CN117498762A - Temperature prediction method and related equipment for permanent magnet synchronous motor on vehicle - Google Patents
Temperature prediction method and related equipment for permanent magnet synchronous motor on vehicle Download PDFInfo
- Publication number
- CN117498762A CN117498762A CN202311260222.2A CN202311260222A CN117498762A CN 117498762 A CN117498762 A CN 117498762A CN 202311260222 A CN202311260222 A CN 202311260222A CN 117498762 A CN117498762 A CN 117498762A
- Authority
- CN
- China
- Prior art keywords
- data
- input feature
- feature set
- set data
- temperature
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 52
- 230000001360 synchronised effect Effects 0.000 title claims abstract description 49
- 230000004927 fusion Effects 0.000 claims abstract description 39
- 238000012545 processing Methods 0.000 claims abstract description 31
- 238000007781 pre-processing Methods 0.000 claims abstract description 11
- 230000000875 corresponding effect Effects 0.000 claims description 36
- 230000002159 abnormal effect Effects 0.000 claims description 19
- 239000012634 fragment Substances 0.000 claims description 19
- 230000002596 correlated effect Effects 0.000 claims description 17
- 238000004364 calculation method Methods 0.000 claims description 16
- 238000010606 normalization Methods 0.000 claims description 14
- 238000003062 neural network model Methods 0.000 claims description 13
- 238000001914 filtration Methods 0.000 claims description 11
- 230000006870 function Effects 0.000 claims description 11
- 230000008569 process Effects 0.000 claims description 10
- 230000015654 memory Effects 0.000 claims description 9
- 238000005457 optimization Methods 0.000 claims description 6
- 230000004044 response Effects 0.000 claims description 6
- 239000006185 dispersion Substances 0.000 claims description 5
- 238000012549 training Methods 0.000 description 9
- 238000010586 diagram Methods 0.000 description 6
- 239000002131 composite material Substances 0.000 description 5
- 230000008878 coupling Effects 0.000 description 3
- 238000010168 coupling process Methods 0.000 description 3
- 238000005859 coupling reaction Methods 0.000 description 3
- 238000012360 testing method Methods 0.000 description 3
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 3
- 241000282326 Felis catus Species 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 2
- 230000015572 biosynthetic process Effects 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 238000004590 computer program Methods 0.000 description 2
- 238000009795 derivation Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000007613 environmental effect Effects 0.000 description 2
- 238000009499 grossing Methods 0.000 description 2
- 230000007774 longterm Effects 0.000 description 2
- 238000003786 synthesis reaction Methods 0.000 description 2
- 230000032683 aging Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000008030 elimination Effects 0.000 description 1
- 238000003379 elimination reaction Methods 0.000 description 1
- 238000005265 energy consumption Methods 0.000 description 1
- 238000004880 explosion Methods 0.000 description 1
- 239000011810 insulating material Substances 0.000 description 1
- 230000007787 long-term memory Effects 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000013021 overheating Methods 0.000 description 1
- 230000002035 prolonged effect Effects 0.000 description 1
- 230000000306 recurrent effect Effects 0.000 description 1
- 230000004043 responsiveness Effects 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
- 230000006403 short-term memory Effects 0.000 description 1
- 238000012800 visualization Methods 0.000 description 1
- 238000004804 winding 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
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
- G06F18/253—Fusion techniques of extracted features
-
- 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
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02P—CONTROL OR REGULATION OF ELECTRIC MOTORS, ELECTRIC GENERATORS OR DYNAMO-ELECTRIC CONVERTERS; CONTROLLING TRANSFORMERS, REACTORS OR CHOKE COILS
- H02P25/00—Arrangements or methods for the control of AC motors characterised by the kind of AC motor or by structural details
- H02P25/02—Arrangements or methods for the control of AC motors characterised by the kind of AC motor or by structural details characterised by the kind of motor
- H02P25/022—Synchronous motors
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02P—CONTROL OR REGULATION OF ELECTRIC MOTORS, ELECTRIC GENERATORS OR DYNAMO-ELECTRIC CONVERTERS; CONTROLLING TRANSFORMERS, REACTORS OR CHOKE COILS
- H02P29/00—Arrangements for regulating or controlling electric motors, appropriate for both AC and DC motors
- H02P29/60—Controlling or determining the temperature of the motor or of the drive
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02P—CONTROL OR REGULATION OF ELECTRIC MOTORS, ELECTRIC GENERATORS OR DYNAMO-ELECTRIC CONVERTERS; CONTROLLING TRANSFORMERS, REACTORS OR CHOKE COILS
- H02P2207/00—Indexing scheme relating to controlling arrangements characterised by the type of motor
- H02P2207/05—Synchronous machines, e.g. with permanent magnets or DC excitation
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/60—Other road transportation technologies with climate change mitigation effect
- Y02T10/64—Electric machine technologies in electromobility
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Power Engineering (AREA)
- General Health & Medical Sciences (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Control Of Electric Motors In General (AREA)
- Health & Medical Sciences (AREA)
Abstract
The application discloses a temperature prediction method of a permanent magnet synchronous motor on a vehicle. The method comprises the following steps: acquiring current operation data of the permanent magnet synchronous motor when the vehicle runs; preprocessing the current operation data to obtain first input feature set data; performing feature processing on the first input feature set data to obtain second input feature set data; obtaining target input feature set data from the first input feature set data and the second input feature set data; and inputting the target input feature set data into a fusion model to obtain the current predicted temperature of the permanent magnet synchronous motor so as to realize the temperature prediction of the permanent magnet synchronous motor, wherein the fusion model is constructed based on a tree model and a door control circulation unit. The application also discloses related equipment. The temperature prediction method and the temperature prediction system improve accuracy, stability and robustness of the temperature prediction model, and achieve accurate temperature prediction of the permanent magnet synchronous motor.
