CN117614323B - Brushless motor rotation speed control method, device, equipment and storage medium - Google Patents
Brushless motor rotation speed control method, device, equipment and storage medium Download PDFInfo
- Publication number
- CN117614323B CN117614323B CN202410095741.6A CN202410095741A CN117614323B CN 117614323 B CN117614323 B CN 117614323B CN 202410095741 A CN202410095741 A CN 202410095741A CN 117614323 B CN117614323 B CN 117614323B
- Authority
- CN
- China
- Prior art keywords
- motor
- target
- motor operation
- feature
- data
- 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 79
- 238000003860 storage Methods 0.000 title claims abstract description 15
- 230000005540 biological transmission Effects 0.000 claims abstract description 59
- 238000005457 optimization Methods 0.000 claims abstract description 54
- 238000011217 control strategy Methods 0.000 claims abstract description 28
- 230000009467 reduction Effects 0.000 claims abstract description 25
- 238000012360 testing method Methods 0.000 claims abstract description 17
- 230000008030 elimination Effects 0.000 claims abstract description 14
- 238000003379 elimination reaction Methods 0.000 claims abstract description 14
- 239000013598 vector Substances 0.000 claims description 86
- 238000004422 calculation algorithm Methods 0.000 claims description 51
- 230000015654 memory Effects 0.000 claims description 39
- 238000004458 analytical method Methods 0.000 claims description 33
- 230000007246 mechanism Effects 0.000 claims description 27
- 238000003066 decision tree Methods 0.000 claims description 25
- 238000000605 extraction Methods 0.000 claims description 24
- 230000002068 genetic effect Effects 0.000 claims description 21
- 238000012545 processing Methods 0.000 claims description 21
- 238000007637 random forest analysis Methods 0.000 claims description 19
- 238000006243 chemical reaction Methods 0.000 claims description 17
- 230000000694 effects Effects 0.000 claims description 16
- 230000006870 function Effects 0.000 claims description 14
- 230000003044 adaptive effect Effects 0.000 claims description 10
- 239000011159 matrix material Substances 0.000 claims description 10
- 230000002457 bidirectional effect Effects 0.000 claims description 9
- 230000004927 fusion Effects 0.000 claims description 8
- 230000035772 mutation Effects 0.000 claims description 7
- 238000010187 selection method Methods 0.000 claims description 6
- 238000004140 cleaning Methods 0.000 claims description 5
- 238000013507 mapping Methods 0.000 claims description 4
- 238000012163 sequencing technique Methods 0.000 claims description 3
- 230000006403 short-term memory Effects 0.000 claims description 3
- 238000013461 design Methods 0.000 abstract description 8
- 238000013135 deep learning Methods 0.000 abstract description 2
- 230000008569 process Effects 0.000 description 33
- 230000002829 reductive effect Effects 0.000 description 9
- 238000004364 calculation method Methods 0.000 description 8
- 238000012544 monitoring process Methods 0.000 description 5
- 230000008901 benefit Effects 0.000 description 3
- 230000008859 change Effects 0.000 description 3
- 238000010606 normalization Methods 0.000 description 3
- 230000000717 retained effect Effects 0.000 description 3
- 239000002699 waste material Substances 0.000 description 3
- 238000013459 approach Methods 0.000 description 2
- 238000003491 array Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000007613 environmental effect Effects 0.000 description 2
- 230000004907 flux Effects 0.000 description 2
- 238000009472 formulation Methods 0.000 description 2
- 230000007774 longterm Effects 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 239000000203 mixture Substances 0.000 description 2
- 238000013021 overheating Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 238000007619 statistical method Methods 0.000 description 2
- 230000009466 transformation Effects 0.000 description 2
- 230000001131 transforming effect Effects 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 238000001816 cooling Methods 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 238000013136 deep learning model Methods 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 238000005265 energy consumption Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000007499 fusion processing Methods 0.000 description 1
- 230000008303 genetic mechanism Effects 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000000670 limiting effect 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
- 238000007500 overflow downdraw method Methods 0.000 description 1
- 108090000623 proteins and genes Proteins 0.000 description 1
- 230000000306 recurrent effect Effects 0.000 description 1
- 230000002441 reversible effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- 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
- H02P6/00—Arrangements for controlling synchronous motors or other dynamo-electric motors using electronic commutation dependent on the rotor position; Electronic commutators therefor
- H02P6/34—Modelling or simulation for control purposes
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B13/00—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
- G05B13/02—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric
- G05B13/0265—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric the criterion being a learning criterion
- G05B13/027—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric the criterion being a learning criterion using neural networks only
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B13/00—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
- G05B13/02—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric
- G05B13/04—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators
- G05B13/042—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators in which a parameter or coefficient is automatically adjusted to optimise the performance
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B13/00—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
- G05B13/02—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric
- G05B13/04—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators
- G05B13/048—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators using a predictor
-
- 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
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/24323—Tree-organised classifiers
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- 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
- G06N3/0442—Recurrent networks, e.g. Hopfield networks characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]
-
- 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
- G06N3/12—Computing arrangements based on biological models using genetic models
- G06N3/126—Evolutionary algorithms, e.g. genetic algorithms or genetic programming
-
- 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
- H02P23/00—Arrangements or methods for the control of AC motors characterised by a control method other than vector control
- H02P23/0004—Control strategies in general, e.g. linear type, e.g. P, PI, PID, using robust control
- H02P23/0018—Control strategies in general, e.g. linear type, e.g. P, PI, PID, using robust control using neural 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
- H02P23/00—Arrangements or methods for the control of AC motors characterised by a control method other than vector control
- H02P23/0004—Control strategies in general, e.g. linear type, e.g. P, PI, PID, using robust control
- H02P23/0022—Model reference adaptation, e.g. MRAS or MRAC, useful for control or parameter estimation
-
- 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
- H02P23/00—Arrangements or methods for the control of AC motors characterised by a control method other than vector control
- H02P23/14—Estimation or adaptation of motor parameters, e.g. rotor time constant, flux, speed, current or voltage
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- General Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Biophysics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Power Engineering (AREA)
- Molecular Biology (AREA)
- Biomedical Technology (AREA)
- Computational Linguistics (AREA)
- Automation & Control Theory (AREA)
- General Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Medical Informatics (AREA)
- Mathematical Physics (AREA)
- Genetics & Genomics (AREA)
- Physiology (AREA)
- Control Of Motors That Do Not Use Commutators (AREA)
Abstract
The application relates to the technical field of deep learning, and discloses a rotating speed control method, device and equipment of a brushless motor and a storage medium. The method comprises the following steps: initializing motor parameters to obtain an initial motor parameter set; the operation test and the data acquisition obtain initial motor operation data and perform data dimension reduction to obtain target motor operation data; extracting features to obtain a plurality of original motor operation features, and performing recursive feature elimination to obtain a plurality of target motor operation features; performing feature classification to obtain a first motor operation feature set and a second motor operation feature set; carrying out transmission performance prediction and Halbach array performance prediction through a motor performance prediction model to obtain a transmission performance predicted value and a Halbach array performance predicted value; bayesian optimization is performed to obtain a target motor parameter set and create a self-adaptive rotating speed control strategy, so that the design parameters of the brushless motor are optimized, and the rotating speed control accuracy of the brushless motor is improved.
Description
Technical Field
The present disclosure relates to the field of deep learning technologies, and in particular, to a method, an apparatus, a device, and a storage medium for controlling a rotational speed of a brushless motor.
Background
In the technical field of motor driving, brushless motors are widely applied to the fields of industrial manufacture, electric vehicles, aerospace and the like, and the performance quality of the brushless motors directly influences the efficiency and stability of a system. Conventional motor control methods often rely on manually adjusted parameters or model-based control strategies, which are difficult to achieve optimal performance in complex and diverse operating environments. Therefore, the self-adaptive rotating speed control strategy becomes a research hot spot, and the performance of the motor is optimized in a data driving mode so as to adapt to different working conditions, and the self-adaptive rotating speed control method has a wide application prospect.
However, current adaptive speed control methods still face some challenges and problems. Initialization and optimization of motor parameters typically requires extensive computation and data acquisition, which increases system complexity and cost. Secondly, how to accurately select the feature with the most information content is still a problem to be solved in the process of feature extraction and selection by considering the influence of a plurality of original features. The design and optimization of motor performance predictive models also requires extensive research to ensure their accuracy and generalization capability, particularly under different operating conditions. That is, the accuracy of the prior art is low.
