CN113162375B - Modeling method for switch loss prediction in IGBT dynamic process - Google Patents
Modeling method for switch loss prediction in IGBT dynamic process Download PDFInfo
- Publication number
- CN113162375B CN113162375B CN202110477808.9A CN202110477808A CN113162375B CN 113162375 B CN113162375 B CN 113162375B CN 202110477808 A CN202110477808 A CN 202110477808A CN 113162375 B CN113162375 B CN 113162375B
- Authority
- CN
- China
- Prior art keywords
- krill
- turn
- igbt
- data
- loss
- 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.)
- Expired - Fee Related
Links
- 238000000034 method Methods 0.000 title claims abstract description 61
- 241000239366 Euphausiacea Species 0.000 claims abstract description 135
- 238000012360 testing method Methods 0.000 claims abstract description 59
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 46
- 238000005457 optimization Methods 0.000 claims abstract description 8
- 238000012545 processing Methods 0.000 claims abstract description 7
- 241000239370 Euphausia superba Species 0.000 claims description 39
- 235000009508 confectionery Nutrition 0.000 claims description 27
- 230000002431 foraging effect Effects 0.000 claims description 12
- 239000011159 matrix material Substances 0.000 claims description 11
- OKTJSMMVPCPJKN-UHFFFAOYSA-N Carbon Chemical group [C] OKTJSMMVPCPJKN-UHFFFAOYSA-N 0.000 claims description 9
- 230000006698 induction Effects 0.000 claims description 9
- 238000012216 screening Methods 0.000 claims description 9
- 238000004364 calculation method Methods 0.000 claims description 8
- 238000009792 diffusion process Methods 0.000 claims description 8
- 238000010606 normalization Methods 0.000 claims description 8
- 241000238557 Decapoda Species 0.000 claims description 6
- 238000011156 evaluation Methods 0.000 claims description 6
- 230000000694 effects Effects 0.000 claims description 4
- 241000237537 Ensis Species 0.000 claims description 3
- 241000239368 Euphausia Species 0.000 claims description 3
- 239000004784 Superba Substances 0.000 claims description 3
- NAWXUBYGYWOOIX-SFHVURJKSA-N (2s)-2-[[4-[2-(2,4-diaminoquinazolin-6-yl)ethyl]benzoyl]amino]-4-methylidenepentanedioic acid Chemical compound C1=CC2=NC(N)=NC(N)=C2C=C1CCC1=CC=C(C(=O)N[C@@H](CC(=C)C(O)=O)C(O)=O)C=C1 NAWXUBYGYWOOIX-SFHVURJKSA-N 0.000 claims description 2
- 102100029469 WD repeat and HMG-box DNA-binding protein 1 Human genes 0.000 claims description 2
- 101710097421 WD repeat and HMG-box DNA-binding protein 1 Proteins 0.000 claims description 2
- 235000013305 food Nutrition 0.000 claims description 2
- 230000002093 peripheral effect Effects 0.000 claims description 2
- 241000238565 lobster Species 0.000 claims 1
- 230000017525 heat dissipation Effects 0.000 abstract description 3
- 238000013021 overheating Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 3
- 238000013178 mathematical model Methods 0.000 description 3
- OAICVXFJPJFONN-UHFFFAOYSA-N Phosphorus Chemical compound [P] OAICVXFJPJFONN-UHFFFAOYSA-N 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 229910052698 phosphorus Inorganic materials 0.000 description 2
- 239000011574 phosphorus Substances 0.000 description 2
- 238000013459 approach Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000002349 favourable effect Effects 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 238000000691 measurement method Methods 0.000 description 1
- 238000004806 packaging method and process Methods 0.000 description 1
- 238000010248 power generation Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 230000003685 thermal hair damage Effects 0.000 description 1
Images
Classifications
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02M—APPARATUS FOR CONVERSION BETWEEN AC AND AC, BETWEEN AC AND DC, OR BETWEEN DC AND DC, AND FOR USE WITH MAINS OR SIMILAR POWER SUPPLY SYSTEMS; CONVERSION OF DC OR AC INPUT POWER INTO SURGE OUTPUT POWER; CONTROL OR REGULATION THEREOF
- H02M1/00—Details of apparatus for conversion
- H02M1/08—Circuits specially adapted for the generation of control voltages for semiconductor devices incorporated in static converters
-
- 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/30—Circuit design
- G06F30/36—Circuit design at the analogue level
- G06F30/367—Design verification, e.g. using simulation, simulation program with integrated circuit emphasis [SPICE], direct methods or relaxation methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Computer Hardware Design (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Software Systems (AREA)
- Computing Systems (AREA)
- Geometry (AREA)
- Molecular Biology (AREA)
- Data Mining & Analysis (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Mathematical Physics (AREA)
- Biomedical Technology (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Power Engineering (AREA)
- Microelectronics & Electronic Packaging (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Medical Informatics (AREA)
- Feedback Control In General (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to a modeling method for predicting switching loss in an IGBT dynamic process, which is an IGBT switching loss prediction method for optimizing an extreme learning machine based on a good krill swarm algorithm, and comprises the following steps: firstly, acquiring dynamic characteristic test data of the IGBT; and secondly, after the test data processing and basic parameter setting of the extreme learning machine and krill cluster algorithm are completed, optimizing the initial krill clusters by using a good point set algorithm, taking the initial krill clusters as weight threshold values of the extreme learning machine, and calculating the fitness of the good point krill. In the optimizing process, continuously updating the position of the krill with the Levy flight and cosine control factors as wings, and calculating the fitness of the krill until the optimizing process is finished; and finally, predicting and outputting IGBT (insulated gate bipolar translator) on-off loss values according to the optimal weight threshold value of the extreme learning machine found by the krill at the best point. The method has the advantages that the algorithm optimization is carried out, the dynamic adjustment is carried out, the prediction accuracy of the prediction model is high, the prediction speed is high, and the prediction result has better guiding significance for engineers to improve the heat dissipation system of the IGBT module and the like.