Description
Technical Field
The disclosed embodiments of the present application relate to the field of motor control technology, and more particularly, to a temperature prediction method and related apparatus for a permanent magnet synchronous motor on a vehicle.
Background
Owing to the advantages of high efficiency, high performance, wide working range, high responsiveness and the like, the permanent magnet synchronous motor is widely applied to a core driving component of a new energy electric automobile, but in actual use, high power density can cause serious temperature rise, and high temperature can cause ageing and damage of components such as windings, insulating materials and the like, so that the reliability and the service life of the motor are reduced. Therefore, the temperature of the permanent magnet synchronous motor is accurately predicted, measures can be taken in time to reduce the temperature, damage to the motor caused by overheating is avoided, and the service life of the motor is prolonged.
Disclosure of Invention
According to an embodiment of the present application, a method and related apparatus for predicting a temperature of a permanent magnet synchronous motor on a vehicle are provided to solve the above-mentioned problems.
The first aspect of the application discloses a temperature prediction method of a permanent magnet synchronous motor on a vehicle, comprising the following steps: acquiring current operation data of the permanent magnet synchronous motor when the vehicle runs; preprocessing the current operation data to obtain first input feature set data; performing feature processing on the first input feature set data to obtain second input feature set data; obtaining target input feature set data from the first input feature set data and the second input feature set data; and inputting the target input feature set data into a fusion model to obtain the current predicted temperature of the permanent magnet synchronous motor so as to realize the temperature prediction of the permanent magnet synchronous motor, wherein the fusion model is constructed based on a tree model and a door control circulation unit.
In some embodiments, the target input feature set data includes first target input feature subset data derived from the first input feature set data and second target input feature subset data derived from the second input feature set data, the fusion model includes a first layer tree model, a second layer tree model, and a third layer neural network model; inputting the target input feature set data into a fusion model to obtain the current predicted temperature of the permanent magnet synchronous motor, wherein the method comprises the following steps: inputting the first target input feature subset data into the first layer tree model, and outputting first prediction result data; inputting the first prediction result data and the second target input feature subset data into the second layer tree model, and outputting second prediction result data; inputting the first target input feature subset data and the second target input feature subset data into a third layer neural network model to obtain third prediction result data; and carrying out weighted fusion on the first predicted result data, the second predicted result data and the third predicted result data to obtain the predicted temperature of the permanent magnet synchronous motor.
In some embodiments, the first input feature set data includes temperature feature data, and performing feature processing on the first input feature set data to obtain second input feature set data includes: performing feature diffraction based on the first input feature set data to generate derivative feature set data; and calculating by using the temperature characteristic data to obtain temperature difference characteristic data, and further forming the second input characteristic set data by the derivative characteristic set data and the temperature difference characteristic data.
In some embodiments, the obtaining target input feature set data from the first input feature set data and the second input feature set data includes: filtering the data features in the first input feature set data and the second input feature set data by using an exponential weighted moving average to obtain feature data to be correlated; in response to completion of the filtering process, performing correlation coefficient calculation to obtain a correlation coefficient corresponding to the feature data to be correlated; and judging based on the correlation coefficient, and carrying out normalization processing on the characteristic data corresponding to the correlation coefficient meeting the preset condition to obtain target input characteristic set data.
In some embodiments, the correlation coefficient calculation includes: taking the predicted temperature characteristic as a reference sequence, and calculating a correlation coefficient between a characteristic sequence corresponding to the characteristic data to be correlated and the reference sequence; the judging based on the correlation coefficient and the normalizing processing of the characteristic data corresponding to the correlation coefficient meeting the preset condition comprises the following steps: sorting the correlation coefficients according to the values to determine characteristic data corresponding to the correlation coefficients meeting preset conditions; and carrying out normalization processing on the characteristic data corresponding to the correlation coefficient meeting the preset condition by using a dispersion normalization method, so as to obtain target input characteristic set data.
In some embodiments, the preprocessing the current operation data to obtain first input feature set data includes: filling the missing value of the current operation data by using a Lagrangian interpolation method; dividing the current operation data to obtain a plurality of driving fragments in response to the completion of filling the missing values; and removing abnormal values in the plurality of driving fragments to obtain the first input feature set data.
In some embodiments, the dividing the current running data to obtain a plurality of running segments includes: deleting the operation fragments corresponding to the target parameters of the current operation data as the first preset value, and dividing the current operation data into a plurality of driving fragments by taking the operation fragments as intervals; the removing the outliers in the plurality of driving segments includes: deleting continuous unchanged operation data in a preset time period in the plurality of driving fragments; calculating characteristic data abnormal values of each driving segment in the plurality of driving segments, and deleting the driving segment corresponding to the data abnormal value larger than a second preset value.
