CN110008525A - Automobile form characteristic crossover Evolution Forecasting method based on INGBM (1,1) - Google Patents
Automobile form characteristic crossover Evolution Forecasting method based on INGBM (1,1) Download PDFInfo
- Publication number
- CN110008525A CN110008525A CN201910187609.7A CN201910187609A CN110008525A CN 110008525 A CN110008525 A CN 110008525A CN 201910187609 A CN201910187609 A CN 201910187609A CN 110008525 A CN110008525 A CN 110008525A
- Authority
- CN
- China
- Prior art keywords
- vehicle
- parameter
- model
- evolution
- ingbm
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/10—Geometric CAD
- G06F30/15—Vehicle, aircraft or watercraft design
-
- 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
-
- 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
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T90/00—Enabling technologies or technologies with a potential or indirect contribution to GHG emissions mitigation
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Geometry (AREA)
- General Physics & Mathematics (AREA)
- Biophysics (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Computer Hardware Design (AREA)
- Molecular Biology (AREA)
- Automation & Control Theory (AREA)
- Data Mining & Analysis (AREA)
- Computing Systems (AREA)
- Computational Linguistics (AREA)
- Biomedical Technology (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- General Health & Medical Sciences (AREA)
- Aviation & Aerospace Engineering (AREA)
- Genetics & Genomics (AREA)
- Computational Mathematics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Physiology (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The present invention provides be based on INGBM (1,1) automobile form characteristic crossover Evolution Forecasting method, the following steps are included: step A: the influence of cross-platform vehicle or concept vehicle being taken into account, they, which reconfigure and construct with vehicle on longitudinal evolution axis, intersects evolution axis;Step B: influence vehicle is handled with the data for being affected vehicle using form homogenizing method, obtains the initial ordered series of numbers of new changing features;Step C: it proposes to improve Nonlinear Grey Bernoulli Jacob model INGBM (1,1), in traditional NGBM (1,1) system delay parameter and time effect parameter are added in model, and parameter is optimized using Genetic Algorithms, make its be more suitable oscillation type number it was predicted that the present invention can solve initial number train wave it is dynamic big and the problem of influence precision of prediction, the prediction effect of concussion data can be substantially improved.
Description
Technical field
The present invention relates to prediction technique technical fields, and specially the automobile form characteristic crossover based on INGBM (1,1) is evolved
Prediction technique.
Background technique
The prediction of automobile form trend of evolution is always the research hotspot of Automobile Design.Mature automobile brand can its product more
New regenerate retains certain Morphological continuity in the process, promotes brand recognition with this, therefore study automobile form trend of evolution not
Only design information can be provided for brand itself, design reference can also be provided for other brands.Determine based on evolution algorithm
Amount analysis.Qualitative analysis such as uses shape grammar research product and automobile form design, or establishes product feature variation and disappear
Mapping relations between the person's of expense image carry out qualitative forecasting.But compared with qualitative analysis, quantitative analysis can provide more for designer
Specific design information.
With the globalization of automobile manufacture industry, depot is dedicated to forcing down research and development cost and avoiding market tightening, each cart
Factory reduces cost by component sharing strategy one after another.Many new passenger car design elements are used on concept car first, are borrowed
This sounds out public acceptance level, reapplies in the design of volume production vehicle.Many riding brands are during Morphological continuity, same to vehicle
Can have both at the same time with the continuity of standdle carrier type feature, continue with vehicle and refer to that feature of the brand in a vehicle platform continues, and
The reference of cross-platform vehicle or the borrow of concept vehicle characteristic feature is added in the continuity of standdle carrier type on the basis of the former, referred to as hands over
Fork is evolved.Therefore the vehicle on longitudinal evolution axis also will receive the influence of concept vehicle.Therefore to passenger car morphological feature carry out across
Vehicle is intersected Evolution Forecasting and is had great theoretical and practical significance to the construction of product system, passenger car family.
Summary of the invention
The purpose of the present invention is to provide the automobile form characteristic crossover Evolution Forecasting methods for being based on INGBM (1,1), with solution
Certainly the problems mentioned above in the background art.
