CN116842322B - Electric motor operation optimization method and system based on data processing - Google Patents
Electric motor operation optimization method and system based on data processing Download PDFInfo
- Publication number
- CN116842322B CN116842322B CN202310886720.1A CN202310886720A CN116842322B CN 116842322 B CN116842322 B CN 116842322B CN 202310886720 A CN202310886720 A CN 202310886720A CN 116842322 B CN116842322 B CN 116842322B
- Authority
- CN
- China
- Prior art keywords
- electric motor
- data
- operation parameter
- value
- debugging
- 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 148
- 238000012545 processing Methods 0.000 title claims abstract description 44
- 238000005457 optimization Methods 0.000 title claims abstract description 22
- 230000002159 abnormal effect Effects 0.000 claims abstract description 44
- 238000001914 filtration Methods 0.000 claims abstract description 31
- 230000008569 process Effects 0.000 claims abstract description 23
- 238000007781 pre-processing Methods 0.000 claims abstract description 9
- 238000013527 convolutional neural network Methods 0.000 claims description 28
- 238000012549 training Methods 0.000 claims description 24
- 238000012360 testing method Methods 0.000 claims description 18
- 238000012216 screening Methods 0.000 claims description 12
- 238000011176 pooling Methods 0.000 claims description 11
- 230000033228 biological regulation Effects 0.000 claims description 7
- 238000001514 detection method Methods 0.000 claims description 7
- 230000005856 abnormality Effects 0.000 claims description 6
- 238000004140 cleaning Methods 0.000 claims description 6
- 238000003062 neural network model Methods 0.000 claims description 6
- 230000006870 function Effects 0.000 claims description 5
- 230000011218 segmentation Effects 0.000 claims description 5
- 238000004088 simulation Methods 0.000 claims description 3
- 230000008901 benefit Effects 0.000 description 4
- 230000007547 defect Effects 0.000 description 4
- 230000004044 response Effects 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000010354 integration Effects 0.000 description 2
- 230000002411 adverse Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000012217 deletion Methods 0.000 description 1
- 230000037430 deletion Effects 0.000 description 1
- 230000005672 electromagnetic field Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000017525 heat dissipation Effects 0.000 description 1
- 230000005389 magnetism Effects 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000013450 outlier detection Methods 0.000 description 1
- 238000012163 sequencing technique Methods 0.000 description 1
- 238000000638 solvent extraction Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/10—Pre-processing; Data cleansing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/903—Querying
- G06F16/9035—Filtering based on additional data, e.g. user or group profiles
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2337—Non-hierarchical techniques using fuzzy logic, i.e. fuzzy clustering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/24323—Tree-organised classifiers
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/2433—Single-class perspective, e.g. one-against-all classification; Novelty detection; Outlier detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/0464—Convolutional networks [CNN, ConvNet]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Abstract
The invention discloses an electric motor operation optimization method and system based on data processing, comprising the following steps: and acquiring operation parameter data in the operation process of the electric motor, preprocessing the data to obtain operation parameter filtering data, classifying the operation parameter filtering data and detecting data outliers to obtain final operation parameter classification data, constructing an electric motor operation prediction model based on the final operation parameter classification data, acquiring abnormal parameters of the electric motor through comparison of the electric motor operation prediction model and a standard electric motor operation prediction model, and selecting a proper debugging method to perform operation optimization processing on the electric motor.
Description
Technical Field
The invention relates to the field of data processing, in particular to an electric motor operation optimization method and system based on data processing.
Background
An electric motor is a device for converting electric energy into mechanical energy, and is mainly used for driving various mechanical equipment and systems, and the electric motor generates force and torque in an electromagnetic field through current to enable a driving shaft to rotate, so that energy conversion and power output are achieved. The electric motor can have the poor heat dissipation in the course of working and lead to the overheat condition, also can have the operation and maintenance overload and lead to the electric motor overheated, output diminish even damage electric motor's condition. The output power of the electric motor is reduced, so that the working efficiency of the electric motor is reduced, the economic benefit is not facilitated, and the electric motor is scrapped due to overheat of the electric motor, and even personal safety is affected, so that the problem of acquiring the electric motor based on data processing is required, and the operation of the electric motor is optimized.
Disclosure of Invention
The invention overcomes the defects of the prior art and provides an electric motor operation optimization method and system based on data processing.
In order to achieve the above purpose, the invention adopts the following technical scheme:
the first aspect of the invention provides an electric motor operation optimization method based on data processing, which comprises the following steps:
s102: collecting operation parameter data in the operation process of the electric motor through a sensor, and carrying out data preprocessing on the operation parameter data to obtain operation parameter filtering data;
s104: constructing an operation parameter data sample based on the operation parameter filtering data, and classifying the operation parameter data sample by using an FCM algorithm to obtain operation parameter classification data;
s106: detecting outliers of the operation parameter classification data by using an isolated forest algorithm, and processing the outliers of the operation parameter classification data to obtain final operation parameter classification data;
s108: based on a convolutional neural network model, combining the final operation parameter classification data and the historical data to construct an electric motor operation prediction model, and generating electric motor abnormal parameters through the electric motor operation prediction model;
S110: and searching big data based on the abnormal parameters of the electric motor to obtain a debugging method for correcting the abnormality of the electric motor, screening the debugging method and outputting the debugging method with highest regulation and control efficiency.
Further, in a preferred embodiment of the present invention, S102 is specifically:
a sensor is arranged in the electric motor, and the sensor is used for directly collecting the operation parameter data in the operation process of the electric motor;
calculating the mean value and standard deviation of the operation parameter data, presetting a standard operation parameter data mean value and a standard deviation threshold value based on a normal distribution principle, defining the operation parameter data with the mean value and the standard deviation outside the standard operation parameter data mean value and the standard deviation threshold value as abnormal values, deleting the abnormal values, performing repeated value detection on the operation parameter data with the abnormal values deleted, deleting the repeated values, and obtaining the operation parameter data with the data cleaned;
the operation parameter data after the data cleaning is imported into a median filter to obtain an operation parameter data sequence, a sliding window is arranged in the median filter, the sliding window slides on the data sequence, and the data in the sliding window are ordered in the sliding process;
Selecting the sorting intermediate value in the sliding window as the unique value in the sliding window, replacing other numerical values in the sliding window with the unique value in the sliding window, repeating the steps, and filtering the operation parameter data after data cleaning to obtain operation parameter filtering data.
Further, in a preferred embodiment of the present invention, S104 is specifically:
randomly selecting K operation parameter data samples as initial clustering centers, and taking the initial clustering centers as centers of various data sets respectively;
calculating Manhattan distances between all the operation parameter data samples and the initial clustering center, and calculating membership values of each operation parameter data sample to various data sets after obtaining the Manhattan distances between the operation parameter data samples and the initial clustering center;
calculating the average value of the operation parameter data samples in various data sets, taking the average value as a new clustering center, presetting a clustering center membership standard value, and repeating the steps if the average value of the operation parameter data samples does not meet the clustering center membership standard value, and continuously classifying the operation parameter data samples by using an FCM algorithm;
and if the average value of the operation parameter data samples meets the membership standard value of the clustering center, outputting a clustering result to obtain operation parameter classification data.
