CN106296436A - Quantitative evaluation method for tobacco variety breeding - Google Patents

Abstract

The invention discloses a quantitative evaluation method for tobacco variety breeding, which comprises the following steps: 1) acquiring biological characters, economic characters and disease index indexes of the bred tobacco variety; 2) calculating the average value of each index, and carrying out averaging treatment; 3) calculating the optimal resolution coefficient of all indexes of the tobacco variety characters; 4) calculating the correlation coefficient of each index of different varieties respectively; 5) calculating the weight of each index; 6) calculating the comprehensive score of the bred tobacco varieties and sequencing; the quantitative evaluation method for tobacco variety breeding accurately quantifies and scores tobacco varieties, and quickly selects ideal flue-cured tobacco varieties.

Description

A kind of method for quantitatively evaluating of tobacco bred selection-breeding

Technical field

The invention belongs to tobacco bred selection-breeding field, be specifically related to the method for quantitatively evaluating of a kind of tobacco bred selection-breeding.

Background technology

At present, China Nicotiana tabacum L. (Nicotiana tabacum L.) kind plantation unification problem is serious, Yun yan85, cloud and mist The cultivated area of 3 kinds such as 87 and K326 reaches more than 70%, causes pest and disease damage to take place frequently, and tobacco style characteristic is single, serious shadow Ring the quality of Nicotiana tabacum L..In consideration of it, China tobacco breeding worker constantly introduces characteristic product from the country such as the U.S. and Zimbabwe Kind, develop self-fertile kind, try hard to the demand meeting Cigarette Industrial Enterprise to high-quality characteristic raw material.

During new product of tobacco selection-breeding, generally require filter out from tens or tens new flue-cured tobacco varieties several The preferable improved seeds of Comprehensive Traits, carry out plot experiment and field demonstration the most again.Existing breed breeding evaluation, is usually The factors such as selection Agronomic trait, economic characters, disease resistance and aesthetic quality, some key index in the analyzing influence factor, Such as yield and disease index etc..This method can catch the principal character of new varieties, but there is deficiency: one is easily to ignore newly Some details of kind, it is impossible to reflect whole trait information of this kind；Two is that some trait expression of new varieties is preferable, and another When some trait expression outer is poor, easily bring puzzlement to people, it is difficult to the quality of this kind of accurate evaluation；Three is the most energetic product Plant and evaluate score, bigger by experience influence fluctuation.

Summary of the invention

The present invention is directed to prior art problem, it is provided that tobacco bred is carried out quantifying marking by one accurately, quickly selects Select the method for quantitatively evaluating of the tobacco bred selection-breeding of preferable flue-cured tobacco cultivars.

Technology and method in order to solve the employing of the above-mentioned problem present invention are as follows:

The method for quantitatively evaluating of a kind of tobacco bred selection-breeding, comprises the following steps:

1) biological character of educated tobacco bred, Economic Characters and disease index index are obtained；

2) calculate the meansigma methods of each index, and carry out equalization process；

3) calculate tobacco bred character and all refer to the optimal resolution ratio of target；

${\mathrm{\Δ}}_{\stackrel{\‾}{X}}=\frac{1}{n\·m}\underset{i=1}{\overset{m}{\Σ}}\underset{k=1}{\overset{n}{\Σ}}|x_0\left(k\right)-x_i\left(k\right)|$

$\mathrm{\ρ}=1.5\frac{{\mathrm{\Δ}}_{\stackrel{\‾}{X}}}{{\mathrm{\Δ}}_{max}}$

Wherein, x₀(k)、x_iK () is index x₀、x_iFirst value sequence, k=1,2 ..., n, represent kth index；I= 1,2 ..., m；Represent i-th kind；It it is the average of all absolute difference；Δ_maxAbsolute value for maximum difference；ρ is Good resolution ratio；

4) coefficient of association of each index of different cultivars is calculated respectively；

${\mathrm{\ξ}}_{i\left(k\right)}=\frac{{\min }_i\min k|r_{0\left(k\right)}-r_{i\left(k\right)}|+{\mathrm{\ρmax}}_i{\max }_k|r_{0\left(k\right)}-r_{i\left(k\right)}|}{|r_{0\left(k\right)}-r_{i\left(k\right)}|+{\mathrm{\ρmax}}_i{\max }_k|r_{0\left(k\right)}-r_{i\left(k\right)}|}$

Wherein, ξ_i(k)It it is the coefficient of association of i-th kind kth index；

5) weight of each index is calculated；

6) calculate the comprehensive score of educated tobacco bred, and be ranked up.