To achieve the above object, the invention provides the following technical scheme: the automobile form feature based on INGBM (1,1) is handed over
Pitch Evolution Forecasting method, comprising the following steps:
Step A: the influence of cross-platform vehicle or concept vehicle is taken into account, and vehicle is again on they and longitudinal evolution axis
Group, which merges to construct, intersects evolution axis;
Step B: influence vehicle is handled with the data for being affected vehicle using form homogenizing method, obtains new feature
Change initial ordered series of numbers;
Step C: it proposes to improve Nonlinear Grey Bernoulli Jacob model INGBM (1,1), in traditional NGBM (1,1) model
System delay parameter and time effect parameter is added, and parameter is optimized using Genetic Algorithms, it is made to be more suitable shake
Swing the prediction of type data.
Further, the form is homogenized method are as follows:
Assuming that evolving in certain vehicle to intersect mode of evolution, the three-dimensional coordinate of each characteristic point is expressed as pij=(xij, yij,
zij), characteristic point variation initial number is classified as Xi={ xi1, xi2..., xik, Yi={ yi1, yi2..., yik, Zi={ zi1, zi2...,
zik, in formula: i=1,2,3 ..., m, j=1,2,3 ..., k, m are characterized a number, and k is passenger car evolutionary generation;
Cross-platform vehicle characteristic point coordinate is denoted asIt on longitudinal evolution axis certain for vehicle generate
It influences;
In formula: d is to be affected vehicle and cross-platform upper influence vehicle weight ratio, obtained p ' on longitudinal evolution axisij=
(x′ij, y 'ij, z 'ij) p before substitutionij=(xij, yij, zij)。
Further, steps are as follows for the improvement of improvement Nonlinear Grey Bernoulli Jacob's model INGBM (1,1):
(1) X is set(0), X(1)And Z(0)Respectively original series, single order Accumulating generation sequence and background value sequence establish ash
The color differential equation is as follows:
x(0)(k)+a tan p(k-r)z(1)(k-r)=b sin p (k-r) (z(1)(k-r))u (2)
This is improved model INGBM (1,1), wherein-a tan p (k-r) is development coefficient, b sin p (k-r) is grey
Actuating quantity, r are system delay time, and p is time effect parameter, μ ≠ 1;
(2) least-squares estimation is made to parameter a and b:
(a, b)T=(BTB)-1BTY (3)
In formula:
BTY=[- E D]T,
(3) model predication value are as follows:
Index μ and time effect parameter p is unknown in formula, and system delay time r can take r=1 respectively, 2 ... and substitutes into mould
Type, more each model accuracy, and precision the higher person is taken, therefore the determination of μ and p parameter can be converted into minimize the error as mesh
Target Non-linear Optimal Model Solve problems are minimised as optimization aim with average relative error MAPE, the relationship between parameter
For constraint condition, μ and p are unknown variable, are solved by Genetic Algorithms;
Optimized model are as follows:
In formula:
Further, the operation of the Genetic Algorithms mainly includes three parts: selection intersects and makes a variation, Optimization Steps
It is as follows:
S1: μ and p are set as parameter to be optimized;
S2: encoding parameter, and initial population is randomly generated;
S3: using the objective function in formula (5) as the fitness function of algorithm;
S4: crossover operation is completed in parent population, and makes a variation and generates filial generation;
S5: being iteratively repeated S3 and S4, stops until individual optimization aim MAPE≤0.01. is operated;
By using GA algorithm, optimized parameter p is obtainedbestAnd μbest, thus obtain abestAnd bbest, therefore have optimal mould
Analog values are as follows:
Compared with prior art, the beneficial effects of the present invention are:
The present invention takes into account the influence of cross-platform vehicle or concept vehicle, and vehicle is again on they and longitudinal evolution axis
Group, which merges to construct, intersects evolution axis.Influence vehicle is handled with the data for being affected vehicle using form homogenizing method, is obtained
To the new initial ordered series of numbers of changing features.To solve the problems, such as that initial number train wave is dynamic big and influences precision of prediction, propose to improve non-thread
Property grey Bernoulli Jacob's model INGBM (1,1), system delay parameter and time effect are added in traditional NGBM (1,1) model
Parameter, and using the Solve problems of genetic algorithm solution optimized parameter, so that it is more suitable oscillation type number it was predicted that can be significantly
Improve the prediction effect of concussion data.
Detailed description of the invention
Fig. 1 is that Accord of the present invention intersects evolution axis schematic diagram;
Fig. 2 is INGBM (1,1) predicted value and initial value comparison diagram under the conditions of different parameters of the present invention;
Fig. 3 is the tenth generation of Accord headlight schematic diagram of the present invention.