Further, in a preferred embodiment of the present invention, S106 is specifically:
constructing an isolated forest through an isolated forest algorithm, and determining the number of binary trees in the isolated forest and the maximum depth value of each binary tree;
performing data processing on the operation parameter classification data to obtain an operation parameter classification data set, randomly selecting a characteristic value of the operation parameter classification data in the operation parameter classification data set as a characteristic value of a current node, and randomly selecting a segmentation value from a characteristic value range of the characteristic value of the current node;
classifying the operation parameter classification data points with the characteristic values smaller than the dividing values of the current nodes in the operation parameter classification data sets into a left subset according to the characteristic values and the dividing values of the selected current nodes, and classifying the operation parameter classification data points with the characteristic values larger than the dividing values of the current nodes into a right subset;
if the depth value of the current node is smaller than the maximum depth value and the number of the operation parameter classification data points in the left subset and the right subset is not zero, continuously creating the subsets on the basis of the left subset and the right subset;
repeating the steps by taking the left subset and the right subset as new data sets respectively, adding 1 to the depth value of the current node to obtain a new depth value, and marking the current node as a leaf node if the depth value of the current node reaches the maximum depth value or the number of the operation parameter classification data points in the data sets is not greater than a preset threshold value;
Determining the position of a root node of a binary tree, and calculating the path lengths from all leaf nodes to the root node in all binary trees, wherein the path length from one leaf node to the root node corresponds to the path length of one operation parameter classification data point;
setting each state parameter path standard threshold value of the operation parameter classification data, marking the operation parameter classification data points with path lengths which are not in each state parameter path standard threshold value range of the operation parameter classification data as outliers, eliminating the outliers, collecting the rest operation parameter classification data points, and generating final operation parameter classification data.
Further, in a preferred embodiment of the present invention, S108 is specifically:
according to the big data retrieval, standard operation parameter classification data of the electric motors of the same model are obtained, and the final operation parameter classification data of the electric motors are divided into a training set and a testing set;
constructing a convolutional neural network model, importing the training set into a convolutional layer of the convolutional neural network model, defining two convolutional kernels of power and temperature of an electric motor in the convolutional layer, sliding the convolutional kernels in data of the training set, performing dot multiplication on weights of the convolutional kernels and training set data corresponding to sliding passing positions in the sliding process to obtain dot multiplication results, and accumulating the dot multiplication results to obtain a convolutional value;
Inputting the convolution values into pooling layers of a convolution neural network model for maximum pooling treatment, selecting the maximum value in each convolution layer as the characteristic value of a sliding region where the current convolution value is located, fusing all the selected characteristic values, and performing reverse training through a cross entropy loss function until the error converges to a preset value;
testing the reverse trained convolutional neural network model through a test set, defining the reverse trained convolutional neural network model as an electric motor operation prediction model when a test result meets a preset value, and constructing a standard electric motor operation prediction model based on standard operation parameter classification data through the same steps;
respectively inputting electric motor operation condition parameters to the electric motor operation prediction model and the standard electric motor operation prediction model, wherein the electric motor operation condition parameters comprise stator current magnitude, power supply frequency magnitude and rotating sub-load magnitude;
the method comprises the steps of obtaining electric motor prediction parameters generated by an electric motor operation prediction model, defining the electric motor prediction parameters as first type prediction parameters, obtaining standard electric motor operation prediction parameters generated by a standard electric motor operation prediction model, defining the electric motor operation prediction parameters as second type prediction parameters, wherein the prediction parameters comprise power and temperature of an electric motor, and carrying out gray correlation processing on the first type prediction parameters and the second type prediction parameters based on a gray correlation method to obtain abnormal parameters of the electric motor.
Further, in a preferred embodiment of the present invention, S110 is specifically:
based on abnormal parameters of the electric motor, searching in a big data network to obtain all debugging methods of the electric motor, and importing all the debugging methods into the electric motor operation prediction model for simulation debugging;
recording the output power and the running temperature of the electric motor running prediction model under various debugging methods, presetting an output power standard value and a running temperature threshold value, screening to obtain all the debugging methods for enabling the output power of the electric motor running prediction model to reach the output power standard value and the running temperature within the running temperature threshold value, and defining the debugging methods as a first type of debugging method set;
obtaining debugging steps of all the debugging methods in the first type of debugging method set, and eliminating the debugging methods in the debugging methods, the debugging steps of which do not meet preset conditions, so as to obtain a second type of debugging method set;
obtaining the debugging time of all the debugging methods in the second type of debugging method set, and eliminating the debugging methods with the debugging time outside the preset time to obtain a third type of debugging method set;
and calculating the debugging efficiency of the debugging methods in the third type of debugging method set to obtain a debugging efficiency sorting table, selecting the debugging method with the highest debugging efficiency in the debugging efficiency sorting table, using the debugging method with the highest debugging efficiency on the electric motor, and optimizing the operation of the electric motor.
The second aspect of the present invention also provides an electric motor operation optimization system based on data processing, the electric motor operation optimization system comprising a memory and a processor, the memory storing an electric motor operation optimization program, the electric motor operation optimization program when executed by the processor implementing the steps of
Collecting operation parameter data in the operation process of the electric motor through a sensor, and carrying out data preprocessing on the operation parameter data to obtain operation parameter filtering data;
constructing an operation parameter data sample based on the operation parameter filtering data, and classifying the operation parameter data sample by using an FCM algorithm to obtain operation parameter classification data;
detecting outliers of the operation parameter classification data by using an isolated forest algorithm, and processing the outliers of the operation parameter classification data to obtain final operation parameter classification data;
based on a convolutional neural network model, combining the final operation parameter classification data and the historical data to construct an electric motor operation prediction model, and generating electric motor abnormal parameters through the electric motor operation prediction model;
and searching big data based on the abnormal parameters of the electric motor to obtain a debugging method for correcting the abnormality of the electric motor, screening the debugging method and outputting the debugging method with highest regulation and control efficiency.