Further, described biological character includes plant height, plant height of pinching, the actual number of blade, the number of blade of pinching, has The effect number of blade, stem girth, pitch, maximum leaf length and maximum width of blade.

Further, described Economic Characters includes yield, the output value, average price and better-than-average cigarette ratio.

Further, described disease index includes climacteric spot disease, anthrax, rust, powdery mildew, tobacco ordinary flower Leaf disease, balck shank and root black rot index.

Further, in step 5) in, the weighing computation method of each index is as follows:

Data principal component analysis after equalization is carried out factorial analysis, analyzes the characteristic root of main constituent and each main constituent Equation contribution rate, extract main constituent, obtain being rotated into sub matrix；By factor loading divided by the characteristic root of main constituent, obtain each Index coefficient in main constituent linear combination, is normalized, and calculates the weight of each index；

Z_j=β_1j*X₁+β_2j*X₂+…+β_nj*X_k

$Q_i=\frac{(m*\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{X}_i){\mathrm{\β}}_{ij}*{\mathrm{\γ}}_j}{(n*\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{X}_i)(m*\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{X}_i){\mathrm{\β}}_{ij}*{\mathrm{\γ}}_j}$

Wherein, Z_jFor main constituent linear combination (j=1,2 ..., p), X₁、X₂、…、X_nFor each index, β_1j、β_2j、 β_3j、……、β_njFor the principal component scores of each index, γ_jRepresent Z_jEquation contribution rate, Q_iIt it is index X_iWeight.

Further, in step 6) in, the comprehensive score computation method of tobacco bred is as follows；

D_i=ξ_i(1)*Q_i+ξ_i(2)*Q_i+…+ξ_i(k)*Q_i

Wherein, D_iComprehensive score by educated tobacco bred.

The invention have the benefit that according to gray system theory, it is provided that a kind of based on optimum resolution ratio objective All character of educated tobacco bred, before quality of tobacco overall merit, are carried out Comprehensive Assessment and quantitatively marking by evaluation methodology, The field growing situation of institute's selection-breeding tobacco bred can be judged, quickly select preferable flue-cured tobacco cultivars.

Accompanying drawing explanation

Fig. 1 is the schematic flow sheet of the present invention.

Fig. 2 is the schematic flow sheet of each index weights computational methods in the present invention.

Detailed description of the invention

Below in conjunction with embodiment, the present invention is described in further detail with accompanying drawing, but they are not to skill of the present invention The restriction of art scheme, based on present invention teach that any conversion made, all falls within protection scope of the present invention.

Refering to shown in Fig. 1-2, the method for quantitatively evaluating of a kind of tobacco bred selection-breeding, comprise the following steps:

1) biological character of educated tobacco bred, Economic Characters and disease index index are obtained；

2) calculate the meansigma methods of each index, and carry out equalization process；

3) calculate tobacco bred character and all refer to the optimal resolution ratio of target；

${\mathrm{\Δ}}_{\stackrel{\‾}{X}}=\frac{1}{n\·m}\underset{i=1}{\overset{m}{\Σ}}\underset{k=1}{\overset{n}{\Σ}}|x_0\left(k\right)-x_i\left(k\right)|$

$\mathrm{\ρ}=1.5\frac{{\mathrm{\Δ}}_{\stackrel{\‾}{X}}}{{\mathrm{\Δ}}_{max}}$

Wherein, x₀(k)、x_iK () is index x₀、x_iFirst value sequence, k=1,2 ..., n, represent kth index；I= 1,2 ..., m；Represent i-th kind；It it is the average of all absolute difference；Δ_maxAbsolute value for maximum difference；ρ is Good resolution ratio；

4) coefficient of association of each index of different cultivars is calculated respectively；

${\mathrm{\ξ}}_{i\left(k\right)}=\frac{{\min }_i\min k|r_{0\left(k\right)}-r_{i\left(k\right)}|+{\mathrm{\ρmax}}_i{\max }_k|r_{0\left(k\right)}-r_{i\left(k\right)}|}{|r_{0\left(k\right)}-r_{i\left(k\right)}|+{\mathrm{\ρmax}}_i{\max }_k|r_{0\left(k\right)}-r_{i\left(k\right)}|}$

Wherein, ξ_i(k)It it is the coefficient of association of i-th kind kth index；

5) weight of each index is calculated；

6) calculate the comprehensive score of educated tobacco bred, and be ranked up.

