0.000 claims description 3
- 210000002569 neuron Anatomy 0.000 claims description 3
- 238000012545 processing Methods 0.000 abstract description 2
- 230000008859 change Effects 0.000 description 15
- 238000013528 artificial neural network Methods 0.000 description 14
- 230000001276 controlling effect Effects 0.000 description 9
- 238000012549 training Methods 0.000 description 9
- 238000005457 optimization Methods 0.000 description 4
- 230000007246 mechanism Effects 0.000 description 3
- 238000004891 communication Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000003068 static effect Effects 0.000 description 2
- PCTMTFRHKVHKIS-BMFZQQSSSA-N (1s,3r,4e,6e,8e,10e,12e,14e,16e,18s,19r,20r,21s,25r,27r,30r,31r,33s,35r,37s,38r)-3-[(2r,3s,4s,5s,6r)-4-amino-3,5-dihydroxy-6-methyloxan-2-yl]oxy-19,25,27,30,31,33,35,37-octahydroxy-18,20,21-trimethyl-23-oxo-22,39-dioxabicyclo[33.3.1]nonatriaconta-4,6,8,10 Chemical compound C1C=C2C[C@@H](OS(O)(=O)=O)CC[C@]2(C)[C@@H]2[C@@H]1[C@@H]1CC[C@H]([C@H](C)CCCC(C)C)[C@@]1(C)CC2.O[C@H]1[C@@H](N)[C@H](O)[C@@H](C)O[C@H]1O[C@H]1/C=C/C=C/C=C/C=C/C=C/C=C/C=C/[C@H](C)[C@@H](O)[C@@H](C)[C@H](C)OC(=O)C[C@H](O)C[C@H](O)CC[C@@H](O)[C@H](O)C[C@H](O)C[C@](O)(C[C@H](O)[C@H]2C(O)=O)O[C@H]2C1 PCTMTFRHKVHKIS-BMFZQQSSSA-N 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000013136 deep learning model Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000001788 irregular Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000035772 mutation Effects 0.000 description 1
- 238000010248 power generation Methods 0.000 description 1
- 230000001105 regulatory effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B11/00—Automatic controllers
- G05B11/01—Automatic controllers electric
- G05B11/36—Automatic controllers electric with provision for obtaining particular characteristics, e.g. proportional, integral, differential
- G05B11/42—Automatic controllers electric with provision for obtaining particular characteristics, e.g. proportional, integral, differential for obtaining a characteristic which is both proportional and time-dependent, e.g. P. I., P. I. D.
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02B—BOARDS, SUBSTATIONS OR SWITCHING ARRANGEMENTS FOR THE SUPPLY OR DISTRIBUTION OF ELECTRIC POWER
- H02B1/00—Frameworks, boards, panels, desks, casings; Details of substations or switching arrangements
- H02B1/26—Casings; Parts thereof or accessories therefor
- H02B1/30—Cabinet-type casings; Parts thereof or accessories therefor
Landscapes
- Engineering & Computer Science (AREA)
- Power Engineering (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Feedback Control In General (AREA)
Abstract
The invention relates to the technical field of data processing. In particular to an intelligent control system and method for a switch cabinet. The method comprises the following steps: collecting operation data of the intelligent switch cabinet in real time, and setting a sliding window, wherein the sliding window slides rightwards once every time the operation data is collected; calculating the quality of the operation data of the sliding window center according to all the operation data in the sliding window, and adjusting the learning rate of the BP neural network model according to the quality of the operation data of the sliding window center so as to optimize the BP neural network model; and optimizing a PID algorithm by using the optimized BP neural network model, and controlling the operation data of the intelligent switch cabinet by using the optimized PID algorithm. The method can control the operation data of the switch cabinet more accurately, and can improve the self-adaptability and the robustness of the intelligent control system of the switch cabinet.
Description
Technical Field
The present invention relates generally to the field of data processing technology. More particularly, the invention relates to an intelligent control system and method for a switch cabinet.
Background
The switch cabinet is an electrical equipment and control system, is mainly used in the power generation, transmission, distribution and electric energy conversion processes of a power system, and is used as an important component of the power system, and the accuracy and reliability of current distribution and control of the switch cabinet directly influence the safe and stable operation of the power system. PID (proportional-integral-derivative) control algorithms are one of the most widely used feedback control methods. The PID algorithm is simple and effective, and is regulated、/>And/>Three parameters control the system output to achieve the desired control effect. However, conventional PID control algorithms, when faced with complex or nonlinear systems, often require experience and trial and error for parameter adjustment, which may be inefficient and inaccurate in some cases. With the development of artificial intelligence technology, especially the application of neural networks, new possibilities are provided for solving the limitations of traditional PID control in complex systems. BP (back propagation) neural networks, as a classical deep learning model, are able to optimize parameters by learning complex relationships between inputs and outputs. Application of BP neural network to optimization of PID control parameters, i.e., learning and tuning of PID controller/>, through neural network、/>And/>The three parameters can realize more intelligent and accurate control.