The invention solves the technical defects in the background technology, and has the following beneficial effects: and acquiring operation parameter data in the operation process of the electric motor, preprocessing the data to obtain operation parameter filtering data, classifying the operation parameter filtering data and detecting data outliers to obtain final operation parameter classification data, constructing an electric motor operation prediction model based on the final operation parameter classification data, acquiring abnormal parameters of the electric motor through comparison of the electric motor operation prediction model and a standard electric motor operation prediction model, and selecting a proper debugging method to perform operation optimization processing on the electric motor. The operation optimization of the electric motor can improve the working efficiency of the electric motor, reduce the power loss, reduce the failure probability of the electric motor, protect the personal safety, prevent the property loss and meet the economic benefit.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, it being obvious that the drawings in the following description are only some embodiments of the invention, and that other embodiments of the drawings can be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 illustrates a flow chart of a method of optimizing operation of an electric motor based on data processing;
FIG. 2 illustrates a flow chart for detecting outliers of operating parameter classification data using an isolated forest algorithm and deriving final operating parameter classification data;
FIG. 3 illustrates a flow chart for constructing an electric motor operation prediction model from final operating parameters and generating abnormal parameters for the electric motor;
fig. 4 shows a program diagram of an electric motor operation optimization system based on data processing.
Detailed Description
In order that the above-recited objects, features and advantages of the present invention will be more clearly understood, a more particular description of the invention will be rendered by reference to the appended drawings and appended detailed description. It should be noted that, in the case of no conflict, the embodiments of the present application and the features in the embodiments may be combined with each other.
In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present invention, however, the present invention may be practiced in other ways than those described herein, and therefore the scope of the present invention is not limited to the specific embodiments disclosed below.
Fig. 1 shows a flow chart illustrating a data processing based method of optimizing the operation of an electric motor, comprising the steps of:
S102: collecting operation parameter data in the operation process of the electric motor through a sensor, and carrying out data preprocessing on the operation parameter data to obtain operation parameter filtering data;
s104: constructing an operation parameter data sample based on the operation parameter filtering data, and classifying the operation parameter data sample by using an FCM algorithm to obtain operation parameter classification data;
s106: detecting outliers of the operation parameter classification data by using an isolated forest algorithm, and processing the outliers of the operation parameter classification data to obtain final operation parameter classification data;
s108: based on a convolutional neural network model, combining the final operation parameter classification data and the historical data to construct an electric motor operation prediction model, and generating electric motor abnormal parameters through the electric motor operation prediction model;
s110: and searching big data based on the abnormal parameters of the electric motor to obtain a debugging method for correcting the abnormality of the electric motor, screening the debugging method and outputting the debugging method with highest regulation and control efficiency.
Further, in a preferred embodiment of the present invention, S102 is specifically:
a sensor is arranged in the electric motor, and the sensor is used for directly collecting the operation parameter data in the operation process of the electric motor;
Calculating the mean value and standard deviation of the operation parameter data, presetting a standard operation parameter data mean value and a standard deviation threshold value based on a normal distribution principle, defining the operation parameter data with the mean value and the standard deviation outside the standard operation parameter data mean value and the standard deviation threshold value as abnormal values, deleting the abnormal values, performing repeated value detection on the operation parameter data with the abnormal values deleted, deleting the repeated values, and obtaining the operation parameter data with the data cleaned;
the operation parameter data after the data cleaning is imported into a median filter to obtain an operation parameter data sequence, a sliding window is arranged in the median filter, the sliding window slides on the data sequence, and the data in the sliding window are ordered in the sliding process;
selecting the sorting intermediate value in the sliding window as the unique value in the sliding window, replacing other numerical values in the sliding window with the unique value in the sliding window, repeating the steps, and filtering the operation parameter data after data cleaning to obtain operation parameter filtering data.
It should be noted that, the sensor may collect the operation parameter data of the electric motor in the operation process, but the operation parameter data may have a problem of excessive noise in the collection process, and the operation parameter data needs to be subjected to data filtering processing. The purpose of performing outlier detection and deletion processing on the operation parameter data is to clean the data, so that the uniformity of the data is realized, and the operation parameter data is subjected to filtering processing by adopting a median filter, so that the data in a filtering window can be the unique value, and the data is smoother. According to the invention, the operation parameter data in the operation process of the electric motor can be subjected to filtering treatment through the median filter, so that the operation parameter data is smoother, and further data processing is facilitated.
Further, in a preferred embodiment of the present invention, S104 is specifically:
randomly selecting K operation parameter data samples as initial clustering centers, and taking the initial clustering centers as centers of various data sets respectively;
calculating Manhattan distances between all the operation parameter data samples and the initial clustering center, and calculating membership values of each operation parameter data sample to various data sets after obtaining the Manhattan distances between the operation parameter data samples and the initial clustering center;
calculating the average value of the operation parameter data samples in various data sets, taking the average value as a new clustering center, presetting a clustering center membership standard value, and repeating the steps if the average value of the operation parameter data samples does not meet the clustering center membership standard value, and continuously classifying the operation parameter data samples by using an FCM algorithm;
and if the average value of the operation parameter data samples meets the membership standard value of the clustering center, outputting a clustering result to obtain operation parameter classification data.
The collected operation parameter data is a set of various data of the electric motor, and the operation parameter data is classified for data modeling of the electric motor, so as to obtain various operation parameter data of the electric motor. Classifying operation parameter data by using an FCM algorithm, randomly selecting K operation parameter data samples from the operation parameter data as initial clustering centers, wherein the similarity distance between the Manhattan distance data is the closer the Euclidean distance between the operation parameter data samples and the initial clustering centers is, the greater the data membership degree between the operation parameter data samples and the initial clustering centers is, and the data membership degree is the similarity of the data. The membership value of the clustering centers is the average value of the operation parameter data samples, when the membership value of the clustering centers meets the membership standard value of the clustering centers, the classification among the clustering centers is proved to be clear, different kinds of data in the operation parameter data can be clearly distinguished, and when the membership value of the clustering centers does not meet the membership standard value of the clustering centers, the steps are required to be repeated. The invention can classify the operation parameter data of the electric motor through the FCM algorithm, and classify, detect and optimize the classification result.
Further, in a preferred embodiment of the present invention, S110 is specifically:
based on abnormal parameters of the electric motor, searching in a big data network to obtain all debugging methods of the electric motor, and importing all the debugging methods into the electric motor operation prediction model for simulation debugging;
recording the output power and the running temperature of the electric motor running prediction model under various debugging methods, presetting an output power standard value and a running temperature threshold value, screening to obtain all the debugging methods for enabling the output power of the electric motor running prediction model to reach the output power standard value and the running temperature within the running temperature threshold value, and defining the debugging methods as a first type of debugging method set;
obtaining debugging steps of all the debugging methods in the first type of debugging method set, and eliminating the debugging methods in the debugging methods, the debugging steps of which do not meet preset conditions, so as to obtain a second type of debugging method set;
obtaining the debugging time of all the debugging methods in the second type of debugging method set, and eliminating the debugging methods with the debugging time outside the preset time to obtain a third type of debugging method set;
and calculating the debugging efficiency of the debugging methods in the third type of debugging method set to obtain a debugging efficiency sorting table, selecting the debugging method with the highest debugging efficiency in the debugging efficiency sorting table, using the debugging method with the highest debugging efficiency on the electric motor, and optimizing the operation of the electric motor.
