CN114662413B - Intelligent inversion optimization method for transmission chain system - Google Patents
Intelligent inversion optimization method for transmission chain system Download PDFInfo
- Publication number
- CN114662413B CN114662413B CN202210566017.8A CN202210566017A CN114662413B CN 114662413 B CN114662413 B CN 114662413B CN 202210566017 A CN202210566017 A CN 202210566017A CN 114662413 B CN114662413 B CN 114662413B
- Authority
- CN
- China
- Prior art keywords
- transmission chain
- variable
- model
- matrix
- chain system
- 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
- 238000005457 optimization Methods 0.000 title claims abstract description 61
- 238000000034 method Methods 0.000 title claims abstract description 54
- 230000005540 biological transmission Effects 0.000 title claims abstract description 48
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 37
- 238000005070 sampling Methods 0.000 claims abstract description 34
- 238000013461 design Methods 0.000 claims abstract description 33
- 230000004044 response Effects 0.000 claims abstract description 30
- 238000012360 testing method Methods 0.000 claims abstract description 20
- 238000004364 calculation method Methods 0.000 claims abstract description 16
- 239000011159 matrix material Substances 0.000 claims description 38
- 230000000694 effects Effects 0.000 claims description 25
- 230000008569 process Effects 0.000 claims description 24
- 150000001875 compounds Chemical class 0.000 claims description 19
- 230000002068 genetic effect Effects 0.000 claims description 15
- 238000007476 Maximum Likelihood Methods 0.000 claims description 14
- 238000004458 analytical method Methods 0.000 claims description 13
- 238000009826 distribution Methods 0.000 claims description 10
- 238000000540 analysis of variance Methods 0.000 claims description 6
- 230000008859 change Effects 0.000 claims description 6
- 238000002474 experimental method Methods 0.000 claims description 6
- 238000012216 screening Methods 0.000 claims description 6
- 230000035945 sensitivity Effects 0.000 claims description 6
- 238000012163 sequencing technique Methods 0.000 claims description 6
- 238000010586 diagram Methods 0.000 claims description 4
- 238000004088 simulation Methods 0.000 claims description 4
- 239000002245 particle Substances 0.000 claims description 3
- 230000009290 primary effect Effects 0.000 claims description 2
- 239000003795 chemical substances by application Substances 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000001808 coupling effect Effects 0.000 description 1
- 238000006880 cross-coupling reaction Methods 0.000 description 1
- 238000012938 design process Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000035772 mutation Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000008646 thermal stress Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2113/00—Details relating to the application field
- G06F2113/06—Wind turbines or wind farms
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Theoretical Computer Science (AREA)
- Biophysics (AREA)
- Evolutionary Computation (AREA)
- Business, Economics & Management (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- General Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Economics (AREA)
- Artificial Intelligence (AREA)
- Computer Hardware Design (AREA)
- Human Resources & Organizations (AREA)
- Medical Informatics (AREA)
- General Business, Economics & Management (AREA)
- Water Supply & Treatment (AREA)
- Geometry (AREA)
- Public Health (AREA)
- Tourism & Hospitality (AREA)
- Strategic Management (AREA)
- Primary Health Care (AREA)
- Marketing (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physiology (AREA)
- Genetics & Genomics (AREA)
- Biomedical Technology (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
An intelligent inversion optimization method for a transmission chain system comprises the steps of carrying out modeling calculation on electromagnetic response of the transmission chain system under an SMC-SVPWM sliding mode control algorithm, utilizing the model to calculate test sample response generated under an improved Latin hypercube sampling method, and establishing an intelligent inversion Gaussian random process optimization model based on target search, wherein the model carries out inversion iteration by taking transmission chain efficiency as a target so as to improve the precision of the model; in order to ensure the high reliability of the transmission chain system, namely reduce the fluctuation of the output torque of the transmission chain, the efficiency and the torque fluctuation are optimized by adopting a multi-objective optimization algorithm based on range selection, and the high-reliability collaborative optimization design facing the transmission chain system of the large-scale offshore wind power equipment is realized.
Description
Technical Field
The invention relates to the technical field of transmission chains, in particular to an intelligent inversion optimization method for a transmission chain system.
Background
In the application of large offshore wind turbines in deep and open sea, the design of high reliability and high efficiency of a transmission chain system of a direct-drive permanent magnet wind turbine still has larger theoretical blank and manufacturing difficulty. The reliability of a transmission chain directly influences the running state and the maintenance cost of the whole unit, and the efficiency directly influences the utilization efficiency of ocean resources and energy resources in China; therefore, the optimal design of the transmission chain system has important significance for the development and utilization of offshore resources in China.