In some embodiments, further comprising: performing super-parameter adjustment on the fusion model by using Bayesian optimization to optimize and update the fusion model; and using the mean square error and the average absolute value error as a loss function of the fusion model.
A second aspect of the present application discloses an electronic device, including a memory and a processor coupled to each other, where the processor is configured to execute program instructions stored in the memory, so as to implement the method for predicting a temperature of a permanent magnet synchronous motor on a vehicle according to the first aspect.
A fourth aspect of the present application discloses a non-transitory computer readable storage medium having stored thereon program instructions which, when executed by a processor, implement the method of predicting the temperature of a permanent magnet synchronous motor on a vehicle as described in the first aspect.
The beneficial effects of this application are: in the embodiment, the current running data of the permanent magnet synchronous motor is obtained when the vehicle runs, the current running data is processed to obtain the first input feature set data and the second input feature set data, the target input feature set data is obtained from the first input feature set data and the second input feature set data, the target input feature set data is input into a fusion model constructed based on a tree model and a door control circulation unit, the current prediction temperature of the permanent magnet synchronous motor is obtained, the accuracy, stability and robustness of the temperature prediction model are improved, and the accurate temperature prediction of the permanent magnet synchronous motor is realized.
Drawings
The application will be further described with reference to the accompanying drawings and embodiments, in which:
fig. 1 is a flow chart of a temperature prediction method of a permanent magnet synchronous motor on a vehicle according to an embodiment of the present application;
FIG. 2 is a schematic diagram of a fusion model according to an embodiment of the present application;
fig. 3 is a schematic structural diagram of an electronic device according to an embodiment of the present application;
fig. 4 is a schematic structural diagram of a nonvolatile computer-readable storage medium according to an embodiment of the present application.
Detailed Description
Reference in the specification to "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment may be included in at least one embodiment of the application. The appearances of such phrases in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments. Those of skill in the art will explicitly and implicitly appreciate that the embodiments described herein may be combined with other embodiments.
The term "and/or" in this application is merely an association relation describing an associated object, and indicates that three relations may exist, for example, a and/or B may indicate: a exists alone, A and B exist together, and B exists alone. In addition, the character "/" herein generally indicates that the front and rear associated objects are an "or" relationship. Further, "a plurality" herein means two or more than two. In addition, the term "at least one" herein means any one of a plurality or any combination of at least two of a plurality, for example, including at least one of A, B, C, and may mean including any one or more elements selected from the group consisting of A, B and C. Furthermore, the terms "first," "second," and "third" in this application are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated.
In order to enable those skilled in the art to better understand the technical solutions of the present application, the technical solutions of the present application are described in further detail below with reference to the accompanying drawings and the detailed description.
Referring to fig. 1, fig. 1 is a flowchart of a temperature prediction method of a permanent magnet synchronous motor on a vehicle according to an embodiment of the present application. The execution subject of the method can be an electronic device with a computing function, such as a microcomputer, a server, a mobile device such as a notebook computer, a tablet computer, and the like.
It should be noted that, if there are substantially the same results, the method of the present application is not limited to the flow sequence shown in fig. 1.
In some possible implementations, the method may be implemented by a processor invoking computer readable instructions stored in a memory, as shown in fig. 1, and may include the steps of:
s11: and acquiring current operation data of the permanent magnet synchronous motor when the vehicle runs.
And acquiring current operation data of the permanent magnet synchronous motor during running of the vehicle, for example, acquiring temperature data of a rear motor driving sensor during running of the vehicle and corresponding data such as water inlet temperature, motor rotation speed, motor D-axis current feedback and the like.
S12: and preprocessing the current operation data to obtain first input feature set data.
And preprocessing the current operation data, wherein the preprocessing comprises filling of a missing value, removal of an abnormal value and division of a driving fragment. The first input feature set data can comprise a set of D-axis current/voltage of the rear motor, a set of Q-axis current/voltage of the rear motor, a set of rotation speed of the rear motor, a set of water inlet temperature, a set of torque and a set of motor temperature.
S13: and performing feature processing on the first input feature set data to obtain second input feature set data.
And performing feature processing, such as process feature derivation processing, on the first input feature set data to obtain second input feature set data, wherein the second input feature set data can comprise a current synthesis value, a voltage synthesis value, motor apparent power, instantaneous power, temperature difference and the like.
S14: target input feature set data is obtained from the first input feature set data and the second input feature set data.
The method comprises the steps of preprocessing current operation data to obtain first input feature set data, carrying out feature processing on the first input feature set data to obtain second input feature set data, further carrying out data processing on the first input feature set data and the second input feature set data, obtaining target input feature set data from the first input feature set data and the second input feature set data, and enabling the target input feature set data to be used for inputting a temperature prediction model of the permanent magnet synchronous motor.
S15: and inputting the target input feature set data into a fusion model to obtain the current predicted temperature of the permanent magnet synchronous motor so as to realize the temperature prediction of the permanent magnet synchronous motor, wherein the fusion model is constructed based on a tree model and a gating and circulating unit.
The fusion model is constructed based on a tree model and a door control circulation unit, for example, a temperature prediction model of the permanent magnet synchronous motor is obtained by fusing a GRU (Gate Recurrent Unit) neural network model with two tree models, namely LightGBM (Light Gradient Boosting Machine) and Catboost. And inputting the target input feature set data into the fusion model for operation, so that the current predicted temperature of the permanent magnet synchronous motor can be obtained, and the temperature prediction of the permanent magnet synchronous motor is realized.