The problem of the existing BP neural network is that, under the intelligent control scene of a switch cabinet touch screen, the electromagnetic interference in the surrounding environment is more, which causes a great amount of noise in the operation data acquired by the sensor, the noise also has randomness and uncertainty, and the noise affects the selection of the learning rate of the BP neural network in the back propagation, because the quality of the input operation data is uncertain, the fixed learning rate may not be suitable for all current data, thereby causing the obtained PID controller to have the following characteristics that、/>And/>The accuracy of the three parameters is insufficient, so that the accuracy of operation data control can be reduced, unstable and unreliable control of the switch cabinet is further caused, and the control precision and efficiency of the intelligent control system of the touch screen of the switch cabinet are reduced.
Disclosure of Invention
To solve one or more of the above-described technical problems, the present invention provides aspects as follows.
In a first aspect, the present invention provides a method for intelligently controlling a switch cabinet, including:
collecting operation data of the intelligent switch cabinet in real time, presetting a sliding window before collecting, and sliding the sliding window rightwards once when the operation data is collected once;
Calculating the quality of the operation data of the sliding window center according to all the operation data in the sliding window, and adjusting the learning rate of the BP neural network model according to the quality of the operation data of the sliding window center so as to optimize the BP neural network model, wherein the learning rate of the BP neural network model is inversely proportional to the quality of the operation data; the BP neural network model is used for optimizing parameters of a PID algorithm; the PID algorithm is used for controlling the operation data of the intelligent switch cabinet; the quality of the operation data is used for representing the approaching degree of the collected operation data and the actual value of the operation data;
proportional adjustment coefficient of PID algorithm by using optimized BP neural network model Integral adjustment coefficient/>And differential adjustment coefficient/>And optimizing the three parameters, and controlling the operation data of the intelligent switch cabinet by utilizing an optimized PID algorithm.
In one embodiment, the calculating the quality of the operational data of the sliding window center includes:
calculating the noise expression degree of the operation data in the center of the sliding window;
Correcting the noise expression degree of the operation data in the center of the sliding window according to the variance of the absolute operation data in the corresponding time range of the sliding window, wherein the larger the variance is, the smaller the noise expression degree is after correction; the absolute operation data refers to operation data set values on a touch screen of the switch cabinet;
And calculating the quality of the operation data in the center of the sliding window according to the corrected noise expression degree.
In one embodiment, calculating the noise performance level of the operational data of the sliding window center includes:
Calculating the noise expression degree of the operation data of the center of the sliding window in the sliding window according to all the operation data in the sliding window, and recording the noise expression degree as the first noise expression degree of the operation data of the center of the sliding window;
calculating the average value of all operation data in the sliding window;
Removing the data point with the maximum deviation degree from the mean value in the sliding window, calculating the noise expression degree of the operation data of the center of the sliding window in the sliding window according to the residual operation data in the sliding window, and recording the noise expression degree as the second noise expression degree of the operation data of the center of the sliding window;
Calculating the noise expression degree of the operation data of the sliding window center according to the first noise expression degree and the second noise expression degree of the operation data of the sliding window center; the calculation expression is as follows:
In the method, in the process of the invention, Noise expression level of running data representing the center of sliding window,/>First noise expression level of operation data representing the center of a sliding window,/>A second noise performance level representing the operational data at the center of the sliding window.
In one embodiment, calculating the first noise performance level of the operational data of the sliding window center includes:
Fitting all data points in the sliding window to obtain a fitting curve;
Calculating a first noise expression degree of operation data in the center of the sliding window according to the fitting curve and the numerical values of all data points in the sliding window; the calculation expression is as follows:
In the method, in the process of the invention, First noise expression level of operation data representing the center of a sliding window,/>Representing the number of data points within a sliding window,/>Representing the integral of the j-th set of adjacent data points corresponding to the fitted curve,/>Representing the j-th operational data point within the sliding window,/>Then the value representing the j-th operational data point,/>Representing the j+1th operating data point within the sliding window,/>Representing the value of the j+1th operating data point in the sliding window,/>Representing trapezoid area/>, of the original data corresponding to the j-th set of adjacent data pointsRepresenting the loss of two adjacent data points during curve fitting within a sliding window.
In one embodiment, the modified noise performance level calculation expression is as follows:
In the method, in the process of the invention, And/>Respectively representing the ith operating data point/>The noise performance degree before and after correction; /(I)Representing the variance of absolute operational data points over a time range corresponding to the sliding window.