The transmission chain system mainly comprises two parts, namely a converter and a generator, modeling analysis is mainly considered for each part of the transmission chain system in the optimization design process, the multi-physical-field cross coupling effect among the parts is considered, and optimization design is developed aiming at the efficiency maximization of the system while the reliability and the stability are not influenced.
In order to optimize the maximum efficiency of the drive chain system, people develop an optimized design for the generator part, and although the optimized efficiency of the drive chain can be improved to a certain extent, the following disadvantages still exist:
(1) the actual power supply of the generator is not considered as a PWM waveform, and only the optimal design of the generator under an ideal sinusoidal power supply waveform is considered, so that the optimal design and application of an actual transmission chain system are difficult to fit with by the optimization method.
(2) By adopting the traditional fitting agent model, such as polynomial fitting, neural network prediction and other methods, the mathematical logic relationship of high interaction and high-order nonlinearity of efficiency and optimization variables is difficult to accurately establish, and the accuracy of the established model is low.
(3) In the prior art, when optimization design is performed, optimization design is often performed on an individual generator, that is, a power supply mode of the generator is considered to be three-phase sine waveform power supply, and a power supply mode of a motor in an actual drive chain system is PWM (pulse width modulation) waveform. The motor runs under PWM control waveform in the actual operation process promptly, and its electromagnetic property all has great change under comparing the ideal power supply waveform with mechanical properties, and specific change has: the increase of electromagnetic loss, the increase of loss of power elements such as IGBT, the reduction of service life, the increase of thermal stress, the increase of temperature rise and the like. Therefore, the design optimization scheme of the generator is designed optimally without considering the actual operation condition of the generator, the error from the actual design scheme is often large, and the design index is difficult to achieve.
(4) In the prior art, when an agent model is established, a traditional fitting model such as a polynomial fitting model, a neural network and other models is often adopted, and even under the condition that a PWM (pulse-width modulation) control waveform at the converter side is not considered, a plurality of physical fields are often influenced between optimization targets such as efficiency and torque and optimally designed motor structure variables, and the mathematical relationship of the transmission chain system has the characteristics of high interactivity, high-order nonlinearity and the like; therefore, the traditional agent model is difficult to learn the responsible mathematical relationship through the data sample. After the influence of the converter side on the whole system is considered, the mathematical relation is more complex, and the accuracy of the model is often difficult to guarantee.
Disclosure of Invention
The invention aims to solve the technical problem of providing an intelligent inversion optimization method of a transmission chain system,
in order to solve the technical problem, the invention adopts the following technical scheme:
an intelligent inversion optimization design method for a transmission chain system comprises the following steps;
s1, performing modeling calculation on electromagnetic response of a transmission chain system under an SMC-SVPWM sliding mode control algorithm to obtain the overall efficiency of the transmission chain system under the SMC-SVPWM sliding mode control algorithm;
s2, selecting the length variable of the permanent magnet on the basis of S1
Width, width
Tooth space parameters
To optimize the variables; establishing a Morris random sampling matrix of motor parameters, calculating the electromagnetic response of a sample by an S1 modeling calculation method, carrying out primary effect analysis on the motor parameters to quantitatively measure the sensitivity of each parameter and the interactivity between the parameters, establishing an orthogonal test table of variables by a Taguchi method, carrying out signal-to-noise ratio analysis on the orthogonal test table, and preliminarily screening out the sensitive variables of the motor;
s3, selecting the length variable of the permanent magnet through S2
Width, width
Tooth space parameters
For sensitive variables, BS1=4.5mm, HS1 was also determined=1.7mm；
Quantitatively analyzing the influence proportion of each variable on the efficiency of the transmission chain, carrying out variance analysis on the influence proportion, and calculating the variance value of each variable;
finally, the elementary effect and the analysis of variance are combined for use, the difference between the horizontal means of one or more factors is checked, and the variable with larger average value of the elementary effect and larger amplitude change in the analysis of variance is selected as a sensitive variable;
s4, generating a sampling scheme by adopting a Latin hypercube sampling scheme with higher uniformity according to the sensitive variable in S3; the sampling scheme comprises the following steps:
firstly, sampling by randomly generating different hyper-Latin cubes with 6 dimensions and 120 sample points;
establishing a distance norm and score function of two spatial points for the established Latin hypercube sampling scheme:
in the formula (I), the compound is shown in the specification,
is the spatial norm between the two samples,
is a norm of a sample
The number of (2);
the parameters are calculated for the distance and,
is to be treatedThe parameters are optimized, and the parameters are optimized,
is as follows
A first sample of
The number of the variables is one,
is as follows
A first sample of
The number of the variables is one,
is the total number of the samples and is,
is composed of
The number of the variables is one,
is composed of
The number of the variables is one,
is the variable level number;
optimizing the local optimal test design schemes under different q through a genetic algorithm, and sequencing all local sequencing schemes through a maximum minimum criterion after determining the local optimal design schemes under different q to select the scheme with the best spatial uniformity:
in the formula (I), the compound is shown in the specification,
is the spatial norm between the two samples,
is a norm of a sample
The number of (2);
s5, carrying out transmission chain integrated simulation on the sample points in the sampling scheme to obtain efficiency response; selecting 70% of sample points to establish a Gaussian random process model, and obtaining an intelligent inversion optimization model by adopting an inversion optimization design method taking a target as a guide on the basis of establishing the Gaussian random process model;
and S6, on the basis of the intelligent inversion optimization model, considering the integral output torque of the transmission chain system as constraint, and optimizing the constraint by adopting a PESA-II (particle swarm optimization-II) range selection-based multi-objective optimization algorithm.