In the embodiment, the current running data of the permanent magnet synchronous motor is obtained when the vehicle runs, the current running data is processed to obtain the first input feature set data and the second input feature set data, the target input feature set data is obtained from the first input feature set data and the second input feature set data, the target input feature set data is input into a fusion model constructed based on a tree model and a door control circulation unit, the current prediction temperature of the permanent magnet synchronous motor is obtained, the accuracy, stability and robustness of the temperature prediction model are improved, and the accurate temperature prediction of the permanent magnet synchronous motor is realized.
In some embodiments, preprocessing the current operational data to obtain first input feature set data includes: filling the missing value of the current operation data by using a Lagrangian interpolation method; responding to completion of filling of the missing values, dividing current operation data to obtain a plurality of driving fragments; and removing abnormal values in the plurality of driving fragments to obtain first input feature set data.
By Lagrange interpolationFilling the missing value in the current operation data, wherein the missing value is caused by the loss of sensor transmission, filling the missing value by adopting a Lagrange interpolation method, and utilizing n+1 discrete pointsThe interpolation polynomial Ln (x) can be obtained by an interpolation function.
In particular, the method comprises the steps of,
ω n+1 (x)＝(x-x 0 )(x-x 1 )…(x-x n )＝(x i -x 0 )…(x i -x i-1 )(x i -x i+1 )…(x i -x n )
wherein: l (L) i (x) Is an interpolation function; omega n+1 (x) Is an interpolation polynomial. And solving a Lagrange interpolation polynomial by using the known number, and substituting the node to be solved into the polynomial to solve the target value.
In response to completing filling of the missing values in the current operation data, a plurality of running segments can be obtained by dividing the current operation data, and as the data collected by the sensor can be subjected to environmental interference such as electromagnetic interference, mechanical vibration, environmental change and the like, sensor readings under abnormal working conditions with problems can be abnormal in the collecting and processing processes of sensor signals, further abnormal values in the running segments need to be removed, so that first input feature set data can be obtained, for example, the first input feature set data comprise a given current/voltage of a rear motor D shaft, a given current/voltage of a rear motor Q shaft, a given rear motor rotating speed, a water inlet temperature, torque and motor temperature, and the specific table is shown below.
TABLE 1
In some embodiments, dividing the current operational data to obtain a plurality of travel segments includes: deleting the operation fragments corresponding to the target parameters of the current operation data as the first preset value, taking the operation fragments as intervals, and dividing the current operation data into a plurality of driving fragments.
In different running data of the vehicle motor, the running mode and the characteristic performance of the motor have larger difference, and the running segments need to be divided, namely the current running data are divided, so that a plurality of running segments can be obtained. Specifically, deleting the operation segment corresponding to the target parameter of the current operation data as the first preset value, taking the operation segment as an interval, dividing the current operation data into a plurality of driving segments, namely removing the segment corresponding to the point of which the motor rotating speed and the torque are fed back to be 0 as the motor rotating speed and the torque can represent the driving state and the stopping state of the automobile, taking the removed segment as a separation, and dividing the current operation data into a plurality of driving segments.
Removing outliers in the plurality of driving segments, comprising: deleting continuous unchanged operation data in a preset time period in a plurality of driving fragments; calculating characteristic data abnormal values of each driving segment in the driving segments, and deleting the driving segments corresponding to the data abnormal values larger than the second preset value.
And deleting the running data which are continuously unchanged in a preset time period in the plurality of running segments, wherein the preset time period is 2 minutes, namely deleting points in which characteristic variables in the divided running segments are continuously unchanged for more than 2 minutes. Calculating abnormal values of characteristic data of each of the plurality of driving segments, deleting the driving segments corresponding to the abnormal values of the data larger than a second preset value, for example, calculating an average value mu and a standard deviation sigma of any characteristic data of one driving segment, and defining a threshold range of the abnormal values: mu+/-3 sigma, so as to reject the data segment exceeding the threshold, for example, the characteristic data of the driving segment comprises temperature, current, voltage, torque and the like, and then abnormal values corresponding to different characteristic data are calculated respectively, and the abnormal values are removed according to rules.
In some embodiments, the first input feature set data includes temperature feature data, and the feature processing is performed on the first input feature set data to obtain second input feature set data, including: performing feature diffraction based on the first input feature set data to generate derivative feature set data; and calculating by using the temperature characteristic data to obtain temperature difference characteristic data, and further forming second input characteristic set data by the derivative characteristic set data and the temperature difference characteristic data.
Feature derivation is performed based on the first input feature set data, generating derivative feature set data. Specifically, the derivative features are as follows:
S el ＝i s ·u s
P el ＝i d ·u d +i q ·u q
wherein i is s ，u s Is the combined value of current and voltage, S el For the apparent power of the motor, the total load and the energy consumption condition of the circuit are reflected, P el For instantaneous power, the energy actually converted into useful work by the circuit at the current moment is reflected, i.e. the derivative feature set data comprises a current composite value i s Voltage composite value u s Apparent motor power S el Instantaneous power P el 。
The temperature characteristic data is used for calculation to obtain temperature difference characteristic data, for example, the initial motor temperature T of each driving segment can be calculated m Difference from the predicted temperature at the previous time is taken as a temperature difference characteristic T diff For example, the initial motor temperatures of 3 driving segments are acquired, the motor temperature at the i-th moment needs to be predicted, and three temperature difference characteristics are obtained by making the difference between the initial motor temperatures of the 3 segments and the predicted motor temperature at the i-1-th moment.