After obtaining the abnormal parameters of the electric motor through the electric motor operation prediction model, the abnormal parameters need to be debugged to restore the electric motor to normal. The method comprises the steps of obtaining a first type of debugging method set, namely obtaining a second type of debugging method set by searching big data, screening all the debugging methods, and defining the debugging methods which can enable the output power of an electric motor to reach the output power standard value and control the temperature within the operating temperature threshold range as a first type of debugging method set. And eliminating the debugging methods with the debugging time outside the preset time in the second type of debugging method set to obtain a third type of debugging method set, wherein the debugging time is slow, so that the debugging efficiency is influenced, and the economic benefit is greatly influenced. And finally, efficiency sequencing is carried out on the debugging methods in the third type of debugging method set, and the debugging method with the highest efficiency is selected for output. The invention can obtain the electric motor abnormal parameter debugging method with the best effect by screening the debugging method in the aspects of debugging step, debugging time and debugging efficiency.
FIG. 2 shows a flowchart for detecting outliers of operating parameter classification data using an isolated forest algorithm and deriving final operating parameter classification data, comprising the steps of:
s202: constructing an isolated forest by using an isolated forest algorithm, processing the operation parameter data, and dividing a left subset and a right subset;
s204: performing data processing on the left subset and the right subset to generate leaf nodes and root nodes, and acquiring path lengths from all leaf nodes to the root nodes in the binary tree;
s206: and generating final operation parameter classification data according to path lengths from all leaf nodes to root nodes in the binary tree.
Further, in a preferred embodiment of the present invention, S202 is specifically:
constructing an isolated forest through an isolated forest algorithm, and determining the number of binary trees in the isolated forest and the maximum depth value of each binary tree;
performing data processing on the operation parameter classification data to obtain an operation parameter classification data set, randomly selecting a characteristic value of the operation parameter classification data in the operation parameter classification data set as a characteristic value of a current node, and randomly selecting a segmentation value from a characteristic value range of the characteristic value of the current node;
And classifying the operation parameter classification data points with the characteristic values smaller than the dividing values of the current nodes in the operation parameter classification data sets into a left subset according to the characteristic values and the dividing values of the selected current nodes, and classifying the operation parameter classification data points with the characteristic values larger than the dividing values of the current nodes into a right subset.
It should be noted that, the isolated forest algorithm is an algorithm for detecting outliers of data, and in the operation parameter classification data, outlier data may exist in the operation parameter data set of each category, so that the operation parameter data set is incompletely classified, and the detection and rejection of the outliers of the operation parameter classification data need to be performed by using the isolated forest algorithm. The maximum depth value of the binary tree means the number of nodes on the longest path from the root node of the binary tree to the most distant leaf node. The operating parameter data are converted into operating parameter data sets, which means tree nodes that build a binary tree, acting as partitioning data sets. The feature value means the value of the feature for dividing the operation parameter data set; the cut value means a cut threshold point dividing the running parameter data set. And comparing the magnitudes of the characteristic values and the segmentation values of the current node on the current node, and dividing the characteristic values and the segmentation values into a left subset and a right subset. The invention can divide the operation parameter data by comparing the characteristic value and the dividing value of the binary tree node to obtain the left subset and the right subset.
Further, in a preferred embodiment of the present invention, S204 is specifically:
if the depth value of the current node is smaller than the maximum depth value and the number of the operation parameter classification data points in the left subset and the right subset is not zero, continuously creating the subsets on the basis of the left subset and the right subset;
repeating the steps by taking the left subset and the right subset as new data sets respectively, adding 1 to the depth value of the current node to obtain a new depth value, and marking the current node as a leaf node if the depth value of the current node reaches the maximum depth value or the number of the operation parameter classification data points in the data sets is not greater than a preset threshold value;
determining the position of a root node of the binary tree, and calculating the path lengths from all leaf nodes to the root node in all binary trees, wherein the path length from one leaf node to the root node corresponds to the path length of one operation parameter classification data point.
If the depth value of the current node is smaller than the maximum depth value, the number of nodes of the binary tree is insufficient, and the outlier data of the operation parameter data cannot be accurately acquired. Gradually increasing the depth value of the current node until the depth value of the current node is the maximum depth value or the operation parameter classification data point in the operation parameter classification data set is smaller than a preset value, if the current node cannot continue to divide left and right subsets in the binary tree, the node is defined as a leaf node. And obtaining the path length from the leaf node to the root node, and judging state classification data corresponding to the operation parameter data points according to the difference of the path lengths. The invention can obtain the leaf node and the path length from the leaf node to the root node by adjusting the depth value of the node.
Further, in a preferred embodiment of the present invention, S206 specifically includes:
setting each state parameter path standard threshold value of the operation parameter classification data, marking the operation parameter classification data points with path lengths which are not in each state parameter path standard threshold value range of the operation parameter classification data as outliers, eliminating the outliers, collecting the rest operation parameter classification data points, and generating final operation parameter classification data.
It should be noted that, because the state classification data corresponding to the operation parameter data points can be distinguished according to different path lengths, the operation parameter classification data points with path lengths not within the standard threshold range of each state parameter path of the operation parameter classification data are marked as outliers, and the operation parameter classification data need to provide data support for electric motor modeling, so that outliers need to be removed to obtain final operation parameter classification data. The invention can generate final operation parameter classification data by judging the path length of each state parameter of the operation parameter classification data point and removing outliers.
FIG. 3 shows a flowchart for constructing an electric motor operation prediction model from final operating parameters and generating abnormal parameters for the electric motor, comprising the steps of:
s302: constructing a convolutional neural network model, and importing the tissue operation parameter classification data of the electric motor into a convolutional layer of the convolutional neural network model to obtain a convolutional value;
s304: performing pooling treatment and reverse training on the convolutional neural network model according to the convolutional value to obtain an electric motor operation prediction model and constructing a standard electric motor operation prediction model;
s306: and (3) guiding the electric motor operation condition parameters into an electric motor operation prediction model and a standard electric motor operation prediction model for prediction to obtain prediction parameters, and carrying out gray correlation processing on the prediction parameters to obtain abnormal parameters of the electric motor.
Further, the method comprises the following steps. In a preferred embodiment of the present invention, S302 is specifically:
according to the big data retrieval, standard operation parameter classification data of the electric motors of the same model are obtained, and the final operation parameter classification data of the electric motors are divided into a training set and a testing set;
constructing a convolutional neural network model, importing the training set into a convolutional layer of the convolutional neural network model, defining two convolutional kernels of power and temperature of an electric motor in the convolutional layer, sliding the convolutional kernels in data of the training set, performing dot multiplication on weights of the convolutional kernels and training set data corresponding to sliding passing positions in the sliding process to obtain dot multiplication results, and accumulating the dot multiplication results to obtain a convolutional value.