In one embodiment, the quality calculation expression of the operation data of the sliding window center is:
，
Wherein the method comprises the steps of Representing the ith operational data point/>Quality of/(I)For the ith operational data point/>Is included in the noise performance level after correction.
In one embodiment, adjusting the learning rate of the BP neural network model comprises:
According to the quality of the running data in the center of the sliding window, the learning rate of the BP neural network model is calculated, and the calculation expression is as follows:
In the method, in the process of the invention, Represent learning rate of BP neural network model,/>Representing the quality of the running data in the center of the sliding window.
In one embodiment, the number of hidden layers of the BP neural network model is 3, the activation function adopts a ReLU function, the number of neurons of an output layer is 3, the back propagation process adopts a gradient descent algorithm, and the learning rate initial value is 0.5.
In a second aspect, the present invention provides a switchgear intelligent control system, comprising a processor and a memory, the memory storing computer program instructions, which when executed by the processor, implement the switchgear intelligent control method of the present invention.
The invention has the technical effects that: according to the intelligent control method for the switch cabinet, the learning rate is adjusted in a self-adaptive mode by judging the quality of the operation data points, and a smaller learning rate is set for the operation data points with large quality, so that an accurate model training result is ensured; setting a larger learning rate for the operation data points with small quality, and adaptively adjusting the learning rate by analyzing the quality of the operation data used for BP neural network training, wherein compared with the learning rate with a fixed size of a traditional BP neural network, the learning rate can ensure that the optimal solution is sought and the convergence speed is increased in the training process, so that the operation data of the switch cabinet is controlled more accurately; furthermore, the BP neural network can learn a nonlinear relation, the traditional PID controller is linear, and the complex nonlinear relation between current data and PID parameters can be better captured through the neural network, so that the self-adaptability, the robustness, the stability and the reliability of the intelligent control system of the switch cabinet are improved; in addition, as the operation data set value of the switch cabinet has no noise, the operation data set value is absolute operation data, and the collected operation data is relative operation data. If the absolute operating data point has not changed in the corresponding time range, the absolute reliability of the noise performance degree of the relative operating data point at the moment is described, because in the case, the factor causing the change of the relative operating data is certainly caused by noise; whereas a change in the absolute operating data point in the corresponding time range indicates that the change in the relative operating data may be due to a change in the operating data setting, the lower the noise performance level is. Therefore, the noise expression degree of the operation data in the center of the sliding window is corrected according to the variance of the absolute operation data in the corresponding time range of the sliding window, so that the calculated noise expression degree is more accurate, the quality of the calculated operation data point is more accurate, and the accuracy, stability and reliability of the control of the switch cabinet are further improved.
Drawings
The above, as well as additional purposes, features, and advantages of exemplary embodiments of the present invention will become readily apparent from the following detailed description when read in conjunction with the accompanying drawings. In the drawings, embodiments of the invention are illustrated by way of example and not by way of limitation, and like reference numerals refer to similar or corresponding parts and in which:
FIG. 1 is a flow chart schematically illustrating a method of intelligent control of a switchgear according to an embodiment of the present invention;
FIG. 2 is a flow chart that schematically illustrates a quality method of calculating operational data for a sliding window center, in accordance with an embodiment of the present invention;
FIG. 3 is a flow chart schematically illustrating a method of calculating the noise performance level of operational data of a sliding window center according to an embodiment of the present invention;
FIG. 4 is a flow chart schematically illustrating a first noise performance level method of calculating operational data for a sliding window center according to an embodiment of the present invention;
FIG. 5 is a schematic diagram schematically illustrating fitting errors for an embodiment of the present invention;
fig. 6 is a schematic diagram schematically illustrating the structure of an intelligent control system for a switchgear according to an embodiment of the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and fully with reference to the accompanying drawings, in which it is evident that the embodiments described are some, but not all embodiments of the invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.
The embodiment of the intelligent control method of the switch cabinet comprises the following steps:
As shown in fig. 1, the intelligent control method of the switch cabinet of the invention comprises the following steps:
s101, acquiring operation data of an intelligent switch cabinet in real time, wherein a sliding window is preset before acquisition, and the sliding window slides rightwards once when the operation data are acquired once;
The collected operation data of the intelligent switch cabinet can be operation parameters such as current, voltage or power of the intelligent switch cabinet. If the collected operation data is the current of the intelligent switch cabinet, a current sensor can be adopted; if the collected operation data is the voltage of the intelligent switch cabinet, a voltage sensor can be adopted. The collection frequency of the operation data of the intelligent switch cabinet is 10HZ.
The length of the sliding window is an odd number of data points, and preferably, the length of the sliding window is set to 11 data points in this embodiment.