Further, the second input feature set data may be composed of derivative feature set data and temperature difference feature data, and the second input feature set data may include a current synthesized value, a voltage synthesized value, a motor apparent power, an instantaneous power, and a temperature difference, as shown in the following table.
TABLE 2
In some embodiments, deriving target input feature set data from the first input feature set data and the second input feature set data comprises: filtering the data features in the first input feature set data and the second input feature set data by using an exponential weighted moving average to obtain feature data to be correlated; in response to completion of the filtering process, performing correlation coefficient calculation to obtain a correlation coefficient corresponding to the feature data to be correlated; judging based on the correlation coefficient, and carrying out normalization processing on the characteristic data corresponding to the correlation coefficient meeting the preset condition to obtain target input characteristic set data.
Carrying out EWMA filtering processing on the data features in the first input feature set data and the second input feature set data by using an exponential weighted moving average so as to carry out smoothing processing on time sequence data and extract long-term trend, thereby reducing the influence of noise and abnormal values on data analysis, namely carrying out filtering processing on the data features by using the exponential weighted moving average, wherein the formula is as follows:
y t ＝αx t +(1-α)y t-1
wherein y is t Represents an exponentially weighted moving average at time t, y t-1 Representing an exponentially weighted moving average at time t-1, alpha is a smoothing factor, is an intermediate value between [0,1]]Parameters in between, e.g., α=0.7.
And responding to the completion of the filtering processing to obtain feature data to be correlated, wherein the feature data to be correlated is the first input feature set data and the second input feature set data subjected to the filtering processing, and further performing correlation coefficient calculation, such as pearson correlation calculation, to obtain correlation coefficients corresponding to the feature data to be correlated, and judging based on the correlation coefficients to perform normalization processing on the feature data corresponding to the correlation coefficients meeting preset conditions to obtain target input feature set data.
In some embodiments, the correlation coefficient calculation includes: taking the predicted temperature characteristic as a reference sequence, and calculating a correlation coefficient between a characteristic sequence corresponding to the characteristic data to be correlated and the reference sequence; judging based on the correlation coefficient, and carrying out normalization processing on characteristic data corresponding to the correlation coefficient meeting the preset condition, wherein the method comprises the following steps: sorting the correlation coefficients according to the values to determine characteristic data corresponding to the correlation coefficients meeting preset conditions; and carrying out normalization processing on the characteristic data corresponding to the correlation coefficient meeting the preset condition by using a dispersion normalization method, and further obtaining target input characteristic set data.
Carrying out correlation coefficient calculation on the feature data to be correlated, for example, carrying out pearson correlation calculation, taking the predicted temperature feature as a reference sequence, calculating the correlation coefficient between the feature sequence corresponding to the feature data to be correlated and the reference sequence, namely, taking the predicted motor temperature as a reference sequence Y, taking the feature in the feature data to be correlated as the feature sequence, and calculating the pearson correlation coefficient between the feature sequence and the reference sequence Y, wherein the formula is as follows:
wherein x and Y respectively represent the characteristic and the predicted motor temperature values,and r is a correlation coefficient, and the value range is (-1, 1) and is the mean value of the feature sequence. And calculating the correlation coefficient of each characteristic and the motor temperature according to r.
The pearson correlation characteristic calculation results in an embodiment of the present application are shown in the following table.
TABLE 3 Table 3
Further, the methodSequencing the relative numbers according to the values to determine feature data corresponding to the correlation coefficients meeting the preset conditions, wherein the correlation coefficients meeting the preset conditions can be pearson correlation coefficients with larger index values, taking the pearson correlation feature calculation result as an example, the feature with larger correlation coefficient values can be selected as the input feature of the model, and i can be seen in the table s ，p el ，m，i q ，u d ，i d If the correlation with the motor temperature is not strong, deleting the part of the characteristics can be considered, so that the data volume is reduced, and the prediction effect of the model is improved.
Normalizing the feature data corresponding to the correlation coefficient meeting the preset condition by using a dispersion normalization method, so as to obtain target input feature set data, for example, normalizing the feature data before and after the pearson correlation calculation meeting the preset condition to reduce the error of the obtained result, for example, mapping the feature data meeting the condition between [0,1] by using a dispersion normalization method, namely a min-max normalization method, wherein the formula is as follows:
wherein x is max X is the maximum value of the sample data min X' is the normalized sample data value, which is the minimum value of the sample data.
In some embodiments, the target input feature set data includes first target input feature subset data derived from the first input feature set data and second target input feature subset data derived from the second input feature set data.
Filtering data features in the first input feature set data and the second input feature set data to obtain feature data to be correlated, calculating correlation coefficients based on the feature data to be correlated, and normalizing feature data corresponding to the correlation coefficients meeting preset conditions to obtain target input feature set data, wherein the first target input feature is obtained from the first input feature set dataSubset data, e.g. comprising a rear motor speed n with a large pearson correlation coefficient with the predicted temperature mech Temperature T of motor m Giving u to the Q-axis voltage of the rear motor q Second target input feature subset data derived from the second input feature set data, e.g. comprising motor apparent power S with a large pearson correlation coefficient with predicted temperature el Voltage composite value u s And a temperature difference T diff 。
The fusion model comprises a first-layer tree model, a second-layer tree model and a third-layer neural network model.