Description
Technical Field
The technical scheme of the invention belongs to the technical field of reliability of power electronic devices (IGBT), and particularly relates to a modeling method for predicting switching loss in an IGBT dynamic process.
Background
With the increasing energy crisis, the power electronic technology is continuously developed, and the progress and development of the society are effectively promoted. As a modern power electronic switch, the IGBT is widely used in the fields of power systems, electric vehicles, high-speed traction, and the like. However, the failure of an IGBT-based photovoltaic inverter accounts for about 37% of the total failure. In a failure fault of a power electronic system, the fault caused by temperature accounts for about 55% of the total fault, and the temperature of the device, with its safety margin and thermal cycle life, presents a negative correlation. The occurrence of the fault seriously affects the normal operation of the system and reduces the reliability of the system operation. Therefore, it is important to study the overheating fatigue and the overheating loss during the operation of the IGBT.
The study of scholars at home and abroad on the overheating fatigue and the overheating loss of the IGBT mainly focuses on the calculation aspect of the switching loss of the IGBT. The switching loss of the IGBT means power consumed at the IGBT in one cycle. The switching loss has a close relationship with the IGBT collector-emitter voltage, collector current, switching frequency, and resistance and voltage in its driving circuit. Moreover, the reliability of the IGBT is greatly related to temperature rise fluctuation generated by the switching loss of the IGBT, and the temperature rise to a certain threshold value can cause thermal damage to the device. Therefore, if modeling prediction is carried out on the switching loss of the IGBT, the method is beneficial to the packaging heat dissipation design, the driving circuit design and the like of the device, and the failure of the device caused by overlarge temperature rise fluctuation due to the heat generated by the switching loss of the IGBT when the device runs is avoided. Therefore, the prediction of the switching loss of the IGBT has a certain meaning. In recent years, some scholars have achieved a certain level of performance by predicting the switching loss of the IGBT.
The current methods for calculating the switching loss of the IGBT are roughly divided into three types: switching loss calculation based on a physical model, a calculation method based on a mathematical model and switching loss prediction based on an intelligent model. Firstly, the switching loss calculation based on the physical model simulates the dynamic characteristics of an IGBT by using simulation software to further calculate the switching loss of the IGBT, the calculation result has high precision, but the process of constructing the model is complex and the simulation speed is slow; secondly, switching loss calculation is carried out based on a mathematical model, a common method is to consult an IGBT technical manual to calculate the switching loss, but the difference between a calculated value and an actual value is large, and although the prediction precision of the polynomial model is improved, the prediction speed is relatively slow; and finally, the switching loss prediction based on the intelligent model is carried out, the prediction precision and the prediction speed of the intelligent model are improved compared with those of the former two, but in the aspect of parameter selection, if the intelligent model is improperly selected, the intelligent model falls into a local optimal solution, and the model is not favorable for searching a global optimal solution.
Therefore, the method aims at the problems that the prediction precision of the existing measurement method is not high, the optimal intelligent model parameters are selected to predict the IGBT switching loss and the like. The method takes an extreme learning machine as a theoretical main body, takes a good point set, a krill group algorithm and a cosine control factor as filling, takes collector current, direct-current bus voltage, switching frequency and gate voltage as input, and takes turn-on loss and turn-off loss as output to establish a mathematical model for predicting the switching loss in the dynamic process of the IGBT. The invention takes multivariable input into full consideration of various factors influencing the IGBT switching loss, improves the prediction precision of the switching loss, and is not easy to fall into a local optimal solution, so the invention has certain practical value.
Disclosure of Invention
The technical problem to be solved by the invention is as follows: the method comprises the steps of obtaining direct-current bus voltage, collector current, gate voltage and switching frequency data of an IGBT module and turn-on and turn-off loss data of the IGBT module based on a dynamic characteristic test of the IGBT power module, taking the direct-current bus voltage, the collector current, the gate voltage and the switching frequency data as input, taking the turn-on and turn-off loss data as output, and establishing a multi-parameter IGBT switching loss prediction model which is theoretically supported by a extreme learning machine (GKH-ELM) optimized by a good krill group algorithm; compared with the existing method, the krill group algorithm is changed into the krill group algorithm through optimization of the sweet spot set algorithm, and when the algorithm is used for population initialization, the individual distribution of the sweet spot krill is more uniform, so that the global optimal solution can be searched; by introducing cosine control factors and Levy flight, parameters of the good krill mass algorithm are slowly accelerated in the early and middle stages and gradually decelerated in the middle and later stages to gradually approach global convergence, so that dynamic search of the good krill mass is realized; the GKH-ELM overcomes the defect of the prediction precision of the existing model, and has important practical significance for improving the working reliability of the IGBT.