S102, optimizing a BP neural network model, wherein the BP neural network model specifically comprises the following steps: calculating the quality of the operation data of the sliding window center according to all the operation data in the sliding window, and adjusting the learning rate of the BP neural network model according to the quality of the operation data of the sliding window center so as to optimize the BP neural network model, wherein the learning rate of the BP neural network model is inversely proportional to the quality of the operation data; the BP neural network model is used for optimizing parameters of a PID algorithm; the PID algorithm is used for controlling the operation data of the intelligent switch cabinet;
The BP neural network model is input into an operation data acquisition value of the intelligent switch cabinet, an operation data setting value of the intelligent switch cabinet and a proportion adjustment coefficient of a PID algorithm Initial value of (1)/(integral adjustment coefficient)Initial value and differential adjustment coefficient/>Is set to be a constant value. The number of hidden layers of the BP neural network model is 3, the activation function adopts a ReLU function, the number of neurons of an output layer is 3, and the back propagation process adopts a gradient descent algorithm. The initial value of the learning rate of the BP neural network model may be set to 0.5.
The quality of the collected operation data refers to the authenticity of the operation data, i.e. the proximity of the collected operation data to the actual operation data of the switchgear. In general, the authenticity of the collected operation data is mainly affected by noise, and the greater the noise is, the worse the authenticity is, so the quality of the collected operation data can be calculated based on the noise expression degree of the operation data.
Because the data acquisition device can receive electromagnetic interference in the intelligent switch cabinet scene and lead to the noise to appear in the operation data who gathers, when carrying out BP neural network model training simultaneously, the slight change of learning rate can all arouse the stability and the accuracy of model, and because the difference of operation data quality, unified learning rate is different to the accuracy of the operation data training of different qualities to influence the accuracy of the BP neural network model that follows to train, lead to the accuracy of the parameter of its PID algorithm of output to worsen, and then lead to the accuracy to switch cabinet operation parameter control to worsen. The learning rate of the BP neural network model is adaptively adjusted through analyzing the quality of the operation data, so that the accuracy of training the BP neural network model is improved while the stability of the model is ensured.
S103, optimizing a PID algorithm by using the optimized BP neural network model, and controlling operation data of the intelligent switch cabinet by using the optimized PID algorithm, wherein the operation data specifically comprises the following steps: proportional adjustment coefficient of PID algorithm by using optimized BP neural network modelIntegral adjustment coefficient/>And differential adjustment coefficient/>And optimizing the three parameters, and controlling the operation data of the intelligent switch cabinet by utilizing an optimized PID algorithm.
When the intelligent control method of the switch cabinet is used for controlling the switch cabinet, the PID algorithm before the primary optimization is used for controlling the operation data of the intelligent switch cabinet, and the BP neural network model before the primary optimization is used for optimizing the parameters of the PID algorithm.
Before the first optimization, the PID algorithm is provided with a proportional adjustment coefficientIntegral adjustment coefficient/>And differential adjustment coefficient/>Wherein the scaling factor/>The initial value of the integral adjustment coefficient/>, can be set to 0.2The initial value of the differential adjustment coefficient/>, can be set to 0.01The initial value of (2) may be set to 0.2.
According to the intelligent control method for the switch cabinet, the learning rate is adjusted in a self-adaptive mode by judging the quality of the operation data points, and a smaller learning rate is set for the operation data points with large quality, so that an accurate model training result is ensured; setting a larger learning rate for the operation data points with small quality, and adaptively adjusting the learning rate by analyzing the quality of the operation data used for BP neural network training, wherein compared with the learning rate with a fixed size of a traditional BP neural network, the learning rate can ensure that the optimal solution is sought and the convergence speed is increased in the training process, so that the control of the operation data of the switch cabinet is more accurate; in addition, the BP neural network can learn a nonlinear relation, the traditional PID controller is linear, and the complex nonlinear relation between current data and PID parameters can be better captured through the neural network, so that the self-adaptability and the robustness of the intelligent control system of the switch cabinet are improved.
In one embodiment, as shown in fig. 2, the calculating the quality of the operation data of the sliding window center includes:
S201, calculating the noise expression degree of the operation data of the sliding window center;
the noise performance level may be calculated from the fit loss resulting from curve fitting all data points within the sliding window.
If the running data in the center of the sliding window is noise, the value which greatly influences the noise expression degree is necessarily removed from the data points which are most deviated from the mean value of all the running data in the sliding window, otherwise, if the data points are normal data points, the value of the noise expression degree is not greatly changed by removing one data point, so that more accurate noise expression degree can be calculated through the difference between the noise expression degrees before and after the data is removed, and the larger the difference is, the higher the noise expression degree is.