Specific embodiment
In order to make the objectives, technical solutions, and advantages of the present invention clearer, with reference to the accompanying drawings and embodiments, right
The present invention is described in further detail.The specific embodiments are only for explaining the present invention technical solution described herein, and
It is not limited to the present invention.
The present invention provides a kind of technical solution: it is based on the automobile form characteristic crossover Evolution Forecasting method of INGBM (1,1),
The following steps are included:
Step A: the influence of cross-platform vehicle or concept vehicle is taken into account, and vehicle is again on they and longitudinal evolution axis
Group, which merges to construct, intersects evolution axis;
Step B: influence vehicle is handled with the data for being affected vehicle using form homogenizing method, obtains new feature
Change initial ordered series of numbers;
Step C: it proposes to improve Nonlinear Grey Bernoulli Jacob model INGBM (1,1), in traditional NGBM (1,1) model
System delay parameter and time effect parameter is added, and parameter is optimized using Genetic Algorithms, it is made to be more suitable shake
Swing the prediction of type data.
Further, the form is homogenized method are as follows:
Assuming that evolving in certain vehicle to intersect mode of evolution, the three-dimensional coordinate of each characteristic point is expressed as pij=(xij, yij,
zij), characteristic point variation initial number is classified as Xi={ xi1, xi2..., xik, Yi={ yi1, yi2..., yik, Zi={ zi1, zi2...,
zik, in formula: i=1,2,3 ..., m, j=1,2,3 ..., k, m are characterized a number, and k is passenger car evolutionary generation;
Cross-platform vehicle characteristic point coordinate is denoted asIt on longitudinal evolution axis certain for vehicle generate
It influences;
In formula: d is to be affected vehicle and cross-platform upper influence vehicle weight ratio, obtained p ' on longitudinal evolution axisij=
(x′ij, y 'ij, z 'ij) p before substitutionij=(xij, yij, zij)。
Traditional Nonlinear Grey Bernoulli Jacob model NGBM (1,1) is the novel grey forecasting model set up in recent years, it
It is the extension of Traditional GM (1,1) model, its feature is to joined index μ in Traditional GM (1,1) model, which can have
Effect embodies the estimated performance of gray system, and substantially increases the flexibility of model form.
Traditional Nonlinear Grey Bernoulli Jacob model solution process is as follows:
(1) original series X is set(0)={ x(0)(1), x(0)(2) ..., x(0)(m) }, whereinK=
1,2 ..., m.
Separately there is Z(1)={ z(1)(1), z(1)(2) ..., z(1)It (m) }, is sequence X(1)Adjacent average generation sequence, wherein z(1)
(k)=0.5x(1)(k)+0.5x(1)(k-1)。
(2) Grey Differential Equation is established:
x(0)(k)+az(1)(k)=b (z(1)(k))μ, μ ≠ 1, k=2,3 ..., m (7)
This is tradition NGBM (1,1) model, wherein x(0)It (k) is grey derivative, z(1)It (k) is the background value of grey derivative, b is ash
Color actuating quantity.As μ=0, which is Traditional GM (1,1) model, and as μ=2, which is then grey Verhulst mould
Type.
(3) least-squares estimation is made to parameter a and b:
(a, b)T=(BTB)-1BTY (8)
Wherein
(4) albinism differential equation of tradition NGBM (1,1) model is established:
(5) albefaction time response formula is obtained:
(6) rightThe reduction of single order regressive is done, is obtained:
Index μ is added in traditional NGBM (1,1) model in Traditional GM (1,1) model, improves NGBM (1,1) model form
Flexibility, different μ values is conducive to NGBM (1,1) model and adapts to different Nonlinear Time Series.But NGBM (1,1) mould
Type is only applicable to the time series that unimodal variation or growth are obstructed, and is not suitable for oscillation type sequence, the variation of passenger car morphological feature
Sequence data is mostly the great small sample ordered series of numbers of concussion property, it is therefore desirable to be improved to traditional NGBM (1,1).