The technical scheme adopted by the invention for solving the technical problem is as follows: a modeling method for predicting switching loss in an IGBT dynamic process is a method for predicting the IGBT switching loss of an optimal krill swarm optimization extreme learning machine, and comprises the following steps:
step one, obtaining IGBT dynamic characteristic test data
(1.1) acquiring m groups of test data through an IGBT dynamic characteristic test, wherein each group of data comprises data of direct current bus voltage, collector current, gate voltage and switching frequency of an IGBT module, and data of turn-on and turn-off loss of the IGBT module;
step two, carrying out normalization processing and distribution on the IGBT dynamic characteristic test data
(2.1) carrying out normalization processing on the IGBT dynamic characteristic test data by using a normalization formula (1):
(2.2) dividing the normalized test data into learning data and test data, wherein the distribution proportion is as follows:
the number of learning data and the number of testing data are A: B;
therefore, the m groups of dynamic characteristic test data of the IGBT are divided into m × A/(A + B) group study data and m × B/(A + B) group test data; each group comprises six variables, the turn-on and turn-off losses of the IGBT module, and the direct-current bus voltage, the collector current, the gate voltage and the switching frequency of the IGBT module;
step three, setting and initializing extreme learning machine parameters
The parameters of the extreme learning machine need to set the number of nodes of each layer and the weight and threshold of the initialized extreme learning machine, and the method comprises the following steps:
setting the number of nodes of an input layer of the extreme learning machine to num _ in;
setting the number of hidden layer nodes of the extreme learning machine to num _ hid;
setting the number of output layer nodes of the extreme learning machine to num _ out, wherein num _ out is 1 because of respective prediction although the turn-on or turn-off loss of the IGBT is predicted;
step four, basic parameters of the krill group algorithm are set, and distribution of weights and thresholds from the input layer of the extreme learning machine to the hidden layer nodes is completed
(4.1) the parameters of the krill population algorithm are required to set the number of krill in the initial krill population, the maximum number of iterations for the krill population optimization, the dimensions of the data contained in each krill and the initial data of the krill population, including:
setting the number of krill in the initial krill mass to n;
setting the maximum Iteration number of optimizing the shrimp group as Iteration _ max;
setting the dimension of data contained in each krill to dim; the dim is determined by an input layer node and an implicit layer node of the extreme learning machine, and is num _ in × num _ hid + num _ hid;
initializing the population data of the phopenaeus ensis group, and taking the population data as the weight and the threshold value from the input layer of the extreme learning machine to the node of the hidden layer. The method comprises the following steps: using MATLAB software, adopting cyclotomic field method to generate a set X containing n points in a unit space of dim dimensionn(i)，Xn(i) The structural formula of (2):
Xn(i)＝{[(r1×i)，(r2×i)，...，(rj×i)，...，(rdim×i)]，1≤i≤n，1≤j≤dim} (2)
in the formula, dim represents the data dimension of each sweet spot;
and (r)jX i) is rjThe fraction of xi, q is the smallest prime number satisfying 2 xdim +3 ≦ q. The optimal point matrix X is obtained by equation (2):
in matrix X, Xij∈[Lb，Ub](1≤i≤n，1≤j≤dim)，Ub、LbAre respectively a variable xijUpper and lower boundaries of (1). And taking the optimal point matrix X as a data matrix of the first generation krill population in the krill population algorithm, and recording as an optimal point phopsophila shrimp population. Thus, in formula (3), X1，X2，...，Xi，...，XnRepresents the 1 st, 2 nd, 3 rd.. once, i.. once, n euphausiids in the euphausiids, xijRepresenting data on the jth dimension on the ith sweet krill. Therefore, the initialization of the weight and the threshold from the input layer of the extreme learning machine to the node of the hidden layer is finished.
(4.2) in consideration of the number of nodes of the input layer and the hidden layer of the extreme learning machine, distributing the dim-dimensional data of each krill, taking the [1, num _ in × num _ hid ] dimensional data of the krill as the weight value from the input layer of the extreme learning machine to the hidden layer, taking the [ num _ in × num _ hid +1, dim ] dimensional data of the krill as the threshold value from the input layer of the extreme learning machine to the hidden layer, and introducing the distributed data into the extreme learning machine. So far, the parameter setting of the extreme learning machine is basically finished.
Fifthly, optimizing the extreme learning machine by using the best krill swarm algorithm, and calculating the fitness value of each krill
(5.1) establishing a prediction model of the krill goodpoint algorithm-extreme learning machine, namely a GKH-ELM prediction model, and predicting the turn-on loss or turn-off loss of the IGBT. Taking data of each krill in the krill group algorithm as a set of weight and threshold from an input layer node to a hidden layer node of the GKH-ELM prediction model; importing the mxA/(A + B) omic learning data obtained in the step (2.2) into an extreme learning machine, wherein the learning data are divided into two parts: input data and output data. The input data comprises 4 variables which are respectively the direct current bus voltage, the collector current, the gate voltage and the switching frequency of the IGBT module, and the output data is the turn-on loss or the turn-off loss of the IGBT module. After the learning data is imported into the extreme learning machine, n groups are obtained, the predicted values of the turn-on loss or the turn-off loss of each group of m × a/(a + B) IGBTs are obtained, the fitness value of each euphausia superba is calculated through a formula (4), namely the predicted performance evaluation value of the extreme learning machine corresponding to each euphausia superba,
in the formula (4), F is the fitness value of each euphausia superba, and the larger the fitness value is, the better the fitness value is; p is a radical ofkTaking the data obtained by the dynamic characteristic test of the IGBT as the actual value of the turn-on loss or turn-off loss of the IGBT; p'kThe predicted value of the turn-on loss or turn-off loss of the IGBT is obtained; n is the predicted output number of the turn-on loss or the turn-off loss of the extreme learning machine, namely the learning period N is m multiplied by A/(A + B). Similarly, during the test, N ═ m × B/(a + B).
And (5.2) screening the euphausia superba with high fitness, and recording and retaining the position data of the euphausia superba with the highest fitness value in the euphausia superba group algorithm. Screening of krill with high fitness comprises: screening current best krill with the best fitness, comparing the fitness values of the current best krill with the fitness values of historical best krill, and selecting the best. That is, if
Recording the position data of the t movement of the euphausia superba i; if it is
The current generation of the optimal sweet krill location data is recorded and recorded as historical optimal sweet krill.