Further, the step of determining the sensitive variable in S3 is as follows:
screening the sensitivity of the whole efficiency of the transmission chain to each parameter by utilizing an elementary effect distribution diagram, and constructing a Morris statistical matrix to carry out elementary effect analysis on variables, namely:
in the formula (I), the compound is shown in the specification,
is as follows
The distribution of the elementary effects of the individual variables,
in the form of increments of the number of bits,
is the variable of the k-th weft yarn,
in the x-direction, the direction of the X-axis,
for drive chain efficiency;
and (4) constructing a calculation matrix X by a random permutation algorithm, wherein each column of the normalized (k + 1) k sampling matrix has only two rows which are different at the ith position.
Further, the calculation matrix X is formulated as follows:
in the formula (I), the compound is shown in the specification,
is a unit vector of k +1 columns,
in order to randomly initialize the increment of the initialization,
is a row vector generated by random increment and having the same dimension as the variable and each element being a multiple of the random increment not exceeding 1,
is a matrix whose elements are only 0, 1，
For a zero vector with only two elements 1,
is a matrix of the units,
for randomly initialized elements with diagonals of only 1, -1,
performing the sampling process r times, and analyzing factors of each behavior elementary effect of the X matrix;
and (4) performing r times of elementary effect calculation on the X matrix to obtain elementary effect distribution of different variables.
Further, the variance formula in S3 is:
in the formula (I), the compound is shown in the specification,
in order to obtain the number of tests,
the number of the variable levels is the number of the variable levels,
for each number of horizontal experiments the number of experiments,
in order to achieve the target value,
represents the first
A variable of
One level corresponds to all
Is determined by the average value of (a) of (b),
representing all times of test
Average value of (a).
Further, on the basis of establishing a Gaussian random process model, a specific process of obtaining an intelligent inversion optimization model by adopting an inversion optimization design method taking a target as a guide is as follows;
establishing a Gaussian random process model for 70% of sample points by combining maximum likelihood estimation with a genetic algorithm, namely, a data-based maximum likelihood estimation function:
in the formula (I), the compound is shown in the specification,
the number of the samples is the number of the samples,
in response to this, the mobile station is allowed to respond,
is a covariance matrix, X is a matrix of sampled samples, Y is a response vector, wherein,
and
the relationship of the functions can be obtained by combining genetic algorithm with maximum likelihood estimation.
Further, after the model is established, by setting a specific optimization target value to 97%, according to the established target response, determining a new model parameter and a point of addition of the optimal improved model probability by combining a genetic algorithm and maximum likelihood estimation;
wherein, the new likelihood function includes target optimization efficiency in addition to the model parameters, that is:
in the formula (I), the compound is shown in the specification,
the number of the samples is the number of the samples,
in response to this, the mobile station is allowed to respond,
is a covariance matrix, X is a matrix of sampled samples, Y is a response vector,
is the set target value.
Compared with the prior art, the invention has the following beneficial effects:
1. the invention provides an intelligent inversion optimization method for a transmission chain system, which is characterized in that the electromagnetic response of the transmission chain system under the SMC-SVPWM sliding mode control algorithm is modeled and calculated, the model is used for calculating the test sample response generated under the improved Latin hypercube sampling method, an intelligent inversion Gaussian random process optimization model based on target search is established, and the model performs inversion iteration by taking the transmission chain efficiency as the target so as to improve the accuracy of the model; the efficiency is improved, meanwhile, in order to ensure the high reliability of the transmission chain system, namely, reduce the fluctuation of the output torque of the transmission chain, a multi-objective optimization algorithm based on range selection is adopted to optimize the efficiency and the torque fluctuation, and the high-reliability collaborative optimization design facing the transmission chain system of the large-scale offshore wind power equipment is realized.