The convolution value refers to that a convolution operation is used to multiply a filter in a convolution neural network model by a local area of input data element by element, and then the multiplication results are added to obtain a single numerical value, wherein the single numerical value is the convolution value. The convolution value can extract the multi-level characteristics of the input operation parameter classification data. The invention can carry out convolution operation on the training set to obtain the convolution value.
Further, the method comprises the following steps. In a preferred embodiment of the present invention, S304 is specifically:
inputting the convolution values into pooling layers of a convolution neural network model for maximum pooling treatment, selecting the maximum value in each convolution layer as the characteristic value of a sliding region where the current convolution value is located, fusing all the selected characteristic values, and performing reverse training through a cross entropy loss function until the error converges to a preset value;
and testing the reverse trained convolutional neural network model through a test set, defining the reverse trained convolutional neural network model as an electric motor operation prediction model when a test result meets a preset value, and constructing a standard electric motor operation prediction model based on standard operation parameter classification data through the same steps.
The purpose of pooling the convolution values is to reduce the spatial dimension of the convolution characteristics, reduce the dimension and complexity of the data, and improve the performance of the convolution neural network model. The purpose of the reverse training is to enable errors to be converged, when the errors are converged to a preset value, the convolutional neural network model is tested by using a test set, and the test result meets the preset value, so that an electric motor operation prediction model can be obtained, and the electric motor operation prediction model can predict the future state of the electric motor through the current state of the electric motor. In the same way, a standard electric motor operation prediction model is constructed, which is an ideal state of operation of the electric motor, without loss of power, and the temperature is kept constant. The standard electric motor operation prediction model is constructed for analysis and comparison with the electric motor operation prediction model. The invention can obtain the electric motor operation prediction model through pooling treatment and reverse training, and can obtain the standard electric motor operation prediction model in the same way.
Further, the method comprises the following steps. In a preferred embodiment of the present invention, S306 specifically is:
respectively inputting electric motor operation condition parameters to the electric motor operation prediction model and the standard electric motor operation prediction model, wherein the electric motor operation condition parameters comprise stator current magnitude, power supply frequency magnitude and rotating sub-load magnitude;
The method comprises the steps of obtaining electric motor prediction parameters generated by an electric motor operation prediction model, defining the electric motor prediction parameters as first type prediction parameters, obtaining standard electric motor operation prediction parameters generated by a standard electric motor operation prediction model, defining the electric motor operation prediction parameters as second type prediction parameters, wherein the prediction parameters comprise power and temperature of an electric motor, and carrying out gray correlation processing on the first type prediction parameters and the second type prediction parameters based on a gray correlation method to obtain abnormal parameters of the electric motor.
Since the standard electric motor operation prediction model is a lossless ideal model, the same parameter conditions are input into the electric motor operation prediction model and the standard electric motor operation prediction model, the obtained prediction parameters are different, and the abnormal situation of the electric motor which may occur after the electric motor is always operated can be judged according to the difference of the prediction parameters. The first type of predicted parameters and the second type of predicted parameters can be subjected to data comparison through a gray correlation method, and abnormal parameters of the electric motor are obtained. The invention can generate the abnormal parameters of the electric motor by comparing the first type of predicted parameters with the second type of predicted parameters.
In addition, the method for optimizing the operation of the electric motor based on data processing further comprises the following steps:
Splitting an electric motor to obtain parts of the electric motor, wherein the parts of the electric motor comprise a bearing, a magnet, a brake, a connector and a radiator;
carrying out surface image recognition and internal flaw detection recognition on the electric motor part by using a camera and an ultrasonic flaw detection device to acquire a part surface image and part internal information;
performing image graying treatment on the part surface image to obtain a part gray level image, guiding the part gray level image into a median filter to perform median filtering to obtain a part gray level filtering image, and performing image feature extraction on the part gray level filtering image to obtain part surface defect information;
importing the surface defect information of the part and the internal information of the part into three-dimensional modeling software, constructing a three-dimensional model of the part, presetting a three-dimensional model of a standard part, and carrying out data integration comparison on the three-dimensional model of the part and the three-dimensional model of the standard part to generate a deviation value of the part;
combining parts in three-dimensional modeling software to obtain an electric motor three-dimensional model, presetting a standard electric motor three-dimensional model, and carrying out data integration comparison on the electric motor three-dimensional model and the standard electric motor three-dimensional model to generate an overall deviation value;
If the component deviation values of all the electric motor components are smaller than the preset value and the overall deviation value is smaller than the preset value, the electric motor is not required to be operated and optimized temporarily;
if the deviation values of the parts of all the electric motor parts are smaller than the preset value, but the whole deviation value is larger than the preset value, the parts of the electric motor do not need to be replaced, and the operation optimization of the electric motor is realized by adjusting the input current, the input power supply frequency and the load size of the electric motor;
if the electric motor has parts with the deviation value larger than the preset value, replacing the parts with the deviation value larger than the preset value, judging the whole deviation value, and carrying out the next operation optimization scheme.
It should be noted that, the electric motor has loss during operation, and the loss position of the electric motor needs to be found and solved, and the bearing of the electric motor is easily worn during high-speed rotation to cause the bearing damage; the magnet of the electric motor is easily affected by temperature change when driving the rotor of the electric motor to rotate, and loses magnetism; the brakes of electric motors are susceptible to wear and failure during frequent start-stop and braking operations; the connector can corrode when operated for a long time; the heat sink also wears out over time. When the loss of parts of the electric motor is too large, the work of the electric motor is directly influenced, and the electric motor needs to be directly replaced. When the loss of the parts of the electric motor is smaller than the preset value, but the working efficiency of the electric motor after the parts are combined is smaller than the preset value, the electric motor needs to be debugged by adjusting the external input condition of the electric motor.
In addition, by acquiring the surface characteristic information and the internal information of the electric motor part, a three-dimensional model can be constructed, and the degree of loss of the electric motor part can be clearly known. The deviation value of the parts of the electric motor is larger than a preset value, the parts are required to be replaced, otherwise, the electric motor is easy to damage. If the deviation value of the electric motor parts is smaller than a preset value, judging the overall deviation degree, and if the overall deviation degree is smaller than the preset value, the running state of the electric motor is not changed greatly in a short period, so that running optimization processing of the electric motor is not needed temporarily; if the overall deviation degree is greater than the preset value, the operation of the electric motor can be optimized by adjusting the input condition of the electric motor. The invention can make an operation optimization scheme of the electric motor by identifying the loss condition of parts of the electric motor.