Step six, improving the krill goodpoint algorithm and searching a global optimal solution
(6.1) updating the superbus krill group X obtained in the step four or the formula (13). Location updating exercise of individual euphausia superba, mainly comprising induced exercise IiForaging movement FiAnd physical diffusion movement PiThe formulas are respectively shown in (5), (6) and (7):
in the formula (5), the first and second groups,
represents the t-th induced movement of the ith euphausia superba, and the same principle is adopted
Represents the t-1 induced movement of the ith euphausia superba; i ismaxExpressing the maximum induction rate, take Imax＝0.01；
Shows the effect of the t-th time of the peripheral euphausia superba on the ith euphausia superba
Representing the induction effect of the historical optimal krill on the krill i; omegaIInertial weight to induce motion, range is [0, 1 ]]。
In the formula (6), Fi tRepresents the t-th foraging movement of the ith euphausia superba, and has the same principle Fi t-1Represents the t-1 foraging movement of the ith euphausia superba;
indicating the induction intensity of the current food to the ith krill sweet spot,
representing the foraging traction of the historical optimal krill on the movement of the krill i; v. ofFExpressing the foraging speed of krill, vF＝0.02；ωFThe weight for induction of foraging motion is in the range of [0, 1%]。
In the formula (7), the first and second groups,
represents the t-th physical diffusion movement of the ith euphausia superba; pmaxRepresenting the maximum rate of diffusion of the krill at the sweet spot, taking Pmax0.005; t represents the current number of movements; deltaiIs the direction vector of the random diffusion of the current sweet krill, and is deltai∈[0，1]。
Therefore, the amount of change in the positional movement of the euphausia superba i at the t-th movement
See formula (8) and the position after the end of the tth movement see formula (9):
recording the position of the krill i after the end of the tth movement
As the initial position for the t +1 th motion.
(6.2) introducing a cosine control factor and a position updating formula of the Levy flight improvement best point shrimp group. The cosine control factor is used for improving omega in the induction motion formula (5) and foraging motion formula (6) of euphausia superbaIAnd ωFThe formula is improved as shown in formula (10).
The levy flight is used to refine equation (9) and is set. If the optimal fitness value of the superbus group does not change after more than 20 iterations, the Levy flight is started until the optimal fitness value of the krill group changes. See lewy flight formula (11):
where u and epsilon both follow a standard normal distribution and β is 1.5, and the calculation formula of phi is shown in equation (12).
Therefore, the modified superba location obtained by combining equations (9), (10), (11), and (12) is updated to equation (13), and thus the location of each individual in the superba group X is updated
The position is updated according to equation (13) to generate a new position.
And step seven, judging whether the current movement times t of the krill groups reach the maximum Iteration times Iteration _ max set by the krill group algorithm. If t is less than Iteration _ max, returning to the step five; if t is equal to Iteration _ max, step eight is executed.
Step eight, outputting the prediction results of the turn-on loss and the turn-off loss of the IGBT
Taking the historical optimal krill position data obtained in the fifth step as the optimal weight and threshold of the GKH-ELM prediction model; taking m × B/(a + B) data in m groups of dynamic test characteristic test data as test data, wherein each group of test data comprises: the turn-on or turn-off loss of the IGBT module, and the direct current bus voltage, the collector current, the gate voltage and the switching frequency of the IGBT module; taking the direct-current bus voltage, the collector current, the gate voltage and the switching frequency of the IGBT module as input data of a GKH-ELM prediction model, and predicting the turn-on or turn-off loss value of the IGBT; and (3) taking the turn-on or turn-off loss data obtained by the IGBT dynamic characteristic test as a reference value of the GKH-ELM predicted IGBT turn-on or turn-off loss value, and calculating the fitness value of the IGBT turn-on or turn-off loss value by using a formula (4), namely calculating the predicted performance evaluation value of the extreme learning machine at the optimal sweet spot krill position.
Displaying a comparison graph of the predicted turn-on and turn-off loss values and actual turn-on and turn-off loss values of the IGBT in the step eight, a root mean square error RMSE and a decision coefficient R of the prediction of the turn-on and turn-off loss values of the IGBT test data on a display screen of a computer by using MATLAB software2。
The modeling method for predicting the switching loss in the dynamic process of the IGBT comprises the following steps of: the dc bus voltage, collector current, gate voltage, switching frequency and turn-on and turn-off losses of an IGBT module are well known to those skilled in the art.
In the modeling method for predicting the switching loss in the dynamic process of the IGBT, the good point set algorithm, the krill group algorithm, the extreme learning machine, the cosine control factor and the Levy flight formula are the prior art and are well known to those skilled in the art.
In the modeling method for predicting the switching loss in the dynamic process of the IGBT, the input method for inputting the parameter data obtained by the dynamic characteristic test of the IGBT into the computer is a known method, and the computer, the display and MATLAB computer software are all obtained by commerce.
The invention has the beneficial effects that: compared with the prior art, the invention has the following characteristics,
(1) according to the method, the GKH-ELM prediction model constructed by the extreme learning machine is optimized based on the good krill group algorithm, the turn-on and turn-off losses of the IGBT are predicted, the high-precision and high-reliability prediction of the turn-on and turn-off losses of the IGBT is realized, an engineer is facilitated to improve a heat dissipation system of an IGBT module and the like, and the operation reliability of the IGBT is improved;
(2) according to the IGBT switching loss prediction model based on the optimal krill swarm optimization algorithm, the krill swarm optimization initialized by the optimal set algorithm is used as a carrier, global search is conducted on the optimal weight and the threshold of the extreme learning machine, and meanwhile, a cosine control factor and a Levin flight formula are introduced to achieve dynamic search on the optimal solution;
(3) the modeling method for predicting the switching loss in the dynamic process of the IGBT is strong in expansion capability and compatibility, can expand factors influencing the switching loss of the IGBT, and can also be expanded to other fields such as relay service life prediction and photovoltaic power generation prediction.
Drawings
The invention is further illustrated with reference to the following figures and examples.
FIG. 1 is a schematic flow diagram of the process of the present invention.
Fig. 2 is a model block diagram of switching loss prediction in an IGBT dynamic process.
Fig. 3 is a graph showing the reliability prediction of the turn-on loss of the IGBT according to the present invention.
Fig. 4 is a graph showing the turn-off loss reliability prediction of the IGBT according to the present invention.