Further, steps are as follows for the improvement of improvement Nonlinear Grey Bernoulli Jacob's model INGBM (1,1):
(1) X is set(0), X(1)And Z(0)Respectively original series, single order Accumulating generation sequence and background value sequence establish ash
The color differential equation is as follows:
x(0)(k)+a tan p(k-r)z(1)(k-r)=b sin p (k-r) (z(1)(k-r))u (2)
This is improved model INGBM (1,1), wherein-a tan p (k-r) is development coefficient, b sin p (k-r) is grey
Actuating quantity, r are system delay time, and p is time effect parameter, μ ≠ 1;
(2) least-squares estimation is made to parameter a and b:
(a, b)T=(BTB)-1BTY (3)
In formula:
(3) model predication value are as follows:
Index μ and time effect parameter p is unknown in formula, and system delay time r can take r=1 respectively, 2 ... and substitutes into mould
Type, more each model accuracy, and precision the higher person is taken, therefore the determination of μ and p parameter can be converted into minimize the error as mesh
Target Non-linear Optimal Model Solve problems are minimised as optimization aim with average relative error MAPE, the relationship between parameter
For constraint condition, μ and p are unknown variable, are solved by Genetic Algorithms;
Optimized model are as follows:
In formula:
It is theoretical in conjunction with Munsell hereditary variation on the basis of Darwinian theory of biological evolution " survival of the fittest, the survival of the fittest ",
Its filial generation is the ordered arrangement of parent gene on chromosome, and genetic algorithm simulation Biologic evolution process is calculated.
It is embodied in that ability of searching optimum is strong, strong robustness, can carry out multivalue comparison the advantages of genetic algorithm;It is changed using probability mechanism
Generation, randomness are strong;Big with other algorithm combination spaces, scalability is strong.
Further, the operation of the Genetic Algorithms mainly includes three parts: selection intersects and makes a variation, Optimization Steps
It is as follows:
S1: μ and p are set as parameter to be optimized;
S2: encoding parameter, and initial population is randomly generated;
S3: using the objective function in formula (5) as the fitness function of algorithm;
S4: crossover operation is completed in parent population, and makes a variation and generates filial generation;
S5: being iteratively repeated S3 and S4, and until individual optimization aim MAPE≤0.01, operation stops;
By using GA algorithm, optimized parameter P is obtainedbestAnd μbest, thus obtain abestAnd bbest, therefore have optimal mould
Analog values are as follows:
Compared with traditional NGBM (1,1) model, improved model INGBM (1,1) increase system delay parameter r, when
Variable element tan p (k-r) and sin p (k-r), improved model INGBM (1,1) pass through the variation of system delay parameter, make
Its development coefficient and ash are measured and are changed over time, therefore for concussion property data, the adaptability of the improved model
Conventional model should be better than.
Honda Accord (Accord) vehicle system has been proposed 9 generation vehicles since 1976 come out altogether so far, due to it is economical,
Economical, durable and practical feature, the deep favorable comment by the countless consumers in various regions.Last century the nineties duration economic depression, should
Layout strategy is positioned as global differentiation by brand, i.e., different same brand vehicles is released in different countries and regions, with full
The market demands such as foot different region, consumer's habit, regulation.With design globalization propulsion, design differentiation gradually by
Design convergentization substitution.Family gene expression characteristics obtain promotion and application in full vehicle system, product family, and low and middle-end vehicle starts to borrow
With brand deluxe carmodel design element.
As shown in Figure 1, when mid-term changes money, face design have passed through a degree of improvement before Accord, wherein upper grid is again
More chromed ornament elements are incorporated when design, the design element that High Tier Brand is sung the praises of under Honda has then been used for reference in both ends headlight moulding
Element, LED fog lamp seem more there is kinesthesia.LED taillight is also transformed simultaneously, the moulding of rear bumper becomes more
It is radical to add.9th generation Accord in 2014 enters market, and preceding face has used for reference the design sung the praises of, and moulding is unique and technology sense is full, allows whole
A headstock is very vibrant.Succinct design also allow visual effect it is resistance to see it is eye-candy.Therefore it is refined as the 9th generation ILX vehicle will to be sung the praises of
The influence vehicle at pavilion is included in Accord intersection evolution axis.
(1) initial data sequence
It by Accord nine generations vehicle and sings the praises of ILX vehicle front view and is put into Alias, carry out the extraction of headlight characteristic curve
And numeralization.Wherein in the 9th generation of Accord, is handled with the data for singing the praises of ILX using form homogenizing method.In view of Accord the 9th
Generation is to be affected vehicle, sings the praises of ILX as influence vehicle, thus this two groups of data are added with 3: 1 weight proportion.