In addition, the method for optimizing the operation of the electric motor based on data processing further comprises the following steps:
a signal transmitter is arranged in the electric motor sensor, a feedback signal of the electric motor sensor is received through a signal receiver, and whether the signal receiver can receive the feedback signal from the electric motor sensor within preset time is judged;
If the signal receiver cannot receive the feedback signal from the electric motor sensor within the preset time, defining the corresponding electric motor sensor as a fault device;
if the signal receiver can receive the feedback signal from the electric motor sensor within the preset time, the response frequency of the feedback signal is obtained;
if the response frequency of the feedback signal is smaller than the preset response frequency, carrying out signal analysis on the feedback signal to obtain a feedback signal graph;
and judging the attenuation speed of the feedback signal according to the feedback signal graph, and defining the corresponding electric motor sensor as fault equipment when the attenuation speed of the feedback signal is larger than a preset value.
It should be noted that, the function of the electric motor sensor is to monitor the internal parameters of the electric motor, when the electric motor sensor fails, the internal parameters of the electric motor cannot be detected, which has adverse effects on the operation optimization of the subsequent electric motor, and even causes loss of personal safety and property safety, so that the electric motor sensor needs to perform fault detection. Monitoring the electric motor feedback signal can timely acquire the state of the electric motor sensor. The invention can judge the fault of the electric motor sensor through the feedback signal of the electric motor sensor, which is that the staff can overhaul in time and provides a data base for the operation optimization of the electric motor.
As shown in fig. 4, the second aspect of the present invention further provides an electric motor operation optimization system based on data processing, the electric motor operation optimization system includes a memory 41 and a processor 42, the memory 41 stores an electric motor operation optimization program, and when the electric motor operation optimization program is executed by the processor 42, the following steps are implemented:
collecting operation parameter data in the operation process of the electric motor through a sensor, and carrying out data preprocessing on the operation parameter data to obtain operation parameter filtering data;
constructing an operation parameter data sample based on the operation parameter filtering data, and classifying the operation parameter data sample by using an FCM algorithm to obtain operation parameter classification data;
detecting outliers of the operation parameter classification data by using an isolated forest algorithm, and processing the outliers of the operation parameter classification data to obtain final operation parameter classification data;
based on a convolutional neural network model, combining the final operation parameter classification data and the historical data to construct an electric motor operation prediction model, and generating electric motor abnormal parameters through the electric motor operation prediction model;
And searching big data based on the abnormal parameters of the electric motor to obtain a debugging method for correcting the abnormality of the electric motor, screening the debugging method and outputting the debugging method with highest regulation and control efficiency.
The foregoing is merely illustrative embodiments of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art can easily think about variations or substitutions within the technical scope of the present invention, and the invention should be covered. Therefore, the protection scope of the invention is subject to the protection scope of the claims.
Claims (6)
1. A method for optimizing the operation of an electric motor based on data processing, comprising the steps of:
collecting operation parameter data in the operation process of the electric motor through a sensor, and carrying out data preprocessing on the operation parameter data to obtain operation parameter filtering data;
constructing an operation parameter data sample based on the operation parameter filtering data, and classifying the operation parameter data sample by using an FCM algorithm to obtain operation parameter classification data;
detecting outliers of the operation parameter classification data by using an isolated forest algorithm, and processing the outliers of the operation parameter classification data to obtain final operation parameter classification data;
Based on a convolutional neural network model, combining the final operation parameter classification data and the historical data to construct an electric motor operation prediction model, and generating abnormal parameters of the electric motor through the electric motor operation prediction model;
based on the abnormal parameters of the electric motor, retrieving the big data to obtain a debugging method for correcting the abnormality of the electric motor, screening the debugging method, and outputting the debugging method with highest regulation and control efficiency;
the method comprises the steps of constructing an electric motor operation prediction model based on a convolutional neural network model by combining final operation parameter classification data and historical data, and generating abnormal parameters of an electric motor through the electric motor operation prediction model, wherein the abnormal parameters comprise the following specific steps:
according to the big data retrieval, standard operation parameter classification data of the electric motors of the same model are obtained, and the final operation parameter classification data of the electric motors are divided into a training set and a testing set;
constructing a convolutional neural network model, importing the training set into a convolutional layer of the convolutional neural network model, defining two convolutional kernels of power and temperature of an electric motor in the convolutional layer, sliding the convolutional kernels in data of the training set, performing dot multiplication on weights of the convolutional kernels and training set data corresponding to sliding passing positions in the sliding process to obtain dot multiplication results, and accumulating the dot multiplication results to obtain a convolutional value;
Inputting the convolution values into pooling layers of a convolution neural network model for maximum pooling treatment, selecting the maximum value in each convolution layer as the characteristic value of a sliding region where the current convolution value is located, fusing all the selected characteristic values, and performing reverse training through a cross entropy loss function until the error converges to a preset value;
testing the reverse trained convolutional neural network model through a test set, defining the reverse trained convolutional neural network model as an electric motor operation prediction model when a test result meets a preset value, and constructing a standard electric motor operation prediction model based on standard operation parameter classification data through the same steps;
respectively inputting electric motor operation condition parameters to the electric motor operation prediction model and the standard electric motor operation prediction model, wherein the electric motor operation condition parameters comprise stator current magnitude, power supply frequency magnitude and rotating sub-load magnitude;
the method comprises the steps of obtaining electric motor prediction parameters generated by an electric motor operation prediction model, defining the electric motor prediction parameters as first type prediction parameters, obtaining standard electric motor operation prediction parameters generated by a standard electric motor operation prediction model, defining the electric motor operation prediction parameters as second type prediction parameters, wherein the prediction parameters comprise power and temperature of an electric motor, and carrying out gray correlation processing on the first type prediction parameters and the second type prediction parameters based on a gray correlation method to obtain abnormal parameters of the electric motor.
2. The method for optimizing the operation of an electric motor based on data processing according to claim 1, wherein the operation parameter data of the electric motor in the operation process is collected by a sensor, and the operation parameter data is subjected to data preprocessing to obtain operation parameter filtering data, specifically:
a sensor is arranged in the electric motor, and the sensor is used for directly collecting the operation parameter data in the operation process of the electric motor;
calculating the mean value and standard deviation of the operation parameter data, presetting a standard operation parameter data mean value and a standard deviation threshold value based on a normal distribution principle, defining the operation parameter data with the mean value and the standard deviation outside the standard operation parameter data mean value and the standard deviation threshold value as abnormal values, deleting the abnormal values, performing repeated value detection on the operation parameter data with the abnormal values deleted, deleting the repeated values, and obtaining the operation parameter data with the data cleaned;
the operation parameter data after the data cleaning is imported into a median filter to obtain an operation parameter data sequence, a sliding window is arranged in the median filter, the sliding window slides on the data sequence, and the data in the sliding window are ordered in the sliding process;
Selecting the sorting intermediate value in the sliding window as the unique value in the sliding window, replacing other numerical values in the sliding window with the unique value in the sliding window, repeating the steps, and filtering the operation parameter data after data cleaning to obtain operation parameter filtering data.