S202, correcting noise expression degree of operation data in the center of a sliding window, specifically: correcting the noise expression degree of the operation data in the center of the sliding window according to the variance of the absolute operation data in the corresponding time range of the sliding window, wherein the larger the variance is, the smaller the noise expression degree is after correction; the absolute operation data refers to operation data set values on a touch screen of the switch cabinet.
Because the operation data of the switch cabinet can be manually set on the touch screen of the switch cabinet, the operation data set value has no noise, the operation data set value can be recorded as absolute operation data, and the collected operation data can be recorded as relative operation data. Because the change of the set value of the operation data can cause the change of the collected operation data, the reliability of the noise expression degree can be calculated according to the change of the absolute operation data of the relative operation data point in the corresponding time range of the sliding window of the relative operation data point, and the noise expression degree of the relative operation data point is corrected. If the absolute operating data point has not changed in the corresponding time range, the absolute reliability of the noise performance degree of the relative operating data point at the moment is described, because in the case, the factor causing the change of the relative operating data is certainly caused by noise; whereas a change in the absolute operating data point in the corresponding time range indicates that the change in the relative operating data may be due to a change in the operating data setting, the lower the noise performance level is.
Thus, the variance of the absolute operating data over the corresponding time range may characterize the confidence in the noise performance level of the operating data at the center of the sliding window, the greater the variance, the more severe the data change over the local range that the absolute operating data point is, the lower the confidence in the noise performance level of the relative operating data point; conversely, the smaller the variance, the slower the data change over the local range of the absolute operational data point, and the higher the confidence in the noise performance level of the relative operational data point.
S203, calculating the quality of the operation data in the center of the sliding window according to the corrected noise expression degree.
The higher the noise performance level of a data point, the greater the possibility that the data point has noise, and therefore, the quality of the running data in the center of the sliding window is inversely related to the noise performance level after correction.
As can be seen from the above embodiments, the more accurate noise performance level can be calculated by removing the difference between the noise performance levels before and after the data, as shown in fig. 3, in one embodiment, calculating the noise performance level of the running data in the center of the sliding window includes:
S301, calculating a first noise expression degree of operation data of the sliding window center, wherein the first noise expression degree is specifically: calculating the noise expression degree of the operation data of the center of the sliding window in the sliding window according to all the operation data in the sliding window, and recording the noise expression degree as the first noise expression degree of the operation data of the center of the sliding window;
Because the noise has randomness, that is, the noise is represented in the data in a diversity manner, which may be the mutation of individual data points, the irregular variation of the frequency and the amplitude in a local range, and the like, the variation of the data can cause that all the data points are fitted on a fitting curve, that is, the loss in the fitting process is larger in the process of fitting the data, so that the loss in curve fitting of the data points in the sliding window can be used for representing the noise representation degree of the running data in the center of the sliding window in the sliding window, and the larger the loss is, the higher the noise representation degree is.
S302, calculating the average value of all operation data in the sliding window;
S303, calculating a second noise expression degree of the operation data of the sliding window center, wherein the second noise expression degree is specifically: removing the data point with the maximum deviation degree from the mean value in the sliding window, calculating the noise expression degree of the operation data of the center of the sliding window in the sliding window according to the residual operation data in the sliding window, and recording the noise expression degree as the second noise expression degree of the operation data of the center of the sliding window;
in order to improve the reliability of noise expression level calculation, the method removes the point which is most deviated from the average value of the data points in the sliding window, then calculates the noise expression level again in the sliding window after the data points are removed, and calculates the noise expression level according to the difference value of the noise expression levels before and after the data points are removed.
S304, calculating the noise expression degree of the operation data of the sliding window center according to the first noise expression degree and the second noise expression degree of the operation data of the sliding window center; the calculation expression is as follows:
（1）
In the method, in the process of the invention, Noise expression level of running data representing the center of sliding window,/>First noise expression level of operation data representing the center of a sliding window,/>A second noise performance level representing the operational data at the center of the sliding window.
As can be seen from the above embodiments, the larger the variance is, the smaller the noise performance degree after correction is, and in one embodiment, the noise performance degree after correction is calculated as follows:
（2）
In the method, in the process of the invention, And/>Respectively representing the ith operating data point/>The noise performance degree before and after correction; exp () is an exponential function; /(I)Representing the variance of absolute operational data points over a time range corresponding to the sliding window.
As can be seen from the above embodiments, the quality of the operation data at the center of the sliding window is inversely related to the modified noise performance level, and in one embodiment, the quality calculation expression of the operation data at the center of the sliding window is:
（3）
Wherein the method comprises the steps of Representing the ith operational data point/>Quality of/(I)For the ith operational data point/>Is included in the noise performance level after correction.