Disclosure of Invention
The application provides a brushless motor rotational speed control method, device, equipment and storage medium.
In a first aspect, the present application provides a method for controlling a rotational speed of a brushless motor, the method for controlling a rotational speed of a brushless motor including:
initializing motor parameters of a target brushless motor through a preset genetic algorithm to obtain an initial motor parameter set;
performing operation test and data acquisition on the target brushless motor based on the initial motor parameter set to obtain initial motor operation data, and performing data dimension reduction on the initial motor operation data based on two-dimensional inter-class boundary Fisher analysis to obtain target motor operation data;
extracting the characteristics of the target motor operation data to obtain a plurality of original motor operation characteristics, and performing recursive characteristic elimination on the plurality of original motor operation characteristics to obtain a plurality of target motor operation characteristics;
performing feature classification on the plurality of target motor operation features through a random forest algorithm to obtain a first motor operation feature set and a second motor operation feature set;
Inputting the first motor operation characteristic set and the second motor operation characteristic set into a preset motor performance prediction model to predict transmission performance and Halbach array performance, so as to obtain a transmission performance predicted value and a Halbach array performance predicted value;
and performing Bayesian optimization on the initial motor parameter set according to the transmission performance predicted value and the Halbach array performance predicted value to obtain a target motor parameter set, and creating a self-adaptive rotating speed control strategy of the target brushless motor according to the target motor parameter set.
In a second aspect, the present application provides a rotational speed control apparatus of a brushless motor, the rotational speed control apparatus of a brushless motor including:
the initialization module is used for initializing motor parameters of the target brushless motor through a preset genetic algorithm to obtain an initial motor parameter set;
the testing module is used for carrying out operation testing and data acquisition on the target brushless motor based on the initial motor parameter set to obtain initial motor operation data, and carrying out data dimension reduction on the initial motor operation data based on two-dimensional inter-class boundary Fisher analysis to obtain target motor operation data;
The feature extraction module is used for extracting features of the target motor operation data to obtain a plurality of original motor operation features, and performing recursive feature elimination on the plurality of original motor operation features to obtain a plurality of target motor operation features;
the feature classification module is used for classifying the features of the plurality of target motor operation features through a random forest algorithm to obtain a first motor operation feature set and a second motor operation feature set;
the prediction module is used for inputting the first motor operation characteristic set and the second motor operation characteristic set into a preset motor performance prediction model to perform transmission performance prediction and Halbach array performance prediction, so as to obtain a transmission performance prediction value and a Halbach array performance prediction value;
the creating module is used for carrying out Bayesian optimization on the initial motor parameter set according to the transmission performance predicted value and the Halbach array performance predicted value to obtain a target motor parameter set, and creating a self-adaptive rotating speed control strategy of the target brushless motor according to the target motor parameter set.
A third aspect of the present application provides a computer device comprising: a memory and at least one processor, the memory having instructions stored therein; the at least one processor invokes the instructions in the memory to cause the computer device to perform the method of controlling the rotational speed of the brushless motor described above.
A fourth aspect of the present application provides a computer-readable storage medium having instructions stored therein, which when run on a computer, cause the computer to perform the above-described method of controlling the rotational speed of a brushless motor.
According to the technical scheme, the accuracy of brushless motor rotating speed control can be improved through the method for initializing motor parameters and driving data by using the genetic algorithm. The sensor data acquisition and feature extraction further increase accurate monitoring of the motor operating state. The performance of the brushless motor is allowed to be optimized in actual operation through the transmission performance prediction and the Halbach array performance prediction. This means that the motor can achieve higher efficiency and performance under different operating conditions. By using a Bayesian optimization algorithm to adjust motor parameters, an adaptive rotational speed control strategy is realized. This means that the motor can adjust its parameters in real time under different conditions to meet performance requirements without manual intervention. The data dimension can be reduced, the calculation efficiency can be improved, and the most important motor operation characteristics can be selected, so that the model performance can be improved. By optimizing the motor parameters and the control strategy, the energy waste is reduced, the energy utilization rate is improved, and the running cost of the motor is reduced. Based on the intelligent control of data and algorithm, the brushless motor can adapt to different working environments and load requirements, so that the intelligent level of the system is improved, the design parameters of the brushless motor are optimized, and the accuracy of the rotation speed control of the brushless motor is improved.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and other drawings may be obtained based on these drawings without inventive effort for a person skilled in the art.
Fig. 1 is a schematic diagram of an embodiment of a method for controlling a rotational speed of a brushless motor according to an embodiment of the present application;
fig. 2 is a schematic diagram of an embodiment of a rotational speed control apparatus for a brushless motor according to an embodiment of the present application.
Detailed Description
The embodiment of the application provides a rotating speed control method, device and equipment of a brushless motor and a storage medium. The terms "first," "second," "third," "fourth" and the like in the description and in the claims of this application and in the above-described figures, if any, are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate such that the embodiments described herein may be implemented in other sequences than those illustrated or otherwise described herein. Furthermore, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed or inherent to such process, method, article, or apparatus.
For ease of understanding, a specific flow of an embodiment of the present application will be described below, referring to fig. 1, and one embodiment of a method for controlling a rotational speed of a brushless motor in an embodiment of the present application includes:
step 101, initializing motor parameters of a target brushless motor through a preset genetic algorithm to obtain an initial motor parameter set;
it is to be understood that the execution body of the present application may be a rotational speed control device of a brushless motor, and may also be a terminal or a server, which is not limited herein. The embodiment of the present application will be described by taking a server as an execution body.
Specifically, initializing motor parameters of a target brushless motor to obtain pole slot matching parameters, permanent magnet thickness parameters and notch width parameters. The pole and slot matching factor is calculated by the pole and slot matching parameter, and reflects the matching degree between the pole and the slot; the permanent magnet thickness effect factor is calculated through the permanent magnet thickness parameter, and characterizes the influence of the permanent magnet thickness on the motor performance; the slot width influence factor is calculated based on the slot width parameter and is used for evaluating the influence of the slot width on the motor efficiency. On the basis of the factors, the parameters are comprehensively considered and initialized through a preset genetic algorithm, and a plurality of first motor parameter sets are generated. Genetic algorithms optimize problem solutions by modeling natural selection and genetic mechanisms, and are applicable to such complex parameter optimization problems. And carrying out fitness analysis on the first motor parameter set to evaluate the contribution degree of each parameter set to motor performance. Fitness analysis is an important element in genetic algorithms, which determines which parameter sets are more suitable for motor requirements. By calculation of fitness values, the parameter sets are fitness ordered to form a sequence of motor parameter sets. To further optimize the parameters, this sequence was population partitioned, and the first 50% of candidate motor parameter populations were selected. This selection process ensures that only the most potential parameter sets are retained for subsequent operations. And performing crossover operation and mutation operation on the candidate motor parameter group. The crossover operation creates a new solution by combining the features of different parameter sets, while the mutation operation randomly adjusts the existing parameter sets to introduce new genetic diversity. The two operations work together to generate a plurality of second motor parameter sets. The roulette selection method is adopted to carry out parameter selection and updating on the second motor parameter sets, which is a probability selection method, so that the parameter updating process is ensured to take the advantages of the parameters into consideration, and certain randomness is maintained. An initial motor parameter set is obtained.