2. Aiming at the problems, the optimization of the whole system is considered in the optimization design of the transmission chain, namely the influence of a control algorithm of a generator and a control external circuit on the whole system is considered, the sample response points obtained by the whole system are considered, the optimization design can be carried out on the motor in the actual running state, and the result is more reliable. Secondly, the invention adopts a Gaussian process modeling method based on target search intelligent inversion optimization, combines specific numerical values of an optimized target, and calculates parameter numerical values of a model and calculates an optimal point adding position through maximum likelihood estimation and a genetic algorithm so as to improve the accuracy of the model.
Drawings
FIG. 1 is a flow chart of the drive chain system intelligent inversion optimization of the present invention.
FIG. 2 is a diagram of optimized variation of tooth space parameters of the drive chain of the present invention.
FIG. 3 is a flow chart of an intelligent inversion Gaussian random process model according to the present invention.
Detailed Description
In order to make the technical problems, technical solutions and advantages of the present invention more apparent, the following detailed description is given with reference to the accompanying drawings and specific embodiments. In the following description, characteristic details such as specific configurations and components are provided only to help the embodiments of the present invention be fully understood. Thus, it will be apparent to those skilled in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the invention. In addition, descriptions of well-known functions and constructions are omitted for clarity and conciseness.
It should be appreciated that reference throughout this specification to "one embodiment" or "an embodiment" means that a particular feature, structure or characteristic described in connection with the embodiment is included in at least one embodiment of the present invention. Thus, the appearances of the phrases "in one embodiment" or "in an embodiment" in various places throughout this specification are not necessarily all referring to the same embodiment. Furthermore, the particular features, structures, or characteristics may be combined in any suitable manner in one or more embodiments.
In various embodiments of the present invention, it should be understood that the sequence numbers of the following processes do not mean the execution sequence, and the execution sequence of each process should be determined by its function and inherent logic, and should not constitute any limitation to the implementation process of the embodiments of the present invention.
It should be understood that the term "and/or" herein is merely one type of association relationship that describes an associated object, meaning that three relationships may exist, e.g., a and/or B may mean: a exists alone, A and B exist simultaneously, and B exists alone. In addition, the character "/" herein generally indicates that the former and latter related objects are in an "or" relationship.
In the embodiments provided herein, it should be understood that "B corresponding to a" means that B is associated with a from which B can be determined. It should also be understood that determining B from a does not mean determining B from a alone, but may be determined from a and/or other information.
As shown in fig. 1 to 3, the present invention provides a method for intelligent inversion optimization of a drive chain system, comprising the following steps;
s1, performing modeling calculation on electromagnetic response of a transmission chain system under an SMC-SVPWM sliding mode control algorithm to obtain the overall efficiency of the transmission chain system under the SMC-SVPWM sliding mode control algorithm; the SMC-SVPWM is space vector pulse width modulation based on a sliding mode control algorithm;
the system mainly comprises a transmission chain control algorithm, a driving external circuit and a generator electromagnetic model, wherein each part is used for carrying out real-time data communication and interactive calculation through a model interface, so that the simulation of a transmission chain system under a specific control algorithm is completed;
s2, selecting the length variable of the permanent magnet on the basis of S1
Width, width
Tooth space parameters
To optimize the variables; establishing a Morris random sampling matrix of motor parameters, calculating the electromagnetic response of a sample by an S1 modeling calculation method, carrying out elementary effect analysis on the motor parameters to quantitatively measure the sensitivity of each parameter and the interactivity among the parameters, establishing an orthogonal test table of variables by a Taguchi method, carrying out signal-to-noise ratio analysis on the orthogonal test table, and preliminarily screening out the sensitive variables of the motor;
s3, selecting the length variable of the permanent magnet through S2
Width, width
Tooth space parameters
As sensitive variables, BS1=4.5mm, HS1=1.7mm were determined at the same time; the procedure and principle for determining the sensitive variables is as follows:
screening the sensitivity of the whole efficiency of the transmission chain to each parameter by utilizing an elementary effect distribution diagram, and constructing a Morris statistical matrix to carry out elementary effect analysis on variables, namely:
in the formula (I), the compound is shown in the specification,
is as follows
The distribution of the elementary effects of the individual variables,
in the form of increments of the number of bits,
is the variable of the k-th weft yarn,
in the x-direction,
for drive chain efficiency;