The first layer of tree model can be a LightGBM tree model, the second layer of tree model can be a Catboost tree model, and the third layer of neural network model can be a GRU neural network model, wherein the LightGBM uses a histogram algorithm to accelerate the training process, so that the calculation cost is greatly reduced while the accuracy is maintained, the multi-thread parallel training is supported, and the model training speed can be accelerated on a multi-core CPU. The CatBoost can automatically process category characteristics without carrying out preprocessing operations such as thermal independent coding and the like, provides functions such as characteristic importance assessment, tree structure visualization and the like, can help a user to understand a decision process of a model, is a variant of a circulating neural network, is used for processing sequence data and time sequence data, effectively solves the problems of gradient elimination and gradient explosion in a traditional network by introducing a gating mechanism, enables the model to better capture long-term dependency relationship in the sequence, and has simpler structure, fewer parameters and higher training efficiency compared with the traditional long-term and short-term memory neural network. The fusion model is a technology for combining a LightGBM tree model, a Catboost tree model and a GRU neural network model based on model fusion to improve overall prediction performance and generalization capability, as shown in fig. 2, and fig. 2 is a schematic structural diagram of the fusion model according to an embodiment of the present application.
Inputting the target input feature set data into the fusion model to obtain the current predicted temperature of the permanent magnet synchronous motor, wherein the method comprises the following steps: inputting first target input feature subset data into a first layer tree model, and outputting first prediction result data; inputting the first prediction result data and the second target input feature subset data into a second layer tree model, and outputting second prediction result data; inputting the first target input feature subset data and the second target input feature subset data into a third layer neural network model to obtain third prediction result data; and carrying out weighted fusion on the first predicted result data, the second predicted result data and the third predicted result data to obtain the predicted temperature of the permanent magnet synchronous motor.
Inputting the first target input feature subset data into a first layer tree model, outputting first prediction result data, namely the motor rotating speed n mech Temperature T of motor m Giving u to the Q-axis voltage of the rear motor q As an input feature of the first layer LightGBM tree model, obtaining first prediction result data f lgb (x i ) Inputting the first predicted result data and the second target input feature subset data into a second layer tree model, namely the first predicted result data f lgb (x i ) And motor apparent power S el Voltage composite value u s And a temperature difference T diff As an input feature of the second-layer Catboost tree model, outputting second prediction result data f cat (x i ). Inputting the first target input feature subset data and the second target input feature subset data into a third layer neural network model, namely the motor rotating speed n mech Temperature T of motor m Q-axis voltage of rear motor is given by u q Apparent motor power S el Voltage composite value u s And a temperature difference T diff And as an input characteristic of the third-layer GRU neural network model, third prediction result data can be obtained.
Further, the first prediction result data, the second prediction result data and the third prediction result data are subjected to weighted fusion, so that the prediction temperature of the permanent magnet synchronous motor is obtained.
Finally, the predicted results of the LightGBM, catBoost and GRU models are weighted and fused to obtain a final regression predicted model, and the final predicted result is as follows:
wherein,f is a predicted value of the temperature of the motor lgb (x i ) F, the result predicted by the LightGBM tree model cat (x i ) F, predicted by Catboost tree model GRU (x i ) A, b, c are weight coefficients, respectively, and a+b+c=1, which are the results predicted by the GRU neural network model.
In some embodiments, the fusion model is super-parametrically adjusted using bayesian optimization to optimize updating the fusion model; and using the mean square error and the mean absolute value error as a loss function of the fusion model.
And performing super-parameter adjustment on the fusion model by using Bayesian optimization to optimize and update the fusion model, namely performing parameter adjustment on the model super-parameters by using Bayesian optimization, wherein the Bayesian optimization adopts a Gaussian process, and performs updating on the model by using the latest observation result, iterates repeatedly, evaluates an objective function and updates the model so as to achieve an expected target.
The mean square error and the average absolute value error are used as the loss function of the fusion model, the mean square error MSE (mean squared error) and the average absolute value error MAE (mean absolute error) are used as the loss function of the model, and the calculation formula is as follows:
wherein y is i Andthe true value and the predicted value of the motor temperature, respectively,/->The average value of the temperature of the motor is true, and n is the number of samples.
Specifically, the training process of the fusion model of the present application is exemplified, for example, using a motor temperature dataset of 50 vehicles for one month for model training and testing. The dataset includes historical temperature information and corresponding real temperature labels. The data set is divided into a training set for training of the model and a test set for evaluating the performance of the model. The training set contains motor temperature data of 40 vehicles, the test set contains motor temperature data of 10 vehicles, and the obtained prediction results are shown in the following table.
TABLE 4 prediction results for different vehicles under model
It will be appreciated by those skilled in the art that in the above-described method of the specific embodiments, the written order of steps is not meant to imply a strict order of execution but rather should be construed according to the function and possibly inherent logic of the steps.