The first noise performance level of the operation data of the sliding window center is the same as the second noise performance level calculating method, as shown in fig. 4, and in one embodiment, calculating the first noise performance level of the operation data of the sliding window center includes:
S401, fitting all data points in a sliding window to obtain a fitting curve;
and fitting all data points in the sliding window by adopting a least square method to obtain a fitting curve.
S402, calculating a first noise expression degree of operation data in the center of the sliding window according to the fitting curve and the numerical values of all data points in the sliding window; the calculation expression is as follows:
（4）
In the method, in the process of the invention, First noise expression level of operation data representing the center of a sliding window,/>Representing the number of data points within a sliding window,/>Representing the integral of the j-th group of adjacent data points corresponding to the fitted curve, wherein the physical meaning of the integral represents the curve and the abscissa axis and the corresponding x=j, x=j+1 enclose a curved surface trapezoidal area,/>Representing the j-th operational data point within the sliding window,/>Then the value representing the j-th operational data point,/>Representing the j +1 th operational data point within the sliding window,Representing the value of the j+1th operating data point in the sliding window,/>Representing trapezoid area/>, of the original data corresponding to the j-th set of adjacent data pointsRepresenting the loss of two adjacent data points during curve fitting within a sliding window.
As shown in fig. 5, the loss in the curve fitting process can be represented by the difference between the integral between two adjacent points in the fitted curve and the trapezoidal area corresponding to the two adjacent points, wherein C and D are two adjacent points in the figure, the trapezoidal area is the area of the figure ABCD, the integral is the area of the curved trapezoid ABDE, and the difference between the two areas can represent the loss in the fitting process. Thus (2)Representing the loss of two adjacent data points during curve fitting within a sliding window,/>The average of the fit loss for all adjacent data points is represented and can be used to characterize the noise performance level of the running data at the center of the sliding window within the sliding window.
In one embodiment, adjusting the learning rate of the BP neural network model comprises:
According to the quality of the running data in the center of the sliding window, the learning rate of the BP neural network model is calculated, and the calculation expression is as follows:
（5）
In the method, in the process of the invention, Represent learning rate of BP neural network model,/>Representing the quality of the running data in the center of the sliding window.
The learning of the BP neural network model is calculated by adopting the expression, and the learning rate can be ensured to be less than 0.5, so that the divergence problem caused by overlarge learning rate is avoided, the large-amplitude updating of the weight is reduced, and the model is more stable.
Switch cabinet intelligent control system embodiment:
the invention also provides an intelligent control system of the switch cabinet. As shown in fig. 6, the intelligent control system for a switchgear comprises a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the intelligent control method for a switchgear according to the first aspect of the invention is realized; the intelligent control system of the switch cabinet further comprises a touch screen arranged on the outer surface of the switch cabinet, the switch cabinet for controlling the intelligent control system of the switch cabinet comprises a supporting plate arranged in the cabinet, a sliding mechanism is further arranged at the bottom of the supporting plate, the intelligent control system of the switch cabinet is controlled and connected to the sliding mechanism, various electronic elements are arranged on the supporting plate, and after a specific key on the touch screen is clicked, the intelligent control system of the switch cabinet controls the sliding mechanism to act, so that the supporting plate is automatically pulled out from the cabinet.
The intelligent control system of the switch cabinet further comprises a communication bus, a communication interface and other components well known to those skilled in the art, and the arrangement and the function of the intelligent control system are known in the art, so that the intelligent control system is not repeated herein.
In the context of this patent, the foregoing memory may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. For example, the computer-readable storage medium may be any suitable magnetic or magneto-optical storage medium, such as, for example, resistance change Memory RRAM (Resistive Random Access Memory), dynamic Random Access Memory DRAM (Dynamic Random Access Memory), static Random Access Memory SRAM (Static Random-Access Memory), enhanced dynamic Random Access Memory EDRAM (ENHANCED DYNAMIC Random Access Memory), high-Bandwidth Memory HBM (High-Bandwidth Memory), hybrid storage cube HMC (Hybrid Memory Cube), or the like, or any other medium that may be used to store the desired information and that may be accessed by an application, a module, or both. Any such computer storage media may be part of, or accessible by, or connectable to, the device. Any of the applications or modules described herein may be implemented using computer-readable/executable instructions that may be stored or otherwise maintained by such computer-readable media.
In the description of the present specification, the meaning of "a plurality", "a number" or "a plurality" is at least two, for example, two, three or more, etc., unless explicitly defined otherwise.
While various embodiments of the present invention have been shown and described herein, it will be obvious to those skilled in the art that such embodiments are provided by way of example only. Many modifications, changes, and substitutions will now occur to those skilled in the art without departing from the spirit and scope of the invention. It should be understood that various alternatives to the embodiments of the invention described herein may be employed in practicing the invention.