the normalized (k + 1) k sampling matrix, where only two rows per column differ at the ith position, is constructed by a random permutation algorithm to compute the matrix X, i.e.:
performing r times of elementary effect calculation on the X matrix to obtain elementary effect distribution of different variables;
in the formula (I), the compound is shown in the specification,
is a unit vector of k +1 columns,
in order to randomly initialize the increment of the initialization,
is a row vector generated by random increment and having the same dimension as the variable and each element being a multiple of the random increment not exceeding 1,
is a matrix whose elements are only 0, 1,
for a zero vector with only two elements 1,
is a matrix of the units,
for randomly initialized elements with diagonals of only 1, -1,
performing the sampling process r times, and analyzing factors of each behavior elementary effect of the X matrix;
the influence proportion of each variable on the efficiency of the transmission chain is quantitatively analyzed, variance analysis is carried out on the influence proportion, the variance value of each variable is calculated, and the formula of the calculated variance value is as follows:
in the formula (I), the compound is shown in the specification,
in order to obtain the number of tests,
the number of the variable levels is the number of the variable levels,
for each number of horizontal experiments the number of experiments,
in order to achieve the target value,
represents the first
A variable of
One level corresponds to all
Is determined by the average value of (a) of (b),
representing all times of test
Average value of (d);
finally, the elementary effect and the analysis of variance are combined for use, the difference between the horizontal means of one or more factors is checked, and the variable with larger average value of the elementary effect and larger amplitude change in the analysis of variance is selected as a sensitive variable;
s4, generating a sampling scheme by adopting a Latin hypercube sampling scheme with high uniformity according to the sensitive variables in the S3, wherein the Latin hypercube sampling can ensure that the projection of the generated sampling scheme to each variable dimension in a test space is uniformly distributed; the sampling method only needs to ensure that the sampling parameters are uniformly distributed in a test space through optimization design, so that the optimal sampling scheme is obtained by establishing norm functions of two sample points in a sample space and a score function of sampling and combining a genetic algorithm to carry out optimization design.
The sampling scheme comprises the following steps:
first by randomly generating different hyper-latin cubic samples of 6 dimensions, 120 sample points,
establishing a distance norm and score function of two spatial points for the established Latin hypercube sampling scheme:
in the formula (I), the compound is shown in the specification,
is the spatial norm between the two samples,
is a norm of a sample
The number of (2);
the parameters are calculated for the distance and,
in order to optimize the parameters to be optimized,
is as follows
A first sample of
The number of the variables is one,
is as follows
A first sample of
The number of the variables is one,
is the total number of the samples and is,
is composed of
The number of the variable is changed according to the number of the variable,
is composed of
The number of the variables is one,
is the variable level number;
optimizing the local optimal test design schemes under different q through a genetic algorithm, and sequencing all local sequencing schemes through a maximum minimum criterion after determining the local optimal design schemes under different q to select the scheme with the best spatial uniformity:
in the formula (I), the compound is shown in the specification,
is the spatial norm between the two samples,
is a norm of a sample
The number of (2);
s5, carrying out transmission chain integrated simulation on the sample points in the sampling scheme to obtain efficiency response; selecting 70% of sample points to establish a Gaussian random process model, and obtaining an intelligent inversion optimization model by adopting an inversion optimization design method taking a target as a guide on the basis of establishing the Gaussian random process model; the Gaussian random process model can avoid complex nonlinear and high-dimensional mathematical relations between response variables and optimization variables, and is predicted and output by establishing Gaussian joint distribution between a sample and a test sample;
the specific steps and flows are as follows;
establishing a Gaussian random process model for 70% of sample points by combining maximum likelihood estimation with a genetic algorithm, namely, a data-based maximum likelihood estimation function:
in the formula:
the number of the samples is the number of the samples,
in response to this, the mobile station is allowed to respond,
is a covariance matrixThe matrix, X being the matrix of sampled samples, Y being the response vector, where,
and with
The relation of the functions can be obtained by combining a genetic algorithm with maximum likelihood estimation;
after the model is established, the specific optimization target value is set to be 97%, and the addition point of the new model parameter and the optimal improved model probability is determined by combining the genetic algorithm through maximum likelihood estimation according to the set target response;
wherein, the new likelihood function also includes target optimization efficiency besides the model parameters, namely:
in the formula (I), the compound is shown in the specification,
the number of the samples is the number of the samples,
in response to this, the mobile station is allowed to respond,
is a covariance matrix, X is a matrix of sampled samples, Y is a response vector,
is the set target value.