Step 102, performing operation test and data acquisition on a target brushless motor based on an initial motor parameter set to obtain initial motor operation data, and performing data dimension reduction on the initial motor operation data based on two-dimensional inter-class boundary Fisher analysis to obtain target motor operation data;
specifically, an operational test is performed on the target brushless motor based on the initial motor parameter set. These parameters include the rotational speed, torque, input power, etc. of the motor, which together determine the operating state and performance of the motor. During the operation test, the motor is fully operated with a preset sensor group, and the sensors include a temperature sensor, a vibration sensor, a speed sensor and the like, which work together to ensure that all-dimensional data about the operation of the motor is collected. And (3) cleaning the initial motor operation data to remove noise and irrelevant information, and ensuring the accuracy and effectiveness of subsequent analysis. And the first motor operation data is standardized, so that the data has a uniform format and scale, and the subsequent comparison and analysis are facilitated. And carrying out time sequence synchronization processing on the standardized second motor operation data to ensure the consistency of the data on a time axis, which is helpful for analyzing the dynamic characteristics of motor operation. Further, the standardized and time-sequence-synchronized motor operation data are classified, and key information such as motor temperature data, motor vibration data and the like is obtained. These data reflect the physical and environmental conditions of the motor during operation. On the basis, the temperature mean value and the temperature variance of the motor temperature data and the vibration mean value and the vibration variance of the motor vibration data are calculated respectively. These statistics provide a quantitative description of the motor operating conditions for the system, which is the basis for subsequent analysis. These statistics are processed using a two-dimensional inter-class boundary Fisher analysis. Fisher analysis is a powerful statistical method for preserving the most critical information during data dimension reduction. It helps the system identify which data features are most important for distinguishing between different operating states by calculating the degree of distinction between the different categories. By calculating Fisher discriminant criteria for temperature mean and temperature variance, vibration mean and vibration variance, it is revealed which statistics are most effective in distinguishing between different motor operating conditions. And generating a corresponding dimension reduction conversion matrix according to the Fisher discriminant. This transformation matrix is constructed based on the results of Fisher analysis, which is capable of transforming the raw multidimensional data into a much more compact and informative low-dimensional data. And carrying out data dimension reduction on the standard motor operation data by using the dimension reduction conversion matrix, wherein the target motor operation data obtained after dimension reduction contains the most critical information.
Step 103, extracting characteristics of the target motor operation data to obtain a plurality of original motor operation characteristics, and performing recursive characteristic elimination on the plurality of original motor operation characteristics to obtain a plurality of target motor operation characteristics;
specifically, the forward long-short time memory unit of the bidirectional long-short time network is used for analyzing the operation data of the target motor so as to extract the forward hidden feature set. A bi-directional long and short time network is an advanced recurrent neural network that is capable of efficiently processing and analyzing time series data, such as motor operation data. Through processing by the forward long and short duration memory unit, the network is able to capture patterns and trends in motor operation data that change over time, these being referred to as forward hidden features. These features reflect the behavior and performance of the motor under specific conditions and are key to understanding the motor's operating state. The same motor operation data is analyzed by a backward long short time memory unit in the bidirectional long short time network, but from the reverse direction of the time sequence, to extract the backward hidden feature set. This backward analysis can capture data patterns that change over time, with these backward hidden features providing an understanding of the motor operating state from another perspective. Through the combined use of the forward and backward long-short time memory units, the bidirectional long-short time can comprehensively analyze the motor operation data, and capture the complex modes and relations therein. And carrying out feature fusion on the forward hidden feature set and the backward hidden feature set to obtain a plurality of original motor operation features. These raw feature sets contain comprehensive information of the operation of the motor, but contain some less important or redundant features. In order to optimize these original features and to improve the accuracy and efficiency of the rotational speed control method, recursive feature elimination is performed. Recursive feature elimination is a feature selection technique that optimizes feature sets by progressively removing features that are considered least important. At each step, the model is trained and then the least significant features are removed, and the process is iterated until a preset number of features is reached or some stop condition is met. In this way, a plurality of target motor operating characteristics are obtained, which are a compact and efficient representation of the original motor operating data, providing key information for the rotational speed control of the brushless motor.
Step 104, classifying the characteristics of the operation characteristics of the multiple target motors through a random forest algorithm to obtain a first motor operation characteristic set and a second motor operation characteristic set;
specifically, the random forest algorithm is an integrated learning method, and the overall prediction accuracy and stability are improved by constructing a plurality of decision trees and integrating the prediction results of the decision trees. A plurality of target motor operating characteristics are input into a random forest algorithm. Each decision tree works independently in a random forest, analyzing and predicting a respective subset of data, thereby efficiently processing a large number of features and identifying the most important features. Based on the use of the features in the decision trees and their degree of contribution to the predicted results, each decision tree in the random forest performs an importance calculation on each target motor operating feature. The calculation results reveal which features are most critical for judging the running state of the motor. And carrying out weighted fusion on the feature importance obtained by each decision tree to obtain the comprehensive target importance of the operation features of each target motor. The weighted fusion considers the performance and reliability of each decision tree, and ensures that the final obtained target importance is comprehensive and accurate. The target importance is compared to a preset feature threshold to determine which features are sufficiently important. This comparison helps to distinguish which features are more critical to the operation and control of the motor. Based on the target comparison result, features are classified into two categories: the target importance is higher than or equal to the feature threshold and the target importance is lower than the feature threshold. For those features whose target importance is greater than or equal to the feature threshold, it is mapped to a first set of motor operating features. This set contains the features most critical to motor control, which will be used directly in subsequent motor control strategy formulation. And those features whose target importance level is below the feature threshold are mapped to a second set of motor operating features. This set of features, while having less impact on motor operation, may still provide auxiliary information for monitoring and analysis of motor status. The feature classification method based on the random forest algorithm not only improves the accuracy of feature selection, but also enhances the effectiveness of a motor control strategy. Through accurate discernment and the most critical characteristic of utilization, can monitor and control the operation of motor more effectively, and then promote the performance and the efficiency of motor, reduce the energy consumption, prolong the life of motor. In addition, the method has good generalization capability, and can adapt to motor operation under different types and conditions, so that a motor control system is more flexible and robust.
Step 105, inputting the first motor operation characteristic set and the second motor operation characteristic set into a preset motor performance prediction model to predict transmission performance and Halbach array performance, and obtaining a transmission performance predicted value and a Halbach array performance predicted value;
specifically, feature encoding and vector conversion are performed on the first motor operation feature set and the second motor operation feature set, and features are converted into a format which can be processed by a motor performance prediction model. Feature encoding and vector conversion are the conversion of actual operational data into a mathematical representation that can be efficiently processed by algorithms. After this conversion is completed, a first motor operation feature vector and a second motor operation feature vector are obtained. These feature vectors are input into a preset motor performance prediction model. This model includes two predictive task networks: a first predictive task network and a second predictive task network. Each network includes an input layer, an attention mechanism layer, a convolutional long and short time memory network, and a full connection layer. The design of this structure aims to improve the accuracy and adaptability of the predictive model. And carrying out vector normalization processing on the first motor operation feature vector through a first input layer to obtain a first standard operation feature vector. The normalization process is to ensure that the data is of the same scale in different dimensions. And carrying out attention mechanism weighting on the first standard operation feature vector through a first attention mechanism layer to obtain a first attention feature vector. Attention mechanisms were introduced to enable models to focus on those features that are most critical, thereby improving the accuracy of predictions. And the first convolution long-short time memory network performs high-dimensional feature extraction on the first attention feature vector to obtain a first high-dimensional feature vector. The combined use of the convolution long-short-term memory network can effectively capture long-term dependency and local characteristics in time series data, thereby providing support for complex prediction tasks. And carrying out transmission performance prediction on the first high-dimensional feature vector through the first full-connection layer to obtain a transmission performance predicted value. And similarly, for the second motor operation feature vector, carrying out standardization processing through a second input layer, weighting through a second attention mechanism layer, carrying out high-dimensional feature extraction through a second convolution long-short time memory network, and finally carrying out Halbach array performance prediction through a second full-connection layer to obtain a Halbach array performance prediction value.
And 106, performing Bayesian optimization on the initial motor parameter set according to the transmission performance predicted value and the Halbach array performance predicted value to obtain a target motor parameter set, and creating a self-adaptive rotating speed control strategy of the target brushless motor according to the target motor parameter set.