Claims (5)
1. The intelligent control method for the switch cabinet is characterized by comprising the following steps of:
collecting operation data of the intelligent switch cabinet in real time, presetting a sliding window before collecting, and sliding the sliding window rightwards once when the operation data is collected once;
calculating the quality of the operation data of the sliding window center according to all the operation data in the sliding window, and adjusting the learning rate of the BP neural network model according to the quality of the operation data of the sliding window center so as to optimize the BP neural network model, wherein the learning rate of the BP neural network model is inversely proportional to the quality of the operation data; the BP neural network model is used for optimizing parameters of a PID algorithm; the PID algorithm is used for controlling the operation data of the intelligent switch cabinet; the quality of the operation data is used for representing the approaching degree of the collected operation data and the actual value of the operation data; the calculating the quality of the operation data of the sliding window center comprises the following steps: calculating the noise expression degree of the operation data in the center of the sliding window; correcting the noise expression degree of the operation data in the center of the sliding window according to the variance of the absolute operation data in the corresponding time range of the sliding window, wherein the larger the variance is, the smaller the noise expression degree is after correction; the absolute operation data refers to operation data set values on a touch screen of the switch cabinet; calculating the quality of the operation data in the center of the sliding window according to the corrected noise expression degree; adjusting the learning rate of the BP neural network model includes: according to the quality of the running data in the center of the sliding window, the learning rate of the BP neural network model is calculated, and the calculation expression is as follows:
；
In the method, in the process of the invention, Represent learning rate of BP neural network model,/>Representing the quality of the operational data at the center of the sliding window;
proportional adjustment coefficient of PID algorithm by using optimized BP neural network model Integral adjustment coefficient/>And differential adjustment coefficient/>Optimizing the three parameters, and controlling the operation data of the intelligent switch cabinet by utilizing an optimized PID algorithm;
The calculating the noise expression degree of the operation data of the sliding window center comprises the following steps:
Calculating the noise expression degree of the operation data of the center of the sliding window in the sliding window according to all the operation data in the sliding window, and recording the noise expression degree as the first noise expression degree of the operation data of the center of the sliding window;
calculating the average value of all operation data in the sliding window;
Removing the data point with the maximum deviation degree from the mean value in the sliding window, calculating the noise expression degree of the operation data of the center of the sliding window in the sliding window according to the residual operation data in the sliding window, and recording the noise expression degree as the second noise expression degree of the operation data of the center of the sliding window;
Calculating the noise expression degree of the operation data of the sliding window center according to the first noise expression degree and the second noise expression degree of the operation data of the sliding window center; the calculation expression is as follows:
；
In the method, in the process of the invention, Noise expression level of running data representing the center of sliding window,/>First noise expression level of operation data representing the center of a sliding window,/>A second noise expression level representing the operation data of the center of the sliding window;
calculating a first noise performance level of the operational data of the sliding window center includes:
Fitting all data points in the sliding window to obtain a fitting curve;
Calculating a first noise expression degree of operation data in the center of the sliding window according to the fitting curve and the numerical values of all data points in the sliding window; the calculation expression is as follows:
；
In the method, in the process of the invention, Representing the number of data points within a sliding window,/>Representing the integral of the j-th set of adjacent data points corresponding to the fitted curve,/>Representing the j-th operational data point within the sliding window,/>Then the value representing the j-th operational data point,/>Representing the j+1th operating data point within the sliding window,/>Representing the value of the j+1th operating data point in the sliding window,/>The trapezoid area of the j-th set of adjacent data points corresponding to the original data is shown,Representing the loss of two adjacent data points during curve fitting within a sliding window.
2. The intelligent control method of a switchgear according to claim 1, wherein the corrected noise expression level calculation expression is as follows:
；
In the method, in the process of the invention, And/>Respectively representing the ith operating data point/>The noise performance degree before and after correction; /(I)Representing the variance of absolute operational data points over a time range corresponding to the sliding window.
3. The intelligent control method of a switch cabinet according to claim 1, wherein the mass calculation expression of the operation data of the sliding window center is:
；
Wherein the method comprises the steps of Representing the ith operational data point/>Quality of/(I)For the ith operational data point/>Is included in the noise performance level after correction.
4. The intelligent control method of the switch cabinet according to any one of claims 1-3, wherein the number of hidden layers of the BP neural network model is 3, the activation function is a ReLU function, the number of neurons of an output layer is 3, the back propagation process adopts a gradient descent algorithm, and the learning rate initial value is 0.5.