In addition, the design target is parameterized and then is included in the calculation of model parameters of the Gaussian random process, the maximum likelihood estimation is adopted to combine with the genetic algorithm to update the model parameters and calculate the optimal point adding position, the calculated optimal point adding position is synchronously updated to the model, so that the accuracy of the model is improved, for example, in the optimized modeling of the embodiment, 78 points are selected to establish a basic model, and the accuracy of the model can be improved to 99.862% through 21 times of point adding.
And S6, on the basis of the intelligent inversion optimization model, considering the integral output torque of the transmission chain system as constraint, and optimizing the constraint by adopting a PESA-II (particle swarm optimization-II) range selection-based multi-objective optimization algorithm. Specifically, based on the intelligent inversion optimization model, a PESA-II multi-objective optimization algorithm based on range selection is adopted to optimize the model. PESA-II sets an outer population and an inner population. During evolution, non-dominant individuals in the inner population are added to the outer population, and dominant individuals in the outer population are eliminated. And when the crowding coefficients of the individuals are the same, randomly deleting one individual, and repeating the process until the population number meets the upper limit. In each evolution, the internal population is emptied, then individuals are selected from the external population, crossover and mutation are carried out according to certain probability to obtain new individuals, and the new individuals are added into the internal population. The final solution of the algorithm is the individual of the external population, namely, several Pareto optimal solutions. And selecting a feasible non-dominant solution according to the actual design requirement to obtain an optimized design scheme.
Therefore, the electromagnetic performance of the motor under the actual PWM waveform is considered, so that the optimized design scheme has reliability.
Compared with the traditional proxy model, the accuracy of the intelligent inversion optimization method for the transmission chain system can be ensured.
Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the technical solutions of the present invention have been described in detail with reference to the foregoing embodiments, it should be understood by those skilled in the art that the technical solutions described in the foregoing embodiments can be modified or some technical features can be replaced equally; and the modifications or the substitutions do not make the essence of the corresponding technical solutions depart from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims (6)
1. An intelligent inversion optimization method for a transmission chain system is characterized by comprising the following steps: comprises the following steps;
s1, performing modeling calculation on electromagnetic response of a transmission chain system under an SMC-SVPWM sliding mode control algorithm to obtain the overall efficiency of the transmission chain system under the SMC-SVPWM sliding mode control algorithm;
s2, selecting the length variable of the permanent magnet on the basis of S1
Width, width
Tooth space parameters
For optimizing variables, BS0 is the width of the notch, BS1 is the width of the main body portion of the groove near the notch, BS2 is the width of the main body portion of the groove near the bottom of the groove, HS0 is the depth of the notch, HS1 is the depth of the gradual change portion from the notch to the main body portion of the groove, and HS2 is the depth of the main body portion of the groove; establishing a Morris random sampling matrix of motor parameters, calculating the electromagnetic response of a sample by an S1 modeling calculation method, carrying out primary effect analysis on the motor parameters to quantitatively measure the sensitivity of each parameter and the interactivity between the parameters, establishing an orthogonal test table of variables by a Taguchi method, carrying out signal-to-noise ratio analysis on the orthogonal test table, and preliminarily screening out the sensitive variables of the motor;
s3, selecting the length variable of the permanent magnet through S2
Width, width
Tooth space parameters
As sensitive variables, BS1=4.5mm, HS1=1.7mm were determined at the same time;
quantitatively analyzing the influence proportion of each variable on the efficiency of the transmission chain, carrying out variance analysis on the influence proportion, and calculating the variance value of each variable;
finally, the elementary effect and the analysis of variance are combined for use, the difference between the horizontal means of one or more factors is checked, and the variable with larger average value of the elementary effect and larger amplitude change in the analysis of variance is selected as a sensitive variable;
s4, generating a sampling scheme by adopting a Latin hypercube sampling scheme with higher uniformity according to the sensitive variable in S3; the sampling scheme comprises the following steps:
firstly, sampling by randomly generating different hyper-Latin cubes with 6 dimensions and 120 sample points;
establishing a distance norm and score function of two spatial points for the established Latin hypercube sampling scheme:
in the formula (I), the compound is shown in the specification,
is the spatial norm between the two samples,
is a norm of a sample
The number of (2);
the parameters are calculated for the distance and,
in order to optimize the parameters to be optimized,
is as follows
A first sample of
The number of the variables is one,
is as follows
A first sample of
The number of the variables is one,
is the total number of the samples and is,
is composed of
The number of the variables is one,
is composed of
The number of the variables is one,
is the variable level number;
optimizing the local optimal test design schemes under different q through a genetic algorithm, and sequencing all local sequencing schemes through a maximum minimum criterion after determining the local optimal design schemes under different q to select the scheme with the best spatial uniformity:
in the formula (I), the compound is shown in the specification,
is the spatial norm between the two samples,
is a norm of a sample
The number of (2);
s5, carrying out transmission chain integrated simulation on the sample points in the sampling scheme to obtain efficiency response; selecting 70% of sample points to establish a Gaussian random process model, and obtaining an intelligent inversion optimization model by adopting an inversion optimization design method taking a target as a guide on the basis of establishing the Gaussian random process model;
and S6, on the basis of the intelligent inversion optimization model, considering the integral output torque of the transmission chain system as constraint, and optimizing the constraint by adopting a PESA-II (particle swarm optimization-II) range selection-based multi-objective optimization algorithm.