3. The method for optimizing operation of an electric motor based on data processing according to claim 1, wherein the operation parameter data sample is constructed based on the operation parameter filtering data, and the operation parameter data sample is classified by using FCM algorithm to obtain operation parameter classification data, specifically:
randomly selecting K operation parameter data samples as initial clustering centers, and taking the initial clustering centers as centers of various data sets respectively;
calculating Manhattan distances between all the operation parameter data samples and the initial clustering center, and calculating membership values of each operation parameter data sample to various data sets after obtaining the Manhattan distances between the operation parameter data samples and the initial clustering center;
calculating the average value of the operation parameter data samples in various data sets, taking the average value as a new clustering center, presetting a clustering center membership standard value, and repeating the steps if the average value of the operation parameter data samples does not meet the clustering center membership standard value, and continuously classifying the operation parameter data samples by using an FCM algorithm;
And if the average value of the operation parameter data samples meets the membership standard value of the clustering center, outputting a clustering result to obtain operation parameter classification data.
4. The method for optimizing operation of an electric motor based on data processing according to claim 1, wherein the method uses an isolated forest algorithm to detect outliers of the operation parameter classification data, and processes the outliers of the operation parameter classification data to obtain final operation parameter classification data, specifically:
constructing an isolated forest through an isolated forest algorithm, and determining the number of binary trees in the isolated forest and the maximum depth value of each binary tree;
performing data processing on the operation parameter classification data to obtain an operation parameter classification data set, randomly selecting a characteristic value of the operation parameter classification data in the operation parameter classification data set as a characteristic value of a current node, and randomly selecting a segmentation value from a characteristic value range of the characteristic value of the current node;
classifying the operation parameter classification data points with the characteristic values smaller than the dividing values of the current nodes in the operation parameter classification data sets into a left subset according to the characteristic values and the dividing values of the selected current nodes, and classifying the operation parameter classification data points with the characteristic values larger than the dividing values of the current nodes into a right subset;
If the depth value of the current node is smaller than the maximum depth value and the number of the operation parameter classification data points in the left subset and the right subset is not zero, continuously creating the subsets on the basis of the left subset and the right subset;
repeating the steps by taking the left subset and the right subset as new data sets respectively, adding 1 to the depth value of the current node to obtain a new depth value, and marking the current node as a leaf node if the depth value of the current node reaches the maximum depth value or the number of the operation parameter classification data points in the data sets is not greater than a preset threshold value;
determining the position of a root node of a binary tree, and calculating the path lengths from all leaf nodes to the root node in all binary trees, wherein the path length from one leaf node to the root node corresponds to the path length of one operation parameter classification data point;
setting each state parameter path standard threshold value of the operation parameter classification data, marking the operation parameter classification data points with path lengths which are not in each state parameter path standard threshold value range of the operation parameter classification data as outliers, eliminating the outliers, collecting the rest operation parameter classification data points, and generating final operation parameter classification data.
5. The method for optimizing the operation of an electric motor based on data processing according to claim 1, wherein the method for searching big data based on abnormal parameters of the electric motor to obtain all the debugging methods of the electric motor, screening the debugging methods and outputting the debugging method with the highest regulation and control efficiency is specifically as follows:
based on abnormal parameters of the electric motor, searching in a big data network to obtain all debugging methods of the electric motor, and importing all the debugging methods into the electric motor operation prediction model for simulation debugging;
recording the output power and the running temperature of the electric motor running prediction model under various debugging methods, presetting an output power standard value and a running temperature threshold value, screening to obtain all the debugging methods for enabling the output power of the electric motor running prediction model to reach the output power standard value and the running temperature within the running temperature threshold value, and defining the debugging methods as a first type of debugging method set;
obtaining debugging steps of all the debugging methods in the first type of debugging method set, and eliminating the debugging methods in the debugging methods, the debugging steps of which do not meet preset conditions, so as to obtain a second type of debugging method set;
Obtaining the debugging time of all the debugging methods in the second type of debugging method set, and eliminating the debugging methods with the debugging time outside the preset time to obtain a third type of debugging method set;
and calculating the debugging efficiency of the debugging methods in the third type of debugging method set to obtain a debugging efficiency sorting table, selecting the debugging method with the highest debugging efficiency in the debugging efficiency sorting table, using the debugging method with the highest debugging efficiency on the electric motor, and optimizing the operation of the electric motor.
6. An electric motor operation optimization system based on data processing, characterized in that the electric motor operation optimization system comprises a memory and a processor, wherein an electric motor operation optimization program is stored in the memory, and when the electric motor operation optimization program is executed by the processor, the following steps are realized:
collecting operation parameter data in the operation process of the electric motor through a sensor, and carrying out data preprocessing on the operation parameter data to obtain operation parameter filtering data;
constructing an operation parameter data sample based on the operation parameter filtering data, and classifying the operation parameter data sample by using an FCM algorithm to obtain operation parameter classification data;
Detecting outliers of the operation parameter classification data by using an isolated forest algorithm, and processing the outliers of the operation parameter classification data to obtain final operation parameter classification data;
based on a convolutional neural network model, combining the final operation parameter classification data and the historical data to construct an electric motor operation prediction model, and generating abnormal parameters of the electric motor through the electric motor operation prediction model;
based on the abnormal parameters of the electric motor, retrieving the big data to obtain a debugging method for correcting the abnormality of the electric motor, screening the debugging method, and outputting the debugging method with highest regulation and control efficiency;
the method comprises the steps of constructing an electric motor operation prediction model based on a convolutional neural network model by combining final operation parameter classification data and historical data, and generating abnormal parameters of an electric motor through the electric motor operation prediction model, wherein the abnormal parameters comprise the following specific steps:
according to the big data retrieval, standard operation parameter classification data of the electric motors of the same model are obtained, and the final operation parameter classification data of the electric motors are divided into a training set and a testing set;