5. An intelligent control system for a switch cabinet, comprising a processor and a memory, wherein the memory stores computer program instructions, and the intelligent control system is characterized in that the intelligent control method for the switch cabinet according to any one of claims 1-4 is realized when the computer program instructions are executed by the processor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410285435.9A CN117891160B (en) | 2024-03-13 | 2024-03-13 | Intelligent control system and method for switch cabinet |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410285435.9A CN117891160B (en) | 2024-03-13 | 2024-03-13 | Intelligent control system and method for switch cabinet |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117891160A CN117891160A (en) | 2024-04-16 |
| CN117891160B true CN117891160B (en) | 2024-05-31 |
Family
ID=90639839
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410285435.9A Active CN117891160B (en) | 2024-03-13 | 2024-03-13 | Intelligent control system and method for switch cabinet |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117891160B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112445136A (en) * | 2020-12-16 | 2021-03-05 | 北京科技大学 | Thickener prediction control method and system based on continuous time neural network |
| CN113642484A (en) * | 2021-08-18 | 2021-11-12 | 湖南师范大学 | Magnetotelluric signal noise suppression method and system based on BP neural network |
| CN114444375A (en) * | 2021-12-13 | 2022-05-06 | 浙江图盛输变电工程有限公司 | High-voltage switch cabinet insulation aging prediction method based on multi-parameter fusion calculation |
| CN116503282A (en) * | 2023-04-24 | 2023-07-28 | 山东交通学院 | Manifold-based excavator construction environment site point cloud denoising method and system |
-
2024
- 2024-03-13 CN CN202410285435.9A patent/CN117891160B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112445136A (en) * | 2020-12-16 | 2021-03-05 | 北京科技大学 | Thickener prediction control method and system based on continuous time neural network |
| CN113642484A (en) * | 2021-08-18 | 2021-11-12 | 湖南师范大学 | Magnetotelluric signal noise suppression method and system based on BP neural network |
| CN114444375A (en) * | 2021-12-13 | 2022-05-06 | 浙江图盛输变电工程有限公司 | High-voltage switch cabinet insulation aging prediction method based on multi-parameter fusion calculation |
| CN116503282A (en) * | 2023-04-24 | 2023-07-28 | 山东交通学院 | Manifold-based excavator construction environment site point cloud denoising method and system |
Non-Patent Citations (1)
| Title |
|---|
| 任子昱.基于改进反向传播神经网络的风电短期功率预测.《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑 信息科技》.全文. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117891160A (en) | 2024-04-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110535146B (en) | Electric power system reactive power optimization method based on depth determination strategy gradient reinforcement learning | |
| Panda et al. | Comparison of particle swarm optimization and genetic algorithm for FACTS-based controller design | |
| CN111092429B (en) | Optimized scheduling method of flexible interconnected power distribution network, storage medium and processor | |
| Chen et al. | Digital IIR filter design using particle swarm optimisation | |
| CN107017640B (en) | A kind of optimal load flow calculation method of electric system, apparatus and system | |
| CN110120670B (en) | DPV-containing power distribution network reactive voltage optimization method, terminal equipment and storage medium | |
| CN110162739B (en) | RFFKBMS algorithm weight updating and optimizing method based on forgetting factor | |
| CN111766777A (en) | PID controller and PID control method | |
| CN105896575B (en) | Hundred megawatt energy storage power control method and system based on self-adaptive dynamic programming | |
| CN112152221B (en) | Load frequency control device and method suitable for information uncertainty system | |
| CN117891160B (en) | Intelligent control system and method for switch cabinet | |
| CN113890039B (en) | Multi-terminal flexible direct-current power distribution network power flow scheduling optimization method | |
| CN112165095B (en) | Low-voltage distribution network reactive power optimization method based on swarm intelligence | |
| CN113759723A (en) | Optimal hybrid H based on genetic algorithm2/H∞Robust controller design method | |
| Zhang et al. | Data-driven load frequency cooperative control for multi-area power system integrated with VSCs and EV aggregators under cyber-attacks | |
| JP2003256005A (en) | Control method and control device | |
| CN116231747A (en) | Distributed photovoltaic inverter control method, device and storage medium | |
| He et al. | Measurement-Driven Damping Control Based on the Deep Transfer Reinforcement Learning to Suppress Sub-synchronous Oscillations in a Large-Scale Renewable Power System | |
| CN115021276A (en) | Power distribution network voltage reactive power control method based on near-end strategy optimization algorithm | |
| CN108494017A (en) | A kind of Autonomous Model micro-grid system distributed and coordinated control method based on inverter | |
| CN111007721A (en) | Method for dynamically selecting sampling period in linear steady system | |
| CN118017523B (en) | Voltage control method, device, equipment and medium for electric power system | |
| CN110729723A (en) | Method and device for obtaining power quality control model in power distribution network | |
| CN117856284B (en) | Deep reinforcement learning-based power grid frequency control method and device | |
| CN114301633B (en) | Control method of micro-grid controller for dealing with deception network attack |
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 |