Specifically, a transmission ratio influence factor of the target brushless motor is calculated according to the transmission performance predicted value. The influence factor reflects the influence degree of the motor transmission ratio on the overall performance of the motor, and is one of key parameters to be considered in the optimization process. Meanwhile, calculating a Halbach array adjustment factor of the target brushless motor according to the Halbach array performance predicted value. Halbach arrays are a specific arrangement of magnetic fields, the adjustment factors of which indicate the degree of adjustment required for Halbach arrays under different performance requirements. And defining a torque optimization objective function of the target brushless motor by combining the transmission ratio influence factor and the Halbach array adjustment factor. This function is the core of the overall optimization process, which takes into account both the motor drive performance and the Halbach array tuning requirements to find the optimal motor operating parameters. And optimizing the initial motor parameter set through a preset Bayesian optimization algorithm. The Bayesian optimization algorithm is an optimization technology based on a probability model, can efficiently search an optimal solution in a parameter space, and is particularly suitable for solving the problems of complex processing and high calculation cost. Through this optimization process, gear ratio adjustment coefficients and Halbach array adjustment coefficients are obtained, which are key to adjusting motor parameters for specific operating conditions and performance requirements. And generating a corresponding target motor parameter set according to the transmission ratio adjustment coefficient and the Halbach array adjustment coefficient. This set contains optimized motor parameters that enable the motor to achieve optimal performance under specific operating conditions. And calculating torque optimization target values of the target motor parameter sets through a torque optimization target function, and verifying whether the optimized parameter sets can actually realize expected performance improvement. And creating an adaptive rotating speed control strategy of the target brushless motor according to the torque optimization target value and the target motor parameter set. The control strategy is formulated based on optimized parameters, which ensures that the motor maintains optimal performance under various operating conditions, while taking into account a balance of energy efficiency and life.
In the embodiment of the application, the accuracy of brushless motor rotating speed control can be improved by using the genetic algorithm to initialize motor parameters and a data driving method. The sensor data acquisition and feature extraction further increase accurate monitoring of the motor operating state. The performance of the brushless motor is allowed to be optimized in actual operation through the transmission performance prediction and the Halbach array performance prediction. This means that the motor can achieve higher efficiency and performance under different operating conditions. By using a Bayesian optimization algorithm to adjust motor parameters, an adaptive rotational speed control strategy is realized. This means that the motor can adjust its parameters in real time under different conditions to meet performance requirements without manual intervention. The data dimension can be reduced, the calculation efficiency can be improved, and the most important motor operation characteristics can be selected, so that the model performance can be improved. By optimizing the motor parameters and the control strategy, the energy waste is reduced, the energy utilization rate is improved, and the running cost of the motor is reduced. Based on the intelligent control of data and algorithm, the brushless motor can adapt to different working environments and load requirements, so that the intelligent level of the system is improved, the design parameters of the brushless motor are optimized, and the accuracy of the rotation speed control of the brushless motor is improved.
In a specific embodiment, the process of executing step 101 may specifically include the following steps:
(1) Initializing motor parameters of a target brushless motor to obtain pole slot matching parameters, permanent magnet thickness parameters and notch width parameters;
(2) Calculating a pole slot matching factor of a pole slot matching parameter, calculating a permanent magnet thickness effect factor of a permanent magnet thickness parameter, and calculating a slot width effect factor of a slot width parameter;
(3) Initializing parameter sets of pole slot matching parameters, permanent magnet thickness parameters and notch width parameters through a preset genetic algorithm according to the pole slot matching factors, the permanent magnet thickness effect factors and the notch width effect factors to obtain a plurality of first motor parameter sets;
(4) Carrying out adaptability analysis on a plurality of first motor parameter sets to obtain a adaptability value of each first motor parameter set;
(5) Performing fitness sequencing on a plurality of first motor parameter sets according to fitness values to obtain a motor parameter set sequence, and performing population division on the motor parameter set sequence to obtain a candidate motor parameter population of the first 50% in the motor parameter set sequence;
(6) Performing crossover operation and mutation operation on the candidate motor parameter groups to obtain a plurality of second motor parameter sets;
(7) And carrying out parameter selection and parameter updating on the plurality of second motor parameter sets based on the roulette selection method to obtain an initial motor parameter set.
Specifically, the target brushless motor is initialized with motor parameters, and basic operation characteristics of the motor are determined, including pole slot matching parameters, permanent magnet thickness parameters and slot width parameters. These parameters are critical to the performance and efficiency of the motor. For example, pole slot matching parameters affect the magnetic flux and torque production of the motor, permanent magnet thickness parameters relate to magnetic field strength and motor efficiency, and slot width parameters affect the magnetic flux distribution and noise level of the motor. The influencing factors of these parameters, namely the pole slot matching factor, the permanent magnet thickness effect factor and the slot width influencing factor, are calculated. For example, pole slot matching factors may be determined by analyzing the ratio of pole number to slot number and their effect on motor torque and efficiency. Similarly, the permanent magnet thickness effect factor and slot width effect factor also need to be calculated based on the physical characteristics of the permanent magnets and the specific impact of the slot design on motor performance. The parameters are set initialized using a preset genetic algorithm. Genetic algorithm is an optimization algorithm imitating biological evolution process, which finds the optimal solution of the problem by simulating natural selection, genetic and mutation processes. In this process, the pole slot matching parameter, the permanent magnet thickness parameter and the slot width parameter are taken as genes in a genetic algorithm, and a plurality of first motor parameter sets are formed by combinations thereof. These sets represent different operating states of the motor. In order to determine which parameter sets are most suitable for the requirements of the motor, fitness analysis is performed. Fitness analysis is the process of evaluating the impact of each parameter set on motor performance. This process involves evaluating how each set of parameters affects the efficiency, stability, and output of the motor. By means of fitness analysis, fitness values of each parameter set can be obtained, and the fitness values reflect the contribution degree of each parameter set to motor performance. And sequencing the first motor parameter sets according to the fitness value to form a motor parameter set sequence. The sequence is subjected to group division, and the first 50% of candidate motor parameter groups are selected. This selection process ensures that only the most potential parameter sets are retained for subsequent operations. And performing crossover operation and mutation operation on the candidate motor parameter groups. The crossover operation is to exchange the features of different parameter sets with each other to create new parameter combinations. The mutation operation is to randomly adjust the existing parameter set to introduce new genetic diversity. The two operations work together to produce a plurality of second sets of motor parameters. And selecting and updating the second motor parameter sets based on a roulette selection method to obtain initial motor parameter sets. Roulette selection is a probability selection method that determines the probability of being selected based on fitness values for each set of parameters. By the method, the initial motor parameter set finally obtained considers the advantages of parameters and maintains certain randomness and diversity.
In a specific embodiment, the process of executing step 102 may specifically include the following steps:
(1) Performing operation test on the target brushless motor based on the initial motor parameter set, and acquiring operation data of the target brushless motor through a preset sensor group to obtain initial motor operation data;
(2) Performing data cleaning on the initial motor operation data to obtain first motor operation data, and performing standardized processing on the first motor operation data to obtain second motor operation data;
(3) Performing time sequence synchronization processing on the second motor operation data to obtain standard motor operation data, and performing data classification on the standard motor operation data to obtain motor temperature data and motor vibration data;
(4) Respectively calculating a temperature mean value and a temperature variance of motor temperature data and a vibration mean value and a vibration variance of motor vibration data;
(5) Based on Fisher analysis of boundaries between two-dimensional classes, computing Fisher discriminant criteria of a temperature mean and a temperature variance, and a vibration mean and a vibration variance;
(6) And generating a corresponding dimension reduction conversion matrix according to the Fisher criterion, and carrying out data dimension reduction on the standard motor operation data according to the dimension reduction conversion matrix to obtain the target motor operation data.