Detailed Description
Fig. 1 shows a flow of the modeling method for predicting switching loss in an IGBT dynamic process according to the present invention, which includes starting → obtaining test data of IGBT dynamic characteristics → allocating, normalizing test data → setting, initializing parameters of an extreme learning machine → setting basic parameters of a krill group algorithm → initializing initial values of the krill group by using a sweet spot set algorithm → calculating fitness values of the sweet spot krill group, and recording position data of current optimal and historical sweet spot krill, updating positions of the sweet spot krill by using an improved position formula → judging whether the number of iterations of the current sweet spot krill reaches the maximum number of iterations, and if not, continuing calculating the fitness value of the sweet spot krill group and updating the position of the sweet spot krill; if yes, the next step is executed → the IGBT switching loss prediction result is output → the end is reached.
Fig. 2 shows a model block diagram of switching loss prediction in the IGBT dynamic process. The model takes direct-current bus voltage, collector current, gate voltage and switching frequency of an IGBT module as input, IGBT turn-on and turn-off losses as output, an extreme learning machine as a main body, a krill group algorithm initialized by a sweet spot set algorithm as a carrier, and a cosine control factor and a Levin flight formula as wings, and establishes an IGBT switching loss prediction model constructed by optimizing the extreme learning machine based on the sweet spot krill group algorithm.
Examples
The invention adopts a PC as a platform to build a model, wherein the CPU is i5-3230M 2.60GHz, the installation memory is 4GB, the operating system is Windows 7-64 bits, and MATLAB R2016a version is used. The IGBT module selects MMG75S120B6HN of Macmic company, the rated value of the module is 1200V/75A, the module comprises two identical IGBT chips and a freewheeling diode, and the distance between the IGBT chips and the FWD chips is 6.4 mm.
Step one, obtaining IGBT dynamic characteristic test data
(1.1) acquiring 240 groups of test data through an IGBT dynamic characteristic test, wherein each group of data comprises direct current bus voltage, collector current, gate voltage and switching frequency data, and turn-on and turn-off loss data of an IGBT module;
step two, carrying out normalization processing and distribution on the IGBT dynamic characteristic test data
(2.1) carrying out normalization processing on IGBT characteristic test data by using a formula (1);
(2.2) dividing the normalized test data into learning data and test data, wherein the distribution proportion is as follows:
the number of learning data and the number of testing data are 8: 2;
therefore, 240 sets of dynamic characteristic test data of the IGBT were divided into 192 sets of study data, and 48 sets of test data; each group of data comprises six variables, turn-on and turn-off losses of the IGBT module, and direct-current bus voltage, collector current, gate voltage and switching frequency of the IGBT module;
step three, setting and initializing extreme learning machine parameters
Setting the number num _ in of input layer nodes of the extreme learning machine to be 4;
setting the number num _ hid of hidden layer nodes of the extreme learning machine to be 7;
setting the number num _ out of the output layer nodes of the extreme learning machine to be 1;
step four, setting basic parameters of the phosphoshrimp swarm algorithm, and completing the weight and threshold value distribution from the input layer of the extreme learning machine to the nodes of the hidden layer
Setting the number n of krill in the krill group to be 50;
setting the maximum Iteration number Iteration _ max of the optimization of the phopenaeus ensis group as 100;
setting the dimension dim to num _ in × num _ hid + num _ hid to 4 × 7+7 to 35 of the data contained in each krill;
and initializing krill group position data to serve as weights and thresholds from the input layer of the extreme learning machine to the hidden layer node. The method comprises the following steps: using MATLAB software, using a method of rounding to generate a set X containing 50 points in a unit space of 35 dimensions50(i) The structural formula is shown in formula (2). When constructing the best point matrix X by using the formula (2), q and X are determinedijUpper and lower boundary U ofb、LbQ is 73, since q is equal to or more than 2 × 35+3 and q is the smallest prime number in the range; u shapeb、LbRespectively assigned with 1 and-1; and further obtaining a best point matrix X which is used as a first generation population position data matrix of the phosphorus shrimp groups and is marked as the best point phosphorus shrimp groups.
And distributing 35-dimensional data of each euphausia superba based on the number of nodes of the input layer and the hidden layer of the extreme learning machine, taking the [1, 28] dimensional data of the euphausia superba as the weight from the input layer of the extreme learning machine to the hidden layer, and taking the [29, 35] dimensional data of the euphausia superba as the threshold from the input layer of the extreme learning machine to the hidden layer, so that the parameter setting of the extreme learning machine is completely finished.
Fifthly, optimizing the extreme learning machine by using the optimal krill swarm algorithm, and calculating the fitness value of each optimal krill
And (5.1) establishing a prediction model based on the krill goodpasture algorithm-extreme learning machine, namely a GKH-ELM prediction model, and predicting the turn-on or turn-off loss of the IGBT. The prediction process is as follows: taking data of each krill in the krill group algorithm as a set of weight and threshold from an input layer node to a hidden layer node of the GKH-ELM prediction model; importing 192 omics learning data into an extreme learning machine, wherein the learning data comprises: input data and output data. A set of input data comprising: and the direct-current bus voltage, the collector current, the gate voltage and the switching frequency of the IGBT module output data which are the turn-on loss or the turn-off loss of the IGBT module. After the learning data is imported into the extreme learning machine, the turn-on loss or turn-off loss values of the 50 groups of 192 IGBTs are obtained, and the fitness value of each krill is calculated through formula (4), namely the performance evaluation value of the extreme learning machine corresponding to each krill.
And (5.2) screening the euphausia superba with high fitness, and recording and retaining the position data of the euphausia superba with the highest fitness value in the euphausia superba group algorithm. Screening of krill with high fitness comprises: screening current best krill with the best fitness, comparing the fitness values of the current best krill with the fitness values of historical best krill, and selecting the best. That is, if
Recording the position data of the t movement of the euphausia superba i; if it is
The position data of the current generation of the best-point krill is recorded.
And step six, updating the position of the euphausia superba by using the improved euphausia superba group position updating formula (13).