constructing a convolutional neural network model, importing the training set into a convolutional layer of the convolutional neural network model, defining two convolutional kernels of power and temperature of an electric motor in the convolutional layer, sliding the convolutional kernels in data of the training set, performing dot multiplication on weights of the convolutional kernels and training set data corresponding to sliding passing positions in the sliding process to obtain dot multiplication results, and accumulating the dot multiplication results to obtain a convolutional value;
Inputting the convolution values into pooling layers of a convolution neural network model for maximum pooling treatment, selecting the maximum value in each convolution layer as the characteristic value of a sliding region where the current convolution value is located, fusing all the selected characteristic values, and performing reverse training through a cross entropy loss function until the error converges to a preset value;
testing the reverse trained convolutional neural network model through a test set, defining the reverse trained convolutional neural network model as an electric motor operation prediction model when a test result meets a preset value, and constructing a standard electric motor operation prediction model based on standard operation parameter classification data through the same steps;
respectively inputting electric motor operation condition parameters to the electric motor operation prediction model and the standard electric motor operation prediction model, wherein the electric motor operation condition parameters comprise stator current magnitude, power supply frequency magnitude and rotating sub-load magnitude;
the method comprises the steps of obtaining electric motor prediction parameters generated by an electric motor operation prediction model, defining the electric motor prediction parameters as first type prediction parameters, obtaining standard electric motor operation prediction parameters generated by a standard electric motor operation prediction model, defining the electric motor operation prediction parameters as second type prediction parameters, wherein the prediction parameters comprise power and temperature of an electric motor, and carrying out gray correlation processing on the first type prediction parameters and the second type prediction parameters based on a gray correlation method to obtain abnormal parameters of the electric motor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310886720.1A CN116842322B (en) | 2023-07-19 | 2023-07-19 | Electric motor operation optimization method and system based on data processing |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310886720.1A CN116842322B (en) | 2023-07-19 | 2023-07-19 | Electric motor operation optimization method and system based on data processing |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116842322A CN116842322A (en) | 2023-10-03 |
| CN116842322B true CN116842322B (en) | 2024-02-23 |
Family
ID=88165059
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310886720.1A Active CN116842322B (en) | 2023-07-19 | 2023-07-19 | Electric motor operation optimization method and system based on data processing |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116842322B (en) |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109543765A (en) * | 2018-08-23 | 2019-03-29 | 江苏海平面数据科技有限公司 | A kind of industrial data denoising method based on improvement IForest |
| CN111340063A (en) * | 2020-02-10 | 2020-06-26 | 北京华电天仁电力控制技术有限公司 | Coal mill data anomaly detection method |
| CN113420509A (en) * | 2021-07-07 | 2021-09-21 | 华能(浙江)能源开发有限公司清洁能源分公司 | Wind turbine state evaluation method and device and storage medium |
| CN114492578A (en) * | 2021-12-24 | 2022-05-13 | 西南交通大学 | Electrical equipment operation data classification method based on BP neural network |
| CN115021679A (en) * | 2022-08-09 | 2022-09-06 | 国网山西省电力公司大同供电公司 | Photovoltaic equipment fault detection method based on multi-dimensional outlier detection |
| CN115263680A (en) * | 2022-07-04 | 2022-11-01 | 华能吉林发电有限公司镇赉风电厂 | Abnormal temperature rise early warning method for variable pitch motor by combining TF-IDF model and LSTM model |
| CN116227745A (en) * | 2023-05-06 | 2023-06-06 | 中国水产科学研究院南海水产研究所 | Big data-based research and analysis method and system for fishing vessels |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| DE102019135608A1 (en) * | 2019-12-20 | 2021-06-24 | Bayerische Motoren Werke Aktiengesellschaft | Method, device and system for the detection of abnormal operating conditions of a device |
-
2023
- 2023-07-19 CN CN202310886720.1A patent/CN116842322B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109543765A (en) * | 2018-08-23 | 2019-03-29 | 江苏海平面数据科技有限公司 | A kind of industrial data denoising method based on improvement IForest |
| CN111340063A (en) * | 2020-02-10 | 2020-06-26 | 北京华电天仁电力控制技术有限公司 | Coal mill data anomaly detection method |
| CN113420509A (en) * | 2021-07-07 | 2021-09-21 | 华能(浙江)能源开发有限公司清洁能源分公司 | Wind turbine state evaluation method and device and storage medium |
| CN114492578A (en) * | 2021-12-24 | 2022-05-13 | 西南交通大学 | Electrical equipment operation data classification method based on BP neural network |
| CN115263680A (en) * | 2022-07-04 | 2022-11-01 | 华能吉林发电有限公司镇赉风电厂 | Abnormal temperature rise early warning method for variable pitch motor by combining TF-IDF model and LSTM model |
| CN115021679A (en) * | 2022-08-09 | 2022-09-06 | 国网山西省电力公司大同供电公司 | Photovoltaic equipment fault detection method based on multi-dimensional outlier detection |
| CN116227745A (en) * | 2023-05-06 | 2023-06-06 | 中国水产科学研究院南海水产研究所 | Big data-based research and analysis method and system for fishing vessels |
Non-Patent Citations (1)
| Title |
|---|
| 基于MCS-MIFS与LightGBM的燃气轮机功率预测方法;黄伟 等;电力科学与工程;第36卷(第05期);23-31 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116842322A (en) | 2023-10-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Madhusudana et al. | Condition monitoring of face milling tool using K-star algorithm and histogram features of vibration signal | |
| CN110826648B (en) | Method for realizing fault detection by utilizing time sequence clustering algorithm | |
| CN115034248A (en) | Automatic diagnostic method, system and storage medium for equipment | |
| CN113569903B (en) | Method, system, equipment, medium and terminal for predicting cutter abrasion of numerical control machine tool | |
| Louen et al. | A new framework for remaining useful life estimation using support vector machine classifier | |
| CN109597315B (en) | Method, equipment and system for identifying health degradation state of mechanical equipment | |
| CN113505655B (en) | Intelligent bearing fault diagnosis method for digital twin system | |
| CN113798920B (en) | Cutter wear state monitoring method based on variational automatic encoder and extreme learning machine | |
| CN111209934A (en) | Fan fault prediction and alarm method and system | |
| Tayyab et al. | Intelligent fault diagnosis of rotating machine elements using machine learning through optimal features extraction and selection | |
| CN117421684B (en) | Abnormal data monitoring and analyzing method based on data mining and neural network | |
| Atamuradov et al. | Feature selection and fault‐severity classification–based machine health assessment methodology for point machine sliding‐chair degradation | |
| CN117131110B (en) | Method and system for monitoring dielectric loss of capacitive equipment based on correlation analysis | |
| Khan et al. | System design for early fault diagnosis of machines using vibration features | |
| CN114326593B (en) | Cutter life prediction system and method | |
| CN114151293B (en) | Fault early warning method, system, equipment and storage medium of fan variable pitch system | |
| CN117076869B (en) | Time-frequency domain fusion fault diagnosis method and system for rotary machine | |
| CN116842322B (en) | Electric motor operation optimization method and system based on data processing | |
| CN112664410B (en) | Big data-based modeling method for unit online monitoring system | |
| CN111275136B (en) | Fault prediction system based on small sample and early warning method thereof | |
| CN116030955B (en) | Medical equipment state monitoring method and related device based on Internet of things | |
| CN116467652A (en) | Bearing fault prediction method and system based on convolution noise reduction self-encoder | |
| Li et al. | Gear fault detection based on adaptive wavelet packet feature extraction and relevance vector machine | |
| Guerroum et al. | Machine learning for the predictive maintenance of a Jaw Crusher in the mining industry | |
| Park et al. | Prediction of the CNC tool wear using the machine learning technique |
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 |