Specifically, an operational test is performed on the target brushless motor based on the initial motor parameter set. These parameters include the rotational speed of the motor, torque settings, input voltage and current, etc., which together determine the operating state and performance of the motor. And in the running test process, the motor is comprehensively subjected to running data acquisition by using a preset sensor group. These sensors include temperature sensors, vibration sensors, speed sensors, etc. that work together to ensure that all-round data is collected about the operation of the motor. The collected initial motor operation data requires data cleaning to remove noise and irrelevant information. The purpose of the data cleaning is to ensure the accuracy and validity of the subsequent analysis. And the first motor operation data is subjected to standardized processing, so that the data has a uniform format and scale, and subsequent comparison and analysis are facilitated. And carrying out time sequence synchronization processing on the standardized second motor operation data to ensure the consistency of the data on a time axis, which is helpful for analyzing the dynamic characteristics of motor operation. Further, the standardized and time-sequence-synchronized motor operation data are classified, and key information such as motor temperature data, motor vibration data and the like is obtained. These data reflect the physical and environmental conditions of the motor during operation. For example, temperature data of the motor may reveal overheating problems thereof, while vibration data indicates mechanical failure or imbalance problems. On the basis, the temperature mean value and the temperature variance of the motor temperature data and the vibration mean value and the vibration variance of the motor vibration data are calculated respectively. These statistics provide a quantitative description of the motor operating conditions for the system, which is the basis for subsequent analysis. These statistics are processed using a two-dimensional inter-class boundary Fisher analysis. Fisher analysis is a powerful statistical method for preserving the most critical information during data dimension reduction. It helps the system identify which data features are most important for distinguishing between different operating states by calculating the degree of distinction between the different categories. By calculating Fisher discriminant criteria for temperature mean and temperature variance, vibration mean and vibration variance, it is revealed which statistics are most effective in distinguishing between different motor operating conditions. And generating a corresponding dimension reduction conversion matrix according to the Fisher discriminant. This transformation matrix is constructed based on the results of Fisher analysis, which is capable of transforming the raw multidimensional data into a much more compact and informative low-dimensional data. And carrying out data dimension reduction on the standard motor operation data through the dimension reduction conversion matrix, wherein the target motor operation data obtained after dimension reduction contains the most critical information. For example, if the temperature average of the motor is significantly higher than other conditions under a particular operating condition, this means that the risk of overheating the motor under such conditions increases, and the control strategy needs to be adjusted to reduce the temperature. Also, if the vibration variance increases under certain conditions, which indicates that the motor is not operating stably under such conditions, adjustments are needed to reduce vibration.
In a specific embodiment, the process of executing step 103 may specifically include the following steps:
(1) Forward hidden feature extraction is carried out on the operation data of the target motor through a forward long-short-time memory unit in a preset bidirectional long-short-time memory network, so that a forward hidden feature set is obtained;
(2) Backward hidden characteristic extraction is carried out on the operation data of the target motor through a backward long-short-time memory unit in the bidirectional long-short-time memory network, so that a backward hidden characteristic set is obtained;
(3) Performing feature fusion on the forward hidden feature set and the backward hidden feature set to obtain a plurality of original motor operation features;
(4) And performing recursive feature elimination on the plurality of original motor operation features to obtain a plurality of target motor operation features.
Specifically, forward hidden feature extraction is performed on the target motor operation data through a forward long-short-time memory unit in the bidirectional long-short-time network. In this process, the forward LSTM cells read time-series data and process the information step by step in the past to future direction. This enables the network to capture data features that evolve over time, which are referred to as forward hidden features. For example, if the motor gradually warms up over a period of time, this trend and pattern of temperature change will be captured by the forward LSTM unit and encoded as a forward hidden feature. The backward long-short-time memory unit in the bidirectional long-short-time network processes the operation data of the target motor, but the direction is opposite to that of the forward unit, namely the target motor never goes to the past. This approach enables the backward LSTM cells to capture features in the time series data that vary inversely with time, which are referred to as backward concealment features. For example, if the motor gradually cools down after a load reduction, this changing trend and pattern will be captured by the backward LSTM unit and encoded as a backward hidden feature. And carrying out feature fusion on the forward hidden feature set and the backward hidden feature set, and combining the forward and backward features to form a more comprehensive and deep feature representation. This fusion ensures the integrity of the motor operation data, i.e. both past to future time series information and future to past time series information. And carrying out recursive feature elimination on the fused original motor operation features. Recursive feature elimination is a feature selection technique that optimizes feature sets by progressively removing features that are considered least important. In each step, a model is built based on the current feature set and the importance of each feature is evaluated. The least important features are then removed and the process is repeated until a preset number of features is reached or some stop condition is met. By this method, a set of compact and efficient target motor operating characteristics are ultimately obtained, which are an optimized representation of the original motor operating data, capable of providing critical information for further analysis and control of the motor.
In a specific embodiment, the process of executing step 104 may specifically include the following steps:
(1) Respectively inputting a plurality of target motor operation characteristics into a random forest algorithm, wherein the random forest algorithm comprises a plurality of decision trees;
(2) Calculating the importance degree of each target motor operation feature through a plurality of decision trees to obtain the feature importance degree of each decision tree;
(3) Weighting and fusing the feature importance of each decision tree to obtain the target importance of each target motor operation feature;
(4) Comparing the target importance with a preset feature threshold to obtain a target comparison result, wherein the target comparison result comprises the feature threshold with the target importance more than or equal to the feature threshold and the feature threshold with the target importance less than the feature threshold;
(5) If the target importance is more than or equal to the feature threshold, mapping the corresponding target motor operation feature to a first motor operation feature set;
(6) And if the target importance is less than the characteristic threshold, mapping the corresponding target motor operation characteristic to a second motor operation characteristic set.
Specifically, a plurality of target motor operating characteristics are input into a random forest algorithm. These characteristics include temperature, vibration, current, voltage, rotational speed, etc. of the motor, which collectively reflect the operating state and health of the motor. These features are processed independently by each decision tree in a random forest algorithm, analyzed and classified. The decision tree is constructed based on a series of rules or conditions that are learned from the feature data. For example, a decision tree predicts whether the motor is overheated based on the temperature and current of the motor. And respectively carrying out importance calculation on the operation characteristics of each target motor through the decision trees. The usage of each feature in the decision tree is evaluated and their extent of contribution to the predicted outcome is evaluated. For example, if a feature is frequently used in multiple decision trees to successfully predict motor state, the importance of the feature may be high. The importance of a feature generated by each decision tree reflects the importance of the feature in a single model. And carrying out weighted fusion on the feature importance of each decision tree to obtain the comprehensive target importance of the operation features of each target motor. The weighted fusion process considers the performance and reliability of each decision tree, ensuring that the final target importance is comprehensive and accurate. For example, for a better performing decision tree, the feature importance will be weighted higher and vice versa. The weighted fusion method can provide a comprehensive feature importance evaluation, and ensures the identification of the most critical features in motor operation data. These target importance levels are compared to preset feature thresholds to determine which features are important enough to be retained in the final feature set. Based on the comparison, features are divided into two categories: the target importance is higher than or equal to the feature threshold and the target importance is lower than the feature threshold. For those features whose target importance is greater than or equal to the feature threshold, it is mapped to a first set of motor operating features. This set contains the features most critical to motor control and analysis, which will be used directly in subsequent motor control strategy formulation. And those features whose target importance level is below the feature threshold are mapped to a second set of motor operating features. This set of features, while having less impact on motor operation, may still provide auxiliary information for monitoring and analysis of motor status. The most critical motor operation characteristics can be effectively identified and utilized by the characteristic classification and optimization method based on the random forest algorithm. This not only improves the accuracy of feature selection, but also enhances the effectiveness of the motor control strategy.
In a specific embodiment, the process of executing step 105 may specifically include the following steps:
(1) Respectively carrying out feature coding and vector conversion on the first motor operation feature set and the second motor operation feature set to obtain a first motor operation feature vector and a second motor operation feature vector;
(2) Inputting a first motor operation feature vector and a second motor operation feature vector into a preset motor performance prediction model, wherein the motor performance prediction model comprises a first prediction task network and a second prediction task network, and the first prediction task network comprises: the first input layer, the first attention mechanism layer, the first convolution long short-term memory network and the first full-connection layer, and the second prediction task network comprises: the second input layer, the second attention mechanism layer, the second convolution long short-time memory network and the second full-connection layer;
(3) Carrying out vector standardization processing on the first motor operation feature vector through the first input layer to obtain a first standard operation feature vector; carrying out attention mechanism weighting on the first standard operation feature vector through a first attention mechanism layer to obtain a first attention feature vector; high-dimensional feature extraction is carried out on the first attention feature vector through a first convolution long-short time memory network, so that a first high-dimensional feature vector is obtained; carrying out transmission performance prediction on the first high-dimensional feature vector through the first full-connection layer to obtain a transmission performance predicted value;
(4) Carrying out vector standardization processing on the second motor operation feature vector through a second input layer to obtain a second standard operation feature vector; carrying out attention mechanism weighting on the second standard operation feature vector through a second attention mechanism layer to obtain a second attention feature vector; performing high-dimensional feature extraction on the second attention feature vector through a second convolution long-short time memory network to obtain a second high-dimensional feature vector; and predicting the performance of the Halbach array by the second full-connection layer on the second high-dimensional feature vector to obtain a predicted value of the performance of the Halbach array.