2. The intelligent inversion optimization method of the drive chain system according to claim 1, wherein: the steps for determining the sensitive variable in S3 are as follows:
screening the sensitivity of the whole efficiency of the transmission chain to each parameter by utilizing an elementary effect distribution diagram, and constructing a Morris statistical matrix to carry out elementary effect analysis on variables, namely:
in the formula (I), the compound is shown in the specification,
is as follows
The distribution of the elementary effects of the individual variables,
in the form of increments of the number of bits,
is the variable of the k-th weft yarn,
in the x-direction,
for drive chain efficiency;
and (4) constructing a calculation matrix X by a random permutation algorithm, wherein each column of the normalized (k + 1) k sampling matrix has only two rows which are different at the ith position.
3. The drive chain system intelligent inversion optimization method of claim 2, wherein: the calculation matrix X is as follows:
in the formula (I), the compound is shown in the specification,
is a unit vector of k +1 columns,
in order to randomly initialize the increment of the initialization,
is a row vector generated by random increment and having the same dimension as the variable and each element being a multiple of the random increment not exceeding 1,
is a matrix whose elements are only 0, 1,
for a zero vector with only two elements 1,
is a matrix of the unit, and is,
for randomly initialized elements with diagonals of only 1, -1,
performing the sampling process r times, and analyzing factors of each behavior elementary effect of the X matrix;
and (4) performing r times of elementary effect calculation on the X matrix to obtain elementary effect distribution of different variables.
4. The drive chain system intelligent inversion optimization method of claim 3, wherein: the variance SSX in S3 is expressed as:
in the formula (I), the compound is shown in the specification,
in order to obtain the number of tests,
the number of the variable levels is the number of the variable levels,
for each number of horizontal experiments the number of experiments,
is a target value for the amount of time,
represents the first
A variable of
One level corresponds to all
Is determined by the average value of (a) of (b),
representing all times of test
Average value of (a).
5. The drive chain system intelligent inversion optimization method of claim 1, wherein: on the basis of establishing a Gaussian random process model, a specific flow for obtaining an intelligent inversion optimization model by adopting an inversion optimization design method taking a target as a guide is as follows;
establishing a Gaussian random process model for 70% of sample points by combining maximum likelihood estimation with a genetic algorithm, namely, a data-based maximum likelihood estimation function:
in the formula:
the number of the samples is the number of the samples,
in response to this, the mobile station is allowed to respond,
is a covariance matrix, X is a matrix of sampled samples, Y is a response vector, wherein,
and
the relationship of the functions can be obtained by combining genetic algorithm with maximum likelihood estimation.