Referring to fig. 3, fig. 3 is a schematic structural diagram of an electronic device according to an embodiment of the present application. The electronic device 30 comprises a memory 31 and a processor 32 coupled to each other, the processor 32 being configured to execute program instructions stored in the memory 31 to implement the steps of the above-described embodiment of the method for predicting the temperature of a permanent magnet synchronous motor on a vehicle. In one particular implementation scenario, electronic device 30 may include, but is not limited to: the microcomputer and the server are not limited herein.
Specifically, the processor 32 is configured to control itself and the memory 31 to implement the steps of the above-described temperature prediction method embodiment of the permanent magnet synchronous motor on the vehicle. The processor 32 may also be referred to as a CPU (Central Processing Unit ), and the processor 32 may be an integrated circuit chip with signal processing capabilities. The processor 32 may also be a general purpose processor, a digital signal processor (Digital Signal Processor, DSP), an application specific integrated circuit (Application Specific Integrated Circuit, ASIC), a Field programmable gate array (Field-Programmable Gate Array, FPGA) or other programmable logic device, discrete gate or transistor logic device, discrete hardware components. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like. In addition, the processor 32 may be commonly implemented by an integrated circuit chip.
Referring to fig. 4, fig. 4 is a schematic structural diagram of a non-volatile computer readable storage medium according to an embodiment of the present application. The non-transitory computer readable storage medium 40 is used to store a computer program 401, which computer program 401, when executed by a processor, for example by the processor 32 in the above-described embodiment of fig. 3, is used to implement the steps of the above-described embodiment of the temperature prediction method for a permanent magnet synchronous motor on a vehicle.
The foregoing description of various embodiments is intended to highlight differences between the various embodiments, which may be the same or similar to each other by reference, and is not repeated herein for the sake of brevity.
In the several embodiments provided in this application, it should be understood that the disclosed methods and related devices may be implemented in other ways. For example, the above-described embodiments of related devices are merely illustrative, e.g., the division of modules or elements is merely a logical functional division, and there may be additional divisions of actual implementation, e.g., elements or components may be combined or integrated into another system, or some features may be omitted, or not performed. Alternatively, the coupling or direct coupling or communication disconnection between the illustrated or discussed elements may be through some interface, indirect coupling or communication disconnection of a device or element, electrical, mechanical, or other form.
In addition, each functional unit in each embodiment of the present application may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit. The integrated units may be implemented in hardware or in software functional units.
The integrated units, if implemented in the form of software functional units and sold or used as stand-alone products, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present application may be embodied essentially or in part or all or part of the technical solution contributing to the prior art or in the form of a software product stored in a storage medium, including several instructions to cause a computer device (which may be a personal computer, a server, or a network device, etc.) or a processor (processor) to perform all or part of the steps of the methods of the embodiments of the present application. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
Those skilled in the art will readily appreciate that many modifications and variations are possible in the device and method while maintaining the teachings of the present application. Accordingly, the above disclosure should be viewed as limited only by the scope of the appended claims.
Claims (10)
1. A method for predicting the temperature of a permanent magnet synchronous motor on a vehicle, comprising:
acquiring current operation data of the permanent magnet synchronous motor when the vehicle runs;
preprocessing the current operation data to obtain first input feature set data;
performing feature processing on the first input feature set data to obtain second input feature set data;
obtaining target input feature set data from the first input feature set data and the second input feature set data;
and inputting the target input feature set data into a fusion model to obtain the current predicted temperature of the permanent magnet synchronous motor so as to realize the temperature prediction of the permanent magnet synchronous motor, wherein the fusion model is constructed based on a tree model and a door control circulation unit.
2. The method of claim 1, wherein the target input feature set data comprises first target input feature subset data derived from the first input feature set data and second target input feature subset data derived from the second input feature set data, the fusion model comprising a first layer tree model, a second layer tree model, and a third layer neural network model;
inputting the target input feature set data into a fusion model to obtain the current predicted temperature of the permanent magnet synchronous motor, wherein the method comprises the following steps:
inputting the first target input feature subset data into the first layer tree model, and outputting first prediction result data;
inputting the first prediction result data and the second target input feature subset data into the second layer tree model, and outputting second prediction result data;
inputting the first target input feature subset data and the second target input feature subset data into a third layer neural network model to obtain third prediction result data;
and carrying out weighted fusion on the first predicted result data, the second predicted result data and the third predicted result data to obtain the predicted temperature of the permanent magnet synchronous motor.
3. The method of claim 1, wherein the first input feature set data comprises temperature feature data, wherein performing feature processing on the first input feature set data to obtain second input feature set data comprises:
performing feature diffraction based on the first input feature set data to generate derivative feature set data;
and calculating by using the temperature characteristic data to obtain temperature difference characteristic data, and further forming the second input characteristic set data by the derivative characteristic set data and the temperature difference characteristic data.
4. A method according to claim 3, wherein said deriving target input feature set data from said first input feature set data and said second input feature set data comprises:
filtering the data features in the first input feature set data and the second input feature set data by using an exponential weighted moving average to obtain feature data to be correlated;
in response to completion of the filtering process, performing correlation coefficient calculation to obtain a correlation coefficient corresponding to the feature data to be correlated;
and judging based on the correlation coefficient, and carrying out normalization processing on the characteristic data corresponding to the correlation coefficient meeting the preset condition to obtain target input characteristic set data.