And seventhly, judging whether the current movement times t of the superbus krill groups reach the maximum iteration times of 100 times set by the algorithm of the superbus krill groups. If t is less than 100, returning to the step five; if t is 100, step eight is executed.
Step eight, outputting the prediction results of the turn-on loss and the turn-off loss of the IGBT
Taking the historical optimal krill position data obtained in the fifth step as the optimal weight and threshold of the GKH-ELM prediction model; using 48 groups of data in 240 groups of dynamic test characteristic test data as test data, wherein each group of test data comprises: the turn-on or turn-off loss of the IGBT module, and the direct current bus voltage, the collector current, the gate voltage and the switching frequency of the IGBT module; taking the direct-current bus voltage, the collector current, the gate voltage and the switching frequency of the IGBT module as input data of a GKH-ELM prediction model, and predicting the turn-on or turn-off loss value of the IGBT; and (3) taking the turn-on or turn-off loss data obtained by the IGBT dynamic characteristic test as a reference value of the GKH-ELM predicted IGBT turn-on or turn-off loss value, and calculating the fitness value of the IGBT turn-on or turn-off loss value by using a formula (4), namely calculating the predicted performance evaluation value of the extreme learning machine under the optimal sweet spot krill.
Displaying a comparison graph of the IGBT turn-on and turn-off loss values predicted in the step eight and actual turn-on and turn-off loss values on a display screen of a computer by using MATLAB software, wherein the predicted root mean square error RMSE (root mean square error) of the turn-on loss of the test data of the IGBT is 0.3772, and the decision coefficient R is20.9984; root mean square error RMSE of turn-off loss prediction 0.0368, and determination coefficient R2＝0.9991。
TABLE 1 comparison of prediction of on-state loss using the method of the present invention with the prior art method
TABLE 2 comparison of prediction of on-state loss using the method of the present invention and the prior art method
As can be seen from tables 1 and 2, the GKH-ELM model is either in RMSE or R2The aspect is superior to other algorithms.
In the above embodiments, the IGBT parameters, such as the dc bus voltage, the collector current, the gate voltage, the switching frequency of the IGBT module, and the turn-on and turn-off losses thereof, are well known to those skilled in the art; the sweet spot set algorithm, the krill group algorithm, the extreme learning machine, the cosine control factor and the Levy flight formula are the prior art and are well known to those skilled in the art; the input method of inputting the parameter data obtained by the IGBT dynamic characteristic test into the computer is a well-known method, and the computer, the display, and the MATLAB computer software are all commercially available.
The above is only a preferred embodiment of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art should be considered as the technical solutions and the inventive concepts of the present invention within the technical scope of the present invention.
Claims (2)
1. A modeling method for predicting switching loss in an IGBT dynamic process is characterized in that the method is a method for predicting the IGBT switching loss of an extreme learning machine based on a good krill swarm algorithm, and comprises the following steps:
step one, obtaining IGBT dynamic characteristic test data
(1.1) acquiring 240 groups of test data through an IGBT dynamic characteristic test, wherein each group of data comprises data of direct current bus voltage, collector current, gate voltage and switching frequency of an IGBT module, and data of turn-on and turn-off loss of the IGBT module;
step two, carrying out normalization processing and distribution on the IGBT dynamic characteristic test data
(2.1) carrying out normalization processing on the IGBT dynamic characteristic test data by using a normalization formula (1):
(2.2) dividing the normalized test data into learning data and test data, wherein the distribution proportion is as follows:
the number of learning data and the number of testing data are 8: 2;
therefore, 240 sets of dynamic characteristic test data of the IGBT were divided into 192 sets of study data, and 48 sets of test data; each group comprises six variables, the turn-on and turn-off losses of the IGBT module, and the direct-current bus voltage, the collector current, the gate voltage and the switching frequency of the IGBT module;
step three, setting and initializing extreme learning machine parameters
The parameters of the extreme learning machine need to set the number of nodes of each layer and the weight and threshold of the initialized extreme learning machine, and the method comprises the following steps:
setting the number num _ in of input layer nodes of the extreme learning machine to be 4;
setting the number num _ hid of hidden layer nodes of the extreme learning machine to be 7;
setting the number num _ out of the output layer nodes of the extreme learning machine to be 1;
step four, basic parameters of the krill group algorithm are set, and distribution of weights and thresholds from the input layer of the extreme learning machine to the hidden layer nodes is completed
(4.1) the parameters of the krill group algorithm set:
setting the number n of krill in the krill group to be 50;
setting the maximum Iteration number Iteration _ max of the optimization of the phopenaeus ensis group as 100;
setting the dimension dim to num _ in × num _ hid + num _ hid to 4 × 7+7 to 35 of the data contained in each krill;
initializing the population data of the shrimp groups, and taking the population data as the weight and the threshold value from the input layer of the extreme learning machine to the node of the hidden layer, wherein the method comprises the following steps: using MATLAB software, using a method of rounding to generate a set X containing 50 points in a unit space of 35 dimensions50(i)，X50(i) The structural formula of (2):
X50(i)＝{[(r1×i)，(r2×i)，...，(rj×i)，...，(r35×i)]，1≤i≤50，1≤j≤35} (2)
in the formula
And (r)jX i) is rjA decimal part of × i, q is the smallest prime number satisfying 2 × dim +3 ≦ q, which can be calculated as q 73, and the best point matrix X is obtained by equation (2):
in matrix X, Xij∈[Lb，Ub](1≤i≤50，1≤j≤35)，Ub、LbAre respectively a variable xijUpper part ofThe lower boundary is respectively assigned as 1 and-1; taking the optimal point matrix X as a first generation of the krill group data matrix of the krill group algorithm, and marking as an optimal point krill group; thus, in formula (3), X1，X2，...，Xi，...，X50Represents the 1 st, 2 nd, 3 rd.., i., 50 th euphausia superba in the euphausia superba group, xijData representing the ith krill in the jth dimension; therefore, the initialization of the weight and the threshold from the input layer of the extreme learning machine to the node of the hidden layer is finished;