Characteristic point coordinate is described as pij=(xij, yij), every headlight characteristic curve is converted into after nurbs curve form altogether
25 characteristic points are obtained, there is ordered series of numbers are as follows:
Xi={ xI, 1, YI, 2..., xI, 9, xI, ILX, Yi={ yI, 1, yI, 2..., yI, 9, yI, ILX}
In formula: xi1To xi9Indicate the Accord first generation to the 9th generation characteristic point X axis coordinate, xI, ILXILX characteristic point is sung the praises of in expression
X axis coordinate, wherein yI, 1To yI, 9Indicate the Accord first generation to the 9th generation characteristic point Y axis coordinate, yI, ILXILX characteristic point is sung the praises of in expression
Y axis coordinate.
There is x 'I, 9=0.75xI, 9+0.25xI, ILX, y 'I, 9=0.75yI, 9+0.25yI, ILX, x 'I, 9Replace xI, 9, y 'I, 9Replace
yI, 9, with the first eight for coordinate together with form new data sequence, the prediction initiation sequence as INGBM (1,1) later.
1 characteristic point changes in coordinates ordered series of numbers of table
X′i={ xI, 1, xI, 2..., x 'I, 9, Y 'i={ yI, 1, yI, 2..., y 'I, 9, data are as shown in table 1.
(2) process is predicted
With X10For, coordinate sequence indicates the variation of X axis coordinate in the from first to the 9th generation evolutionary process of this o'clock,
X10={ 46.59,84.66,138.67,134.59,94.75,101.72,86.19,72.78,52.11 }.
For the precision of prediction for examining model above, introduces average relative error (ARPE) and be used as validation criteria.
In formula: xiIt is measured value,For predicted value, k is passenger car Morphological evolution algebra.The more low then model accuracy of ARPE more
Office.
According to the different parameters for improving INGBM (1,1), obtained prediction result and ARPE value, as shown in table 2.
2 initial value of table and predicted value
As r=1, pbest=1.3525, μbest=-0.5282, ARPE 3.53%, as shown in Fig. 2 (a);Work as r=2
When, pbest=-1.3171, μbest=-0.5236, ARPE 3.55%, as shown in Fig. 2 (b);As r=3, pbest=
0.0928, μbest=-0.5305, ARPE 4.16%, as shown in Fig. 2 (c);As r=4, pbest=1.3230, μbest=-
0.6777, ARPE 2.93%, as shown in Fig. 2 (d);As r=5, pbest=1.9122, μbest=-0.9451, ARPE is
3.44%, as shown in Fig. 2 (e);As r=6, pbest=0.8504, μbest=0.7571, ARPE 2.58%, such as Fig. 2 (f) institute
Show;As r=7, pbest=0.4174, μbest=0.3958, ARPE 9.63%, as shown in Fig. 2 (g).
As r=6, model obtains minimum ARPE value, therefore the x obtained using INGBM (1,1) model10,10Value is
25.27。
(3) prediction result and analysis
Coordinate data predicted value of 25 headlight characteristic points when evolving to for ten generations is successively obtained, as shown in table 3.It will
Data are placed in Alias, obtain modelling effect, as shown in Figure 3.
Difficult the tenth generation predicted value of pavilion headlight characteristic point of table 3
The influence of cross-platform vehicle or concept vehicle is taken into account, they are reconfigured simultaneously with vehicle on longitudinal evolution axis
It constructs and intersects evolution axis.Influence vehicle is handled with the data for being affected vehicle using form homogenizing method, is obtained new
The initial ordered series of numbers of changing features.It proposes to improve Nonlinear Grey Bernoulli Jacob model INGBM (1,1), in traditional NGBM (1,1) model
Middle addition system delay parameter and time effect parameter, and parameter is optimized using Genetic Algorithms, to solve initial number
Train wave is dynamic big and the problem of influence precision of prediction.Case result proves that INGBM (1, the 1) model intersects in morphological feature and evolves in advance
Validity in survey can also substantially improve the prediction effect of concussion data.This model is in addition to can operate with other car body features
Outside line Study on Evolution, other product design fields are also extended to.
The above only expresses the preferred embodiment of the present invention, and the description thereof is more specific and detailed, but can not be because
This and be interpreted as limitations on the scope of the patent of the present invention.It should be pointed out that for those of ordinary skill in the art,
Under the premise of not departing from present inventive concept, several deformations can also be made, improves and substitutes, these belong to protection of the invention
Range.Therefore, the scope of protection of the patent of the invention shall be subject to the appended claims.