Specifically, feature encoding and vector conversion are performed on the first and second motor operation feature sets to convert the features into a format that can be processed by the deep learning model. The actual operating data (e.g., temperature readings, vibration frequency, amperage, etc.) is converted into a mathematical representation. For example, temperature readings are converted into a continuous sequence of values, while the vibration modes of the machine are encoded as a series of discrete labels. This conversion allows the original operational data to be utilized by the next model. And inputting the first motor operation characteristic vector and the second motor operation characteristic vector into a preset motor performance prediction model. This model consists of two independent networks of predictive tasks: a first predictive task network and a second predictive task network. Each network includes an input layer, an attention mechanism layer, a convolutional long and short time memory network (ConvLSTM), and a fully connected layer. The design of the network structure enables the model to process complex input data and capture time sequence features and deep features in the data. For the first motor operation feature vector, vector normalization processing is performed through the first input layer to ensure that the data have the same scale in different dimensions. The first attention mechanism layer weights these normalized feature vectors. Attention mechanisms were introduced to enable models to focus on those features that are most critical, thereby improving the accuracy of predictions. The first convolution long-short time memory network carries out high-dimensional feature extraction on the attention weighted feature vectors and captures long-term dependency and local features in time series data. And predicting the transmission performance of the high-dimensional feature vectors through the first full-connection layer, so as to obtain a predicted value of the transmission performance. And similarly, for the second motor operation feature vector, carrying out standardization processing through a second input layer, then carrying out weighting through a second attention mechanism layer, then carrying out high-dimensional feature extraction through a second convolution long-short-time memory network, and finally carrying out Halbach array performance prediction through a second full-connection layer to obtain a performance prediction value of the Halbach array. An advantage of this approach is that it can process multiple types of input features simultaneously and capture complex relationships between them.
In a specific embodiment, the process of executing step 106 may specifically include the following steps:
(1) Calculating a transmission ratio influence factor of the target brushless motor according to the transmission performance predicted value, and calculating a Halbach array adjustment factor of the target brushless motor according to the Halbach array performance predicted value;
(2) Defining a torque optimization objective function of the target brushless motor according to the transmission ratio influence factor and the Halbach array adjustment factor;
(3) Carrying out parameter set optimization on an initial motor parameter set through a preset Bayesian optimization algorithm to obtain a transmission ratio adjustment coefficient and a Halbach array adjustment coefficient;
(4) Generating a corresponding target motor parameter set according to the transmission ratio adjustment coefficient and the Halbach array adjustment coefficient;
(5) And calculating a torque optimization target value of the target motor parameter set through the torque optimization target function, and creating an adaptive rotating speed control strategy of the target brushless motor according to the torque optimization target value and the target motor parameter set.
Specifically, a transmission ratio influence factor of the target brushless motor is calculated according to the transmission performance predicted value. The transmission ratio influencing factor refers to the degree of influence of the transmission ratio on the motor performance, such as efficiency, output torque, etc. For example, for a particular brushless motor, if the prediction shows a decrease in motor efficiency at a higher gear ratio, the gear ratio influencing factor will be directed to decreasing the gear ratio to optimize performance. And calculating the Halbach array adjustment factor of the target brushless motor according to the Halbach array performance predicted value. The Halbach array adjustment factor relates to the effect of the magnetic field layout of the Halbach array in the motor on the motor performance, such as magnetic field strength, uniformity, etc. And defining a torque optimization objective function of the target brushless motor according to the transmission ratio influence factor and the Halbach array adjustment factor. This objective function is the core of the optimization process, which combines the effects of the gear ratio and Halbach array to determine the optimal motor parameter settings. The torque optimization objective function may be designed based on factors such as desired output, efficiency, and operational stability of the motor to ensure optimal performance of the motor under various operating conditions. And optimizing the initial motor parameter set through a preset Bayesian optimization algorithm. Bayesian optimization is an efficient global optimization algorithm that searches for optimal parameters by constructing a probabilistic model of the parameters and under the direction of the model. In this process, the algorithm will continuously adjust the search strategy based on the performance of the objective function to find the optimal solution faster. Through bayesian optimization, gear ratio adjustment coefficients and Halbach array adjustment coefficients can be obtained, which represent the amount of parameter adjustment required to achieve optimal motor performance. And generating a corresponding target motor parameter set according to the obtained transmission ratio adjustment coefficient and the Halbach array adjustment coefficient. This set contains adjusted motor parameters such as changed gear ratio, reconfigured Halbach array, etc. to ensure optimal performance of the motor at the new parameter settings. And calculating a torque optimization target value of the target motor parameter set through a torque optimization target function. The optimized parameter set is used to evaluate the expected performance of the motor. The torque optimization target value is calculated based on the expected output and performance targets of the motor, reflecting the best performance that the motor can achieve at a given parameter setting. For example, this involves calculating the maximum torque output and maximum efficiency that can be achieved by the motor for a particular gear ratio and Halbach array configuration. And creating an adaptive rotating speed control strategy of the target brushless motor according to the torque optimization target value and the target motor parameter set. This control strategy is designed based on optimal parameter settings of the motor in order to automatically adjust the rotational speed and other control parameters of the motor under different operating conditions to achieve optimal performance. For example, control strategies include automatically reducing rotational speed as motor load increases to maintain high efficiency, or adjusting Halbach array configuration to optimize cooling effects when motor temperature increases are detected.
The above describes a method for controlling the rotational speed of a brushless motor in the embodiment of the present application, and the following describes a device for controlling the rotational speed of a brushless motor in the embodiment of the present application, referring to fig. 2, one embodiment of the device for controlling the rotational speed of a brushless motor in the embodiment of the present application includes:
an initialization module 201, configured to initialize motor parameters of a target brushless motor through a preset genetic algorithm, so as to obtain an initial motor parameter set;
the testing module 202 is configured to perform operation testing and data acquisition on the target brushless motor based on the initial motor parameter set to obtain initial motor operation data, and perform data dimension reduction on the initial motor operation data based on two-dimensional inter-class boundary Fisher analysis to obtain target motor operation data;
the feature extraction module 203 is configured to perform feature extraction on the target motor operation data to obtain a plurality of original motor operation features, and perform recursive feature elimination on the plurality of original motor operation features to obtain a plurality of target motor operation features;
the feature classification module 204 is configured to perform feature classification on the plurality of target motor operation features through a random forest algorithm to obtain a first motor operation feature set and a second motor operation feature set;
The prediction module 205 is configured to input the first motor operation feature set and the second motor operation feature set into a preset motor performance prediction model to perform transmission performance prediction and Halbach array performance prediction, so as to obtain a transmission performance predicted value and a Halbach array performance predicted value;
the creating module 206 is configured to perform bayesian optimization on the initial motor parameter set according to the transmission performance predicted value and the Halbach array performance predicted value to obtain a target motor parameter set, and create an adaptive rotational speed control strategy of the target brushless motor according to the target motor parameter set.
Through the cooperation of the components, the accuracy of brushless motor rotation speed control can be improved through a method for initializing motor parameters and driving data by using a genetic algorithm. The sensor data acquisition and feature extraction further increase accurate monitoring of the motor operating state. The performance of the brushless motor is allowed to be optimized in actual operation through the transmission performance prediction and the Halbach array performance prediction. This means that the motor can achieve higher efficiency and performance under different operating conditions. By using a Bayesian optimization algorithm to adjust motor parameters, an adaptive rotational speed control strategy is realized. This means that the motor can adjust its parameters in real time under different conditions to meet performance requirements without manual intervention. The data dimension can be reduced, the calculation efficiency can be improved, and the most important motor operation characteristics can be selected, so that the model performance can be improved. By optimizing the motor parameters and the control strategy, the energy waste is reduced, the energy utilization rate is improved, and the running cost of the motor is reduced. Based on the intelligent control of data and algorithm, the brushless motor can adapt to different working environments and load requirements, so that the intelligent level of the system is improved, the design parameters of the brushless motor are optimized, and the accuracy of the rotation speed control of the brushless motor is improved.
The present application also provides a computer device including a memory and a processor, where the memory stores computer readable instructions that, when executed by the processor, cause the processor to perform the steps of the method for controlling a rotational speed of a brushless motor in the above embodiments.