(4.2) in consideration of the number of nodes of an input layer and a hidden layer of the extreme learning machine, distributing 35-dimensional data of each superbus, taking [1, 28] dimensional data of the superbus as a weight value from the input layer to the hidden layer of the extreme learning machine, taking [28, 35] dimensional data of the superbus as a threshold value from the input layer to the hidden layer of the extreme learning machine, and introducing the distributed data into the extreme learning machine; at this point, the parameter setting of the extreme learning machine is completely finished;
fifthly, optimizing the extreme learning machine by using the optimal krill group algorithm, and calculating the fitness value of each krill
(5.1) establishing a prediction model of the krill goodpoint algorithm-extreme learning machine, namely a GKH-ELM prediction model, and predicting the turn-on loss or turn-off loss of the IGBT; taking data of each krill in the krill group algorithm as a set of weight and threshold from an input layer node to a hidden layer node of the GKH-ELM prediction model; and (3) importing 192 omics learning data obtained in the step (2.2) into an extreme learning machine, wherein the learning data is divided into two parts: input data and output data; the input data comprises 4 variables which are respectively the direct current bus voltage, the collector current, the gate voltage and the switching frequency of the IGBT module, and the output data is the turn-on loss or the turn-off loss of the IGBT module; after the learning data are imported into the extreme learning machine, 50 groups of predicted values of the turn-on loss or the turn-off loss of 192 IGBTs in each group are obtained, the fitness value of each krill is calculated through a formula (4), namely the predicted performance evaluation value of the extreme learning machine corresponding to each krill,
in the formula (4), F is the fitness value of each euphausia superba, and the larger the fitness value is, the better the fitness value is; p is a radical ofkTaking the data obtained by the dynamic characteristic test of the IGBT as the actual value of the turn-on loss or turn-off loss of the IGBT; p'kThe predicted value of the turn-on loss or turn-off loss of the IGBT is obtained; n is the predicted output number of the turn-on loss or the turn-off loss of the extreme learning machine, namely N is 192 during the learning period; similarly, during the test, N-48;
(5.2) screening the euphausia superba with high fitness, and recording and retaining the position data of the euphausia superba with the maximum fitness value in an euphausia superba group algorithm; screening of krill with high fitness comprises: screening current generation of superb krill with optimal fitness, and comparing the fitness values of current generation of superb krill with historical optimal superb krill to select optimal, i.e., if
Recording the position data of the t movement of the euphausia superba i; if it is
Recording the position data of the current optimal sweet krill and recording as historical optimal sweet krill;
step six, improving the krill goodpoint algorithm and searching a global optimal solution
(6.1) updating the superbus krill group X obtained in the step four or the formula (13); location updating exercise of individual euphausia superba, mainly comprising induced exercise IiForaging movement FiAnd physical diffusion movement PiThe formulas are respectively shown in (5), (6) and (7):
in the formula (5), the first and second groups,
represents the t-th induced movement of the ith euphausia superba, and the same principle is adopted
Represents the t-1 induced movement of the ith euphausia superba; i ismaxExpressing the maximum induction rate, take Imax＝0.01；
Shows the effect of the t-th time of the peripheral euphausia superba on the ith euphausia superba
Representing the induction effect of the historical optimal krill on the krill i; omegaIInertial weight to induce motion, range is [0, 1 ]]；
In the formula (6), Fi tRepresents the t-th foraging movement of the ith euphausia superba, and has the same principle Fi t-1Represents the t-1 foraging movement of the ith euphausia superba;
indicating the induction intensity of the current food to the ith krill sweet spot,
representing the foraging traction of the historical optimal krill on the movement of the krill i; v. ofFExpressing the foraging speed of krill, vF＝0.02；ωFThe weight for induction of foraging motion is in the range of [0, 1%]；
In the formula (7), Pi tRepresents the t-th physical diffusion movement of the ith euphausia superba; pmaxRepresenting the maximum rate of diffusion of the krill at the sweet spot, taking Pmax0.005; t represents the current number of movements; deltaiIs the direction vector of the random diffusion of the current sweet krill, and is deltai∈[0，1]；
Therefore, the amount of change in the positional movement of the euphausia superba i at the t-th movement
See formula (8) and the position after the end of the tth movement see formula (9):
recording the position of the krill i after the end of the tth movement
As the initial position of the t +1 th motion;
(6.2) introducing a cosine control factor and a position updating formula of the Levy flight improvement Kyoho lobster colony, wherein the cosine control factor is used for improving omega in the Kyoho krill induced motion formula (5) and the foraging motion formula (6)IAnd ωFThe formula is improved as shown in formula (10);
the Lei-dimensional flight is used for improving the formula (9), and is set, if the optimal fitness value of the krill group exceeds 20 iterations and is not changed, the Lei-dimensional flight is started until the optimal fitness value of the krill group is changed, and the Lei-dimensional flight formula is shown in the formula (11):
wherein u and epsilon both follow a standard normal distribution, and β is 1.5, and the calculation formula of phi is shown in formula (12);
therefore, the modified superba location obtained by combining equations (9), (10), (11), and (12) is updated to equation (13), and thus the location of each individual in the superba group X is updated
Updating the position according to a formula (13) to generate a new position;
step seven, judging whether the current movement times t of the krill groups reach the maximum iteration times set by the krill group algorithm for 100 times; if t is less than 100, returning to the step five; if t is 100, executing step eight;
step eight, outputting the prediction results of the turn-on loss and the turn-off loss of the IGBT