5. The method of claim 4, wherein the correlation coefficient calculation comprises:
taking the predicted temperature characteristic as a reference sequence, and calculating a correlation coefficient between a characteristic sequence corresponding to the characteristic data to be correlated and the reference sequence;
the judging based on the correlation coefficient and the normalizing processing of the characteristic data corresponding to the correlation coefficient meeting the preset condition comprises the following steps:
sorting the correlation coefficients according to the values to determine characteristic data corresponding to the correlation coefficients meeting preset conditions;
and carrying out normalization processing on the characteristic data corresponding to the correlation coefficient meeting the preset condition by using a dispersion normalization method, so as to obtain target input characteristic set data.
6. The method of claim 1, wherein preprocessing the current operational data to obtain first input feature set data comprises:
filling the missing value of the current operation data by using a Lagrangian interpolation method;
dividing the current operation data to obtain a plurality of driving fragments in response to the completion of filling the missing values;
and removing abnormal values in the plurality of driving fragments to obtain the first input feature set data.
7. The method of claim 6, wherein the dividing the current operation data to obtain a plurality of driving segments comprises:
deleting the operation fragments corresponding to the target parameters of the current operation data as the first preset value, and dividing the current operation data into a plurality of driving fragments by taking the operation fragments as intervals;
the removing the outliers in the plurality of driving segments includes:
deleting continuous unchanged operation data in a preset time period in the plurality of driving fragments;
calculating characteristic data abnormal values of each driving segment in the plurality of driving segments, and deleting the driving segment corresponding to the data abnormal value larger than a second preset value.
8. The method as recited in claim 1, further comprising:
performing super-parameter adjustment on the fusion model by using Bayesian optimization to optimize and update the fusion model; and
and using the mean square error and the average absolute value error as a loss function of the fusion model.
9. An electronic device comprising a memory and a processor coupled to each other, the processor configured to execute program instructions stored in the memory to implement the method of predicting temperature of a permanent magnet synchronous motor on a vehicle according to any one of claims 1 to 8.
10. A non-transitory computer readable storage medium having stored thereon program instructions, which when executed by a processor, implement the method of predicting the temperature of a permanent magnet synchronous motor on a vehicle according to any one of claims 1 to 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311260222.2A CN117498762A (en) | 2023-09-26 | 2023-09-26 | Temperature prediction method and related equipment for permanent magnet synchronous motor on vehicle |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311260222.2A CN117498762A (en) | 2023-09-26 | 2023-09-26 | Temperature prediction method and related equipment for permanent magnet synchronous motor on vehicle |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117498762A true CN117498762A (en) | 2024-02-02 |
Family
ID=89678923
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311260222.2A Pending CN117498762A (en) | 2023-09-26 | 2023-09-26 | Temperature prediction method and related equipment for permanent magnet synchronous motor on vehicle |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117498762A (en) |
-
2023
- 2023-09-26 CN CN202311260222.2A patent/CN117498762A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112785016B (en) | New energy automobile maintenance and fault monitoring and diagnosis method based on machine learning | |
| CN112257263B (en) | Equipment residual life prediction system based on self-attention mechanism | |
| CN111967535A (en) | Fault diagnosis method and device for temperature sensor in grain storage management scene | |
| CN113608916A (en) | Fault diagnosis method and device, electronic equipment and storage medium | |
| CN113225209A (en) | Network monitoring real-time early warning method based on time series similarity retrieval | |
| CN115062553B (en) | Water pump working condition degradation detection method, equipment and medium based on multi-model fusion | |
| CN116400244B (en) | Abnormality detection method and device for energy storage battery | |
| CN114358077A (en) | Permanent magnet synchronous motor demagnetization fault diagnosis model construction method and fault diagnosis method and system | |
| CN115291545A (en) | Multi-layer control system and method for power electronic product | |
| CN114581261A (en) | Fault diagnosis method, system, equipment and storage medium based on quick graph calculation | |
| WO2024198109A1 (en) | Motor temperature estimation method and device based on gpr-rc model | |
| CN113793076A (en) | Dynamic risk pool monitoring method, system, equipment and readable storage medium | |
| CN111971200A (en) | Method for determining the state of ageing of a battery, computer program, storage device, control device and vehicle | |
| CN117498762A (en) | Temperature prediction method and related equipment for permanent magnet synchronous motor on vehicle | |
| CN116404951A (en) | Motor temperature estimation method and device based on GP-NARX model | |
| KR101490636B1 (en) | Defect classification method and apparatus of induction motors | |
| CN116432117A (en) | Motor fault classification method and device based on multi-source information fusion | |
| Li et al. | A Thermal-Aware Digital Twin Model of Permanent Magnet Synchronous Motors (PMSM) Based on BP Neural Networks | |
| Zuo et al. | Bearing fault dominant symptom parameters selection based on canonical discriminant analysis and false nearest neighbor using GA filtering signal | |
| CN113395031B (en) | Complex system fault diagnosis method and system based on multistage model | |
| CN115392331A (en) | Battery classification method and device, storage medium and electronic equipment | |
| CN118606731B (en) | Tunnel deformation prediction method and system | |
| Hou et al. | Bearing fault diagnosis based on variational mode decomposition and modified CNN | |
| CN118114014A (en) | Model training method for motor temperature prediction, and related method and apparatus | |
| CN117933342A (en) | Method for constructing prediction model and data prediction method |
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 |