Claims (4)
1. being based on the automobile form characteristic crossover Evolution Forecasting method of INGBM (1,1), it is characterised in that: the following steps are included:
Step A: the influence of cross-platform vehicle or concept vehicle is taken into account, they are reconfigured with vehicle on longitudinal evolution axis
And it constructs and intersects evolution axis;
Step B: influence vehicle is handled with the data for being affected vehicle using form homogenizing method, obtains new changing features
Initial ordered series of numbers;
Step C: it proposes to improve Nonlinear Grey Bernoulli Jacob model INGBM (1,1), be added in traditional NGBM (1,1) model
System delay parameter and time effect parameter, and optimized using Genetic Algorithms to parameter make it be more suitable oscillation type
Data prediction.
2. the automobile form characteristic crossover Evolution Forecasting method according to claim 1 based on INGBM (1,1), feature
Be: the form is homogenized method are as follows:
Assuming that evolving in certain vehicle to intersect mode of evolution, the three-dimensional coordinate of each characteristic point is expressed as pij=(xij, yij, zij),
Characteristic point variation initial number is classified as Xi={ xi1, xi2..., xik, Yi={ yi1, yi2..., yik, Zi={ zi1, zi2..., zik,
In formula: i=1,2,3 ..., m, j=1,2,3 ..., k, m are characterized a number, and k is passenger car evolutionary generation;
Cross-platform vehicle characteristic point coordinate is denoted asIt on longitudinal evolution axis certain have an impact for vehicle;
In formula: d is to be affected vehicle and cross-platform upper influence vehicle weight ratio, obtained p ' on longitudinal evolution axisij=(x 'ij,
y′ij, z 'ij) p before substitutionij=(xij, yij, zij)。
3. the automobile form characteristic crossover Evolution Forecasting method according to claim 1 based on INGBM (1,1), feature
Be: steps are as follows for the improvement of improvement Nonlinear Grey Bernoulli Jacob's model INGBM (1,1):
(1) X is set(0), X(1)And Z(0)Respectively original series, single order Accumulating generation sequence and background value sequence, it is micro- to establish grey
Divide equation as follows:
x(0)(k)+atanp(k-r)z(1)(k-r)=bsinp (k-r) (z(1)(k-r))u (2)
This is improved model INGBM (1,1), wherein-atanp (k-r) is development coefficient, bsinp (k-r) is grey actuating quantity, r
For system delay time, p is time effect parameter, μ ≠ 1;
(2) least-squares estimation is made to parameter a and b:
(a, b)T=(BTB)-1BTY (3)
In formula:
BTY=[- E D]T,
(3) model predication value are as follows:
Index μ and time effect parameter p is unknown in formula, and system delay time r can take r=1 respectively, 2 ... and substitutes into model,
Compare each model accuracy, and take precision the higher person, therefore the determination of μ and p parameter can be converted into minimize the error as target
Non-linear Optimal Model Solve problems, optimization aim is minimised as with average relative error MAPE, the relationship between parameter is
Constraint condition, μ and p are unknown variable, are solved by Genetic Algorithms;
Optimized model are as follows:
In formula:
4. the automobile form characteristic crossover Evolution Forecasting method according to claim 1 based on INGBM (1,1), feature
Be: the operation of the Genetic Algorithms mainly includes three parts: selection intersects and makes a variation, and Optimization Steps are as follows:
S1: μ and P are set as parameter to be optimized;
S2: encoding parameter, and initial population is randomly generated;
S3: using the objective function in formula (5) as the fitness function of algorithm;
S4: crossover operation is completed in parent population, and makes a variation and generates filial generation;
S5: being iteratively repeated S3 and S4, and until individual optimization aim MAPE≤0.01, operation stops;
By using GA algorithm, optimized parameter p is obtainedbestAnd μbest, thus obtain abestAnd bbest, therefore have the optimal analogue value
Are as follows:
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910187609.7A CN110008525A (en) | 2019-03-12 | 2019-03-12 | Automobile form characteristic crossover Evolution Forecasting method based on INGBM (1,1) |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910187609.7A CN110008525A (en) | 2019-03-12 | 2019-03-12 | Automobile form characteristic crossover Evolution Forecasting method based on INGBM (1,1) |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110008525A true CN110008525A (en) | 2019-07-12 |
Family
ID=67166945