6. The drive chain system intelligent inversion optimization method of claim 5, wherein: after the model is established, the specific optimization target value is set to be 97%, and the addition point of the new model parameter and the optimal improved model probability is determined by combining the genetic algorithm through maximum likelihood estimation according to the set target response;
wherein, the new likelihood function also includes target optimization efficiency besides the model parameters, namely:
in the formula (I), the compound is shown in the specification,
the number of the samples is the number of the samples,
in response to this, the mobile station is allowed to respond,
is a covariance matrix, X is a matrix of sampled samples, Y is a response vector,
is the set target value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210566017.8A CN114662413B (en) | 2022-05-24 | 2022-05-24 | Intelligent inversion optimization method for transmission chain system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210566017.8A CN114662413B (en) | 2022-05-24 | 2022-05-24 | Intelligent inversion optimization method for transmission chain system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114662413A CN114662413A (en) | 2022-06-24 |
| CN114662413B true CN114662413B (en) | 2022-08-09 |
Family
ID=82036665
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210566017.8A Active CN114662413B (en) | 2022-05-24 | 2022-05-24 | Intelligent inversion optimization method for transmission chain system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114662413B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110555249A (en) * | 2019-08-20 | 2019-12-10 | 湖南大学 | motor parameter design method based on global optimal water pump load annual loss power consumption |
| CN113468682A (en) * | 2021-06-16 | 2021-10-01 | 江苏大学 | Permanent magnet motor robust optimization design method considering magnetic material uncertainty |
| CN113627000A (en) * | 2021-07-30 | 2021-11-09 | 江苏大学 | Permanent magnet motor layered robust optimization design method based on parameter sensitive domain |
| CN113626942A (en) * | 2021-06-03 | 2021-11-09 | 西北工业大学 | Double-amplitude turbine disk fatigue creep life reliability optimization method based on proxy model |
| CN113673129A (en) * | 2021-08-25 | 2021-11-19 | 合肥工业大学 | Multi-objective robust optimization design method for ultra-high efficiency induction motor |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US11195057B2 (en) * | 2014-03-18 | 2021-12-07 | Z Advanced Computing, Inc. | System and method for extremely efficient image and pattern recognition and artificial intelligence platform |
-
2022
- 2022-05-24 CN CN202210566017.8A patent/CN114662413B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110555249A (en) * | 2019-08-20 | 2019-12-10 | 湖南大学 | motor parameter design method based on global optimal water pump load annual loss power consumption |
| CN113626942A (en) * | 2021-06-03 | 2021-11-09 | 西北工业大学 | Double-amplitude turbine disk fatigue creep life reliability optimization method based on proxy model |
| CN113468682A (en) * | 2021-06-16 | 2021-10-01 | 江苏大学 | Permanent magnet motor robust optimization design method considering magnetic material uncertainty |
| CN113627000A (en) * | 2021-07-30 | 2021-11-09 | 江苏大学 | Permanent magnet motor layered robust optimization design method based on parameter sensitive domain |
| CN113673129A (en) * | 2021-08-25 | 2021-11-19 | 合肥工业大学 | Multi-objective robust optimization design method for ultra-high efficiency induction motor |
Non-Patent Citations (1)
| Title |
|---|
| 超超高效感应电机设计及其多目标鲁棒优化设计;尉恬;《中国优秀博硕士学位论文全文数据库(硕士)》;中国学术期刊（光盘版）电子杂志社;20220215;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114662413A (en) | 2022-06-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109117951B (en) | BP neural network-based probability load flow online calculation method | |
| CN102331966A (en) | Software test data evolution generation system facing path | |
| CN108121215B (en) | Process control loops method of evaluating performance and device based on full loop reconstruct emulation | |
| Arias-Rosales et al. | Wind turbine selection method based on the statistical analysis of nominal specifications for estimating the cost of energy | |
| CN109753763B (en) | Wind-solar combined output probability modeling method | |
| CN105931134A (en) | Method for evaluating reliability of system with grid-connected photovoltaic power station | |
| CN110110434A (en) | A kind of initial method that Probabilistic Load Flow deep neural network calculates | |
| Liang et al. | An enhanced slime mould algorithm and its application for digital IIR filter design | |
| CN117909855B (en) | Data monitoring and auxiliary governance method based on electric power model | |
| CN117407802A (en) | Runoff prediction method based on improved depth forest model | |
| CN116894504A (en) | Wind power cluster power ultra-short-term prediction model establishment method | |
| CN109408896B (en) | Multi-element intelligent real-time monitoring method for anaerobic sewage treatment gas production | |
| CN114662413B (en) | Intelligent inversion optimization method for transmission chain system | |
| Zhang et al. | A method based on Lorenz disturbance and variational mode decomposition for wind speed prediction | |
| Bi et al. | Multi-swarm genetic gray wolf optimizer with embedded autoencoders for high-dimensional expensive problems | |
| CN111259340B (en) | Saturation load prediction method based on logistic regression | |
| CN110674460B (en) | E-Seq2Seq technology-based data driving type unit combination intelligent decision method | |
| CN104156462A (en) | Complex network community mining method based on cellular automatic learning machine | |
| CN117057265A (en) | Flow field prediction method based on migration learning and discrete physical information neural network | |
| Ping et al. | A mixed isomorphism approach for kinematic structure enumeration graphs based on intelligent design and manufacturing | |
| CN106651163A (en) | Capacity confidence level evaluation method for multiple wind power plants on the basis of Copula function | |
| CN113205171B (en) | Multi-objective optimization method for discrete binary particle swarm motor | |
| CN113722853B (en) | Group intelligent evolutionary engineering design constraint optimization method for intelligent computation | |
| CN113935556B (en) | Temperature sensor optimal arrangement method based on DNA genetic algorithm | |
| CN113435628B (en) | Medium-long-term runoff prediction method and system based on linear discriminant analysis and IALO-ELM |
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 |