The present application also provides a computer readable storage medium, which may be a non-volatile computer readable storage medium, or may be a volatile computer readable storage medium, where instructions are stored in the computer readable storage medium, where the instructions when executed on a computer cause the computer to perform the steps of the method for controlling the rotational speed of the brushless motor.
It will be clearly understood by those skilled in the art that, for convenience and brevity of description, the specific working processes of the above-described systems, systems and units may refer to the corresponding processes in the foregoing method embodiments, which are not repeated herein.
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 in essence or a part contributing to the prior art or all or part of the technical solution 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.) to perform all or part of the steps of the methods described in 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 (random access memory, RAM), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
The above embodiments are merely for illustrating the technical solution of the present application, and not for limiting the same; although the present application has been described in detail with reference to the foregoing embodiments, it should be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the corresponding technical solutions.
Claims (8)
1. A rotational speed control method of a brushless motor, characterized by comprising:
initializing motor parameters of a target brushless motor through a preset genetic algorithm to obtain an initial motor parameter set;
performing operation test and data acquisition on the target brushless motor based on the initial motor parameter set to obtain initial motor operation data, and performing data dimension reduction on the initial motor operation data based on two-dimensional inter-class boundary Fisher analysis to obtain target motor operation data;
extracting the characteristics of the target motor operation data to obtain a plurality of original motor operation characteristics, and performing recursive characteristic elimination on the plurality of original motor operation characteristics to obtain a plurality of target motor operation characteristics;
Performing feature classification on the plurality of target motor operation features through a random forest algorithm to obtain a first motor operation feature set and a second motor operation feature set;
inputting the first motor operation characteristic set and the second motor operation characteristic set into a preset motor performance prediction model to predict transmission performance and Halbach array performance, so as to obtain a transmission performance predicted value and a Halbach array performance predicted value; the method specifically comprises the following steps: respectively carrying out feature coding and vector conversion on the first motor operation feature set and the second motor operation feature set to obtain a first motor operation feature vector and a second motor operation feature vector; inputting the first motor operation feature vector and the second motor operation feature vector into a preset motor performance prediction model, wherein the motor performance prediction model comprises a first prediction task network and a second prediction task network, and the first prediction task network comprises: the first input layer, the first attention mechanism layer, the first convolution long short-term memory network and the first full-connection layer, wherein the second prediction task network comprises: the second input layer, the second attention mechanism layer, the second convolution long short-time memory network and the second full-connection layer; carrying out vector standardization processing on the first motor operation feature vector through the first input layer to obtain a first standard operation feature vector; performing attention mechanism weighting on the first standard operation feature vector through the first attention mechanism layer to obtain a first attention feature vector; performing high-dimensional feature extraction on the first attention feature vector through the first convolution long-short-term memory network to obtain a first high-dimensional feature vector; carrying out transmission performance prediction on the first high-dimensional feature vector through the first full-connection layer to obtain a transmission performance prediction value; carrying out vector standardization processing on the second motor operation feature vector through the second input layer to obtain a second standard operation feature vector; performing attention mechanism weighting on the second standard operation feature vector through the second attention mechanism layer to obtain a second attention feature vector; performing high-dimensional feature extraction on the second attention feature vector through the second convolution long-short-time memory network to obtain a second high-dimensional feature vector; performing Halbach array performance prediction on the second high-dimensional feature vector through the second full-connection layer to obtain a Halbach array performance prediction value;
Performing Bayesian optimization on the initial motor parameter set according to the transmission performance predicted value and the Halbach array performance predicted value to obtain a target motor parameter set, and creating a self-adaptive rotating speed control strategy of the target brushless motor according to the target motor parameter set; the method specifically comprises the following steps: calculating a transmission ratio influence factor of the target brushless motor according to the transmission performance predicted value, and calculating a Halbach array adjustment factor of the target brushless motor according to the Halbach array performance predicted value; defining a torque optimization objective function of the target brushless motor according to the transmission ratio influence factor and the Halbach array adjustment factor; carrying out parameter set optimization on the initial motor parameter set through a preset Bayesian optimization algorithm to obtain a transmission ratio adjustment coefficient and a Halbach array adjustment coefficient; generating a corresponding target motor parameter set according to the transmission ratio adjustment coefficient and the Halbach array adjustment coefficient; and calculating a torque optimization target value of the target motor parameter set through the torque optimization target function, and creating an adaptive rotating speed control strategy of the target brushless motor according to the torque optimization target value and the target motor parameter set.
2. The method for controlling the rotational speed of a brushless motor according to claim 1, wherein initializing motor parameters of the target brushless motor by a preset genetic algorithm to obtain an initial motor parameter set comprises:
initializing motor parameters of a target brushless motor to obtain pole slot matching parameters, permanent magnet thickness parameters and notch width parameters;
calculating a pole slot matching factor of the pole slot matching parameter, calculating a permanent magnet thickness effect factor of the permanent magnet thickness parameter, and calculating a slot width effect factor of the slot width parameter;
initializing a parameter set according to the pole slot matching factor, the permanent magnet thickness effect factor and the notch width effect factor and through a preset genetic algorithm, so as to obtain a plurality of first motor parameter sets;
performing fitness analysis on the plurality of first motor parameter sets to obtain a fitness value of each first motor parameter set;
performing fitness sequencing on the plurality of first motor parameter sets according to the fitness value to obtain a motor parameter set sequence, and performing group division on the motor parameter set sequence to obtain a candidate motor parameter group of the first 50% in the motor parameter set sequence;
Performing crossover operation and mutation operation on the candidate motor parameter groups to obtain a plurality of second motor parameter sets;
and carrying out parameter selection and parameter updating on the plurality of second motor parameter sets based on a roulette selection method to obtain an initial motor parameter set.
3. The method for controlling a rotational speed of a brushless motor according to claim 1, wherein the performing operation test and data acquisition on the target brushless motor based on the initial motor parameter set to obtain initial motor operation data, and performing data dimension reduction on the initial motor operation data based on two-dimensional inter-class boundary Fisher analysis to obtain target motor operation data includes:
performing operation test on the target brushless motor based on the initial motor parameter set, and acquiring operation data of the target brushless motor through a preset sensor group to obtain initial motor operation data;
performing data cleaning on the initial motor operation data to obtain first motor operation data, and performing standardized processing on the first motor operation data to obtain second motor operation data;
performing time sequence synchronization processing on the second motor operation data to obtain standard motor operation data, and performing data classification on the standard motor operation data to obtain motor temperature data and motor vibration data;
Respectively calculating a temperature mean value and a temperature variance of the motor temperature data and a vibration mean value and a vibration variance of the motor vibration data;
based on Fisher analysis of boundaries between two-dimensional classes, computing Fisher discriminants of the temperature mean and the temperature variance, and the vibration mean and the vibration variance;
and generating a corresponding dimension reduction conversion matrix according to the Fisher criterion, and carrying out data dimension reduction on the standard motor operation data according to the dimension reduction conversion matrix to obtain target motor operation data.
4. The method of claim 1, wherein the performing feature extraction on the target motor operation data to obtain a plurality of original motor operation features, and performing recursive feature elimination on the plurality of original motor operation features to obtain a plurality of target motor operation features, includes:
forward hidden feature extraction is carried out on the operation data of the target motor through a forward long-short-time memory unit in a preset bidirectional long-short-time memory network, so that a forward hidden feature set is obtained;
backward hidden characteristic extraction is carried out on the operation data of the target motor through a backward long-short-time memory unit in the bidirectional long-short-time memory network, so that a backward hidden characteristic set is obtained;
Performing feature fusion on the forward hidden feature set and the backward hidden feature set to obtain a plurality of original motor operation features;
and performing recursive feature elimination on the plurality of original motor operation features to obtain a plurality of target motor operation features.