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910187609.7A Pending CN110008525A (en) | 2019-03-12 | 2019-03-12 | Automobile form characteristic crossover Evolution Forecasting method based on INGBM (1,1) |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110008525A (en) |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104200272A (en) * | 2014-08-28 | 2014-12-10 | 北京工业大学 | Complex network community mining method based on improved genetic algorithm |
CN104504462A (en) * | 2014-12-11 | 2015-04-08 | 南昌大学 | Automobile form evolution trend prediction method based on improved gray BPNN (BP Neural Network) and Markov chain |
CN105930670A (en) * | 2016-04-29 | 2016-09-07 | 浙江大学 | Model parameter uncertainty-based dynamic prediction method for river emergency pollution accident |
CN107730039A (en) * | 2017-10-10 | 2018-02-23 | 中国南方电网有限责任公司电网技术研究中心 | Method and system for predicting load of power distribution network |
-
2019
- 2019-03-12 CN CN201910187609.7A patent/CN110008525A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104200272A (en) * | 2014-08-28 | 2014-12-10 | 北京工业大学 | Complex network community mining method based on improved genetic algorithm |
CN104504462A (en) * | 2014-12-11 | 2015-04-08 | 南昌大学 | Automobile form evolution trend prediction method based on improved gray BPNN (BP Neural Network) and Markov chain |
CN105930670A (en) * | 2016-04-29 | 2016-09-07 | 浙江大学 | Model parameter uncertainty-based dynamic prediction method for river emergency pollution accident |
CN107730039A (en) * | 2017-10-10 | 2018-02-23 | 中国南方电网有限责任公司电网技术研究中心 | Method and system for predicting load of power distribution network |
Non-Patent Citations (1)
Title |
---|
徐秋莹: "基于智能计算的乘用车外观特征继承与变异研究", 《中国优秀博硕士学位论文全文数据库(博士)工程科技Ⅱ辑》 * |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106447024A (en) | Particle swarm improved algorithm based on chaotic backward learning | |
AlGeddawy et al. | Optimum granularity level of modular product design architecture | |
CN111311077B (en) | Method for solving flexible job shop scheduling by particle swarm algorithm based on longicorn stigma search | |
CN102214321B (en) | Method and system for optimizing loading layout of three-dimensional containers | |
McCormack et al. | Speaking the Buick language: capturing, understanding, and exploring brand identity with shape grammars | |
CN108595763A (en) | Die assembly design method based on model index technology | |
CN105788306B (en) | Import list suitable for travel speed fluctuation puts formula arterial highway green wave coordination design method | |
CN112215847B (en) | Method for automatically segmenting overlapped chromosomes based on counterstudy multi-scale features | |
CN108334949A (en) | A kind of tachytelic evolution method of optimization depth convolutional neural networks structure | |
CN106527381B (en) | A kind of fast evaluation method towards parallel batch processing machine dynamic dispatching | |
CN110909787B (en) | Method and system for optimizing multi-target batch scheduling based on cluster evolution algorithm | |
CN110085027B (en) | Method for decomposing large-scale road network group traffic flow induction tasks | |
CN104504462B (en) | Automobile form trend of evolution Forecasting Methodology based on improvement grey BPNN and markov chain | |
CN111445094B (en) | Express vehicle path optimization method and system based on time requirement | |
CN109598742A (en) | A kind of method for tracking target and system based on SSD algorithm | |
CN110473591A (en) | Idiotype network functional module based on quantum calculation is excavated and analysis method | |
CN105512755A (en) | Decomposition-based multi-objective distribution estimation optimization method | |
CN110008525A (en) | Automobile form characteristic crossover Evolution Forecasting method based on INGBM (1,1) | |
Coello | Evolutionary multiobjective optimization: current and future challenges | |
CN111580488B (en) | WBS buffer area vehicle sequencing scheduling method based on improved genetic algorithm | |
CN110097581A (en) | Method based on point cloud registering ICP algorithm building K-D tree | |
CN114298307A (en) | Mixed frog leaping solving method for mobile crowd sensing variable speed multi-task allocation problem | |
CN112749211A (en) | Novel tea yield prediction method based on electric power big data | |
CN104732292B (en) | Principal flow velocity line automatic planning based on profile data | |
CN102902823A (en) | System and method for searching traffic route |
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 | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190712 |