Taking the historical optimal krill position data obtained in the fifth step as the optimal weight and threshold of the GKH-ELM prediction model; using 48 data of 240 sets of dynamic test characteristic test data as test data, wherein each set of test data comprises: the turn-on or turn-off loss of the IGBT module, and the direct current bus voltage, the collector current, the gate voltage and the switching frequency of the IGBT module; taking the direct-current bus voltage, the collector current, the gate voltage and the switching frequency of the IGBT module as input data of a GKH-ELM prediction model, and predicting the turn-on or turn-off loss value of the IGBT; taking the turn-on or turn-off loss data obtained by the IGBT dynamic characteristic test as a reference value of the GKH-ELM prediction IGBT turn-on or turn-off loss value, and calculating the fitness value of the IGBT turn-on or turn-off loss value by using a formula (4), namely calculating the prediction performance evaluation value of the extreme learning machine at the optimal sweet spot krill position;
displaying a comparison graph of the predicted turn-on and turn-off loss values and actual turn-on and turn-off loss values of the IGBT in the step eight, a root mean square error RMSE and a decision coefficient R of the prediction of the turn-on and turn-off loss values of the IGBT test data on a display screen of a computer by using MATLAB software2。
2. The modeling method for predicting switching loss in an IGBT dynamic process according to claim 1, characterized in that: variables affecting the switching loss of the IGBT are fully considered, namely the direct-current bus voltage, the collector current, the gate voltage and the switching frequency of the IGBT module.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110477808.9A CN113162375B (en) | 2021-04-30 | 2021-04-30 | Modeling method for switch loss prediction in IGBT dynamic process |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110477808.9A CN113162375B (en) | 2021-04-30 | 2021-04-30 | Modeling method for switch loss prediction in IGBT dynamic process |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113162375A CN113162375A (en) | 2021-07-23 |
| CN113162375B true CN113162375B (en) | 2022-04-12 |
Family
ID=76872718
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110477808.9A Expired - Fee Related CN113162375B (en) | 2021-04-30 | 2021-04-30 | Modeling method for switch loss prediction in IGBT dynamic process |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113162375B (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107368858A (en) * | 2017-07-28 | 2017-11-21 | 中南大学 | A kind of parametrization measurement multi-model intelligent method for fusing of intelligent environment carrying robot identification floor |
| CN109101738A (en) * | 2018-08-24 | 2018-12-28 | 河北工业大学 | A kind of IGBT module degree of aging appraisal procedure |
| CN109918720A (en) * | 2019-01-23 | 2019-06-21 | 广西大学 | Diagnosis Method of Transformer Faults based on krill group's Support Vector Machines Optimized |
-
2021
- 2021-04-30 CN CN202110477808.9A patent/CN113162375B/en not_active Expired - Fee Related
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107368858A (en) * | 2017-07-28 | 2017-11-21 | 中南大学 | A kind of parametrization measurement multi-model intelligent method for fusing of intelligent environment carrying robot identification floor |
| CN109101738A (en) * | 2018-08-24 | 2018-12-28 | 河北工业大学 | A kind of IGBT module degree of aging appraisal procedure |
| CN109918720A (en) * | 2019-01-23 | 2019-06-21 | 广西大学 | Diagnosis Method of Transformer Faults based on krill group's Support Vector Machines Optimized |
Non-Patent Citations (2)
| Title |
|---|
| "Combination of krill herd algorithm with chaos theory in global optimization problems";Leila Gharavian et al.;《2013 3rd Joint Conference of AI & Robotics and 5th RoboCup Iran Open International Symposium》;20130408;全文 * |
| "鳞虾群算法的改进及其在结构可靠性分析中的应用";程立翔;《中国优秀博硕士学位论文全文数据库(硕士)基础科学辑》;20210215;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113162375A (en) | 2021-07-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113657064B (en) | Multi-physical-field joint simulation method for power semiconductor module | |
| CN112001127B (en) | IGBT junction temperature prediction method | |
| CN110147559B (en) | Converter multidisciplinary optimization design method based on multi-physics coupling | |
| US11913986B2 (en) | Reliability evaluation method and system of microgrid inverter IGBT based on segmented LSTM | |
| CN104779830B (en) | A kind of variable inversion controlling method of dead time | |
| CN1731645A (en) | Voltage control method based on soft partitions in power system | |
| CN110765695B (en) | Simulation calculation method for obtaining crack propagation path of concrete gravity dam based on high-order finite element method | |
| CN110138000B (en) | Hybrid direct-current power transmission system control parameter optimization method based on robust multi-objective optimization algorithm | |
| CN103577896B (en) | A kind of region partitioning method of large scale electric network adaptive setting | |
| CN112329244B (en) | Optimal power loss equivalent modeling method for IGBT junction temperature estimation | |
| CN103036467A (en) | Multi-objective Pareto optimal control method of inverter based on immune algorithm | |
| CN115629553A (en) | High-spatial-temporal-resolution IGBT module electric heating stress calculation method | |
| CN115048804A (en) | Urban pipe network multi-sensor optimized deployment method | |
| CN115270698A (en) | Chip global automatic layout method based on deep reinforcement learning | |
| CN113162375B (en) | Modeling method for switch loss prediction in IGBT dynamic process | |
| CN115912367A (en) | Intelligent generation method for operation mode of power system based on deep reinforcement learning | |
| CN111244944A (en) | AC/DC power distribution network reliability assessment method based on influence increment | |
| CN113468847A (en) | Integrated circuit global layout method based on non-integer multiple line height unit | |
| CN103714211B (en) | Integrated circuit layouts method based on Move Mode sequence Yu multi-agent particle swarm | |
| CN106407492A (en) | Micro-grid real-time simulation method for processing multi-break switches of power systems | |
| CN117269838A (en) | Method and system for determining short-circuit current of network-structured power electronic equipment | |
| CN108121875A (en) | A kind of Topology Optimization Method that platform is solved based on Matlab and CFX joints | |
| CN113344426B (en) | Energy storage station planning method, device and equipment | |
| Tian et al. | Improving simulated annealing algorithm for FPGA placement based on reinforcement learning | |
| CN108197695A (en) | A kind of individualized learning method for optimizing route based on lightning searching algorithm |
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 | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20220412 |