5. The method for controlling a rotational speed of a brushless motor according to claim 1, wherein the classifying the plurality of target motor operation features by a random forest algorithm to obtain a first motor operation feature set and a second motor operation feature set includes:
respectively inputting the operation characteristics of the target motors into a random forest algorithm, wherein the random forest algorithm comprises a plurality of decision trees;
calculating the importance degree of the operation characteristics of each target motor through the decision trees to obtain the characteristic importance degree of each decision tree;
weighting and fusing the feature importance of each decision tree to obtain the target importance of each target motor operation feature;
comparing the target importance with a preset feature threshold to obtain a target comparison result, wherein the target comparison result comprises a feature threshold with the target importance more than or equal to the feature threshold and a feature threshold with the target importance less than the feature threshold;
If the target importance is more than or equal to the feature threshold, mapping the corresponding target motor operation feature to a first motor operation feature set;
and if the target importance is less than the characteristic threshold, mapping the corresponding target motor operation characteristic to a second motor operation characteristic set.
6. A rotational speed control apparatus of a brushless motor, the rotational speed control apparatus comprising:
the initialization module is used for initializing motor parameters of the target brushless motor through a preset genetic algorithm to obtain an initial motor parameter set;
the testing module is used for carrying out operation testing and data acquisition on the target brushless motor based on the initial motor parameter set to obtain initial motor operation data, and carrying out data dimension reduction on the initial motor operation data based on two-dimensional inter-class boundary Fisher analysis to obtain target motor operation data;
the feature extraction module is used for extracting features of the target motor operation data to obtain a plurality of original motor operation features, and performing recursive feature elimination on the plurality of original motor operation features to obtain a plurality of target motor operation features;
the feature classification module is used for classifying the features of the plurality of target motor operation features through a random forest algorithm to obtain a first motor operation feature set and a second motor operation feature set;
The prediction module is used for inputting the first motor operation characteristic set and the second motor operation characteristic set into a preset motor performance prediction model to perform transmission performance prediction and Halbach array performance prediction, so as to obtain a transmission performance prediction value and a Halbach array performance prediction value; the method specifically comprises the following steps: respectively carrying out feature coding and vector conversion on the first motor operation feature set and the second motor operation feature set to obtain a first motor operation feature vector and a second motor operation feature vector; inputting the first motor operation feature vector and the second motor operation feature vector into a preset motor performance prediction model, wherein the motor performance prediction model comprises a first prediction task network and a second prediction task network, and the first prediction task network comprises: the first input layer, the first attention mechanism layer, the first convolution long short-term memory network and the first full-connection layer, wherein the second prediction task network comprises: the second input layer, the second attention mechanism layer, the second convolution long short-time memory network and the second full-connection layer; carrying out vector standardization processing on the first motor operation feature vector through the first input layer to obtain a first standard operation feature vector; performing attention mechanism weighting on the first standard operation feature vector through the first attention mechanism layer to obtain a first attention feature vector; performing high-dimensional feature extraction on the first attention feature vector through the first convolution long-short-term memory network to obtain a first high-dimensional feature vector; carrying out transmission performance prediction on the first high-dimensional feature vector through the first full-connection layer to obtain a transmission performance prediction value; carrying out vector standardization processing on the second motor operation feature vector through the second input layer to obtain a second standard operation feature vector; performing attention mechanism weighting on the second standard operation feature vector through the second attention mechanism layer to obtain a second attention feature vector; performing high-dimensional feature extraction on the second attention feature vector through the second convolution long-short-time memory network to obtain a second high-dimensional feature vector; performing Halbach array performance prediction on the second high-dimensional feature vector through the second full-connection layer to obtain a Halbach array performance prediction value;
The creating module is used for carrying out Bayesian optimization on the initial motor parameter set according to the transmission performance predicted value and the Halbach array performance predicted value to obtain a target motor parameter set, and creating a self-adaptive rotating speed control strategy of the target brushless motor according to the target motor parameter set; the method specifically comprises the following steps: calculating a transmission ratio influence factor of the target brushless motor according to the transmission performance predicted value, and calculating a Halbach array adjustment factor of the target brushless motor according to the Halbach array performance predicted value; defining a torque optimization objective function of the target brushless motor according to the transmission ratio influence factor and the Halbach array adjustment factor; carrying out parameter set optimization on the initial motor parameter set through a preset Bayesian optimization algorithm to obtain a transmission ratio adjustment coefficient and a Halbach array adjustment coefficient; generating a corresponding target motor parameter set according to the transmission ratio adjustment coefficient and the Halbach array adjustment coefficient; and calculating a torque optimization target value of the target motor parameter set through the torque optimization target function, and creating an adaptive rotating speed control strategy of the target brushless motor according to the torque optimization target value and the target motor parameter set.
7. A computer device, the computer device comprising: a memory and at least one processor, the memory having instructions stored therein;
the at least one processor invokes the instructions in the memory to cause the computer device to perform the method of controlling the rotational speed of a brushless motor as claimed in any one of claims 1 to 5.
8. A computer-readable storage medium having instructions stored thereon, which when executed by a processor, implement a method of controlling the rotational speed of a brushless motor according to any one of claims 1-5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410095741.6A CN117614323B (en) | 2024-01-24 | 2024-01-24 | Brushless motor rotation speed control method, device, equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410095741.6A CN117614323B (en) | 2024-01-24 | 2024-01-24 | Brushless motor rotation speed control method, device, equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117614323A CN117614323A (en) | 2024-02-27 |
| CN117614323B true CN117614323B (en) | 2024-04-02 |
Family
ID=89948361
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410095741.6A Active CN117614323B (en) | 2024-01-24 | 2024-01-24 | Brushless motor rotation speed control method, device, equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117614323B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117097227A (en) * | 2023-10-17 | 2023-11-21 | 深圳市华科科技有限公司 | Speed regulation control method and related device for motor |
| CN117390536A (en) * | 2023-12-11 | 2024-01-12 | 深圳市宝腾互联科技有限公司 | Operation and maintenance management method and system based on artificial intelligence |
-
2024
- 2024-01-24 CN CN202410095741.6A patent/CN117614323B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117097227A (en) * | 2023-10-17 | 2023-11-21 | 深圳市华科科技有限公司 | Speed regulation control method and related device for motor |
| CN117390536A (en) * | 2023-12-11 | 2024-01-12 | 深圳市宝腾互联科技有限公司 | Operation and maintenance management method and system based on artificial intelligence |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117614323A (en) | 2024-02-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CA3073676A1 (en) | Systems and methods for operating a data center based on a generated machine learning pipeline | |
| CN111237988B (en) | Control method and system for subway vehicle-mounted air conditioning unit | |
| CN112884008B (en) | Prediction evaluation method and device for running state of power information acquisition system | |
| CN110874665B (en) | Control device and method for wind generating set | |
| CN114744309B (en) | BMS-based battery safety management method, device, equipment and storage medium | |
| US8676356B2 (en) | System and method for motor parameter estimation | |
| CN117318055B (en) | Power load prediction model processing method and device, electronic equipment and storage medium | |
| CN113762591B (en) | Short-term electric quantity prediction method and system based on GRU and multi-core SVM countermeasure learning | |
| CN117614323B (en) | Brushless motor rotation speed control method, device, equipment and storage medium | |
| CN116541778A (en) | Equipment data processing and caching method based on Internet of things | |
| CN117544021B (en) | Rotating speed control method and system of brushless direct current motor | |
| CN117094478B (en) | Energy scheduling management method, device, equipment and storage medium | |
| Symeonidis et al. | A methodology for predicting agent behavior by the use of data mining techniques | |
| Tyas et al. | Implementation of Particle Swarm Optimization (PSO) to improve neural network performance in univariate time series prediction | |
| CN117673580B (en) | Temperature control equipment for new energy battery | |
| CN117498522B (en) | Power supply control method and system for photovoltaic energy storage charging pile | |
| CN116774058B (en) | Battery life prediction method, device, equipment and storage medium | |
| CN117439146B (en) | Data analysis control method and system for charging pile | |
| Zhai et al. | Multi-swarm genetic gray wolf optimizer with embedded autoencoders for high-dimensional expensive problems | |
| CN117910550B (en) | Oil-free ultrahigh-speed centrifugal compressor automatic optimization system based on deep learning | |
| CN117195001A (en) | HPC job power consumption prediction method and system based on power consumption curve and script information | |
| CN117577981A (en) | Photovoltaic power generation energy storage control method and system | |
| CN118012182A (en) | Control method, system, equipment and storage medium for operating environment of high-voltage switch cabinet | |
| CN117217351A (en) | High-precision learning electric power auxiliary decision-making method and system | |
| CN118134210A (en) | Carbon footprint management method and system for steel production |
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 |