Embodiment 1

1. the meansigma methods that 10 are participated in representational 8 character of selection-breeding tobacco lines is listed in the table below.

2. with DPS7.05, initial data is carried out equalization process

When being modeled by gray system, owing to data volume level difference is bigger, it is necessary to carry out nondimensionalization process, make because of Prime sequences planningization, just can calculate.Just value algorithm is applicable to process stable economics phenomenon, and equalization is usually used in place Manage the data sequence without visible trend, therefore initial data is carried out equalization process.Additionally due to disease index value is the biggest, cigarette Grass Performance of cultivar is the poorest, and therefore to such index first inverseization, then equalization processes.

3. calculate optimal resolution ratio

3.1 differences calculating tobacco bred character

3.2 calculate optimal resolution ratio

According to formula (1), draw Δ_X=0.1366, Δ_max=0.9553；Now will affect the whole of tobacco bred character Index, as a gray system, according to formula (2), draws ρ=0.2145.

4. calculate the coefficient of association of each index of different cultivars

According to formula (3), draw the coefficient of association ξ of each index of different cultivars_i(k)。

Strain 1	0.6446	0.5988	0.8609	0.5181	0.5322	0.6053	0.75582	0.1766
									Strain 2	0.9706	0.7177	0.9976	0.7782	0.7085	1	1	1
Strain 3	1	1	0.9976	1	0.8753	0.8457	0.722	1
									Strain 4	0.6385	0.5615	0.9642	0.5416	0.9156	0.9421	0.66527	1
Strain 5	0.8809	0.5859	0.8952	0.572	0.534	0.4492	0.5535	0.232
									Strain 6	0.6322	0.5902	0.8555	0.536	0.8183	0.5599	0.52179	0.1808
Strain 7	0.7617	0.5714	0.8394	0.5288	1	0.8154	0.64233	0.5273
									Strain 8	0.6805	0.6515	0.9242	0.5266	0.7063	0.5296	0.50984	0.3173
Strain 9	0.6503	0.6077	0.9416	0.4932	0.619	0.9557	0.86603	0.1858
									Strain 10	0.6869	0.6241	1	0.6037	0.6925	0.9526	0.83361	0.2507

5. calculate the weight of each index

5.1 extract main constituent

Analyze software with SPSS16.0, the data after equalization carried out principal component analysis, draw the population variance table of explanation, Analyze characteristic root > main constituent of 1, find that the characteristic root of composition 1,2,3 is 4.078,1.563 and 1.366 respectively, and accumulative contribution Rate reaches 87.581%, more than 85%, determines that main constituent is 1,2,3.

5.2 obtain main constituent linear combination

By factor loading divided by the characteristic root of main constituent, obtain each index coefficient in main constituent linear combination, draw Formula (4)

Z₁=0.458X4+0.452X1+0.433X2+0.078X7+0.002X6+0.256X3+0.013X5+ 0.256X8；

Z₂=0.187X4+0.008X1+0.156X2+0.751X7+0.705X6+0.526X3+0.019X5+ 0.161X8；

Z₃=0.172X4+0.049X1+0.082X2-0.107X7+0.385X6+0.083X3+0.836X5+ 0.628X8；

5.3 draw weight

According to formula (5), draw the weight of each index:

6. calculate the comprehensive score of educated tobacco bred

According to formula (6), draw the Comprehensive Traits score of educated tobacco bred, and be ranked up.

Strain	Comprehensive score	Sequence
			Strain 1	0.574755	9
Strain 2	0.895245	2
			Strain 3	0.9469	1
Strain 4	0.767041	3
			Strain 5	0.589473	8
Strain 6	0.572031	10
			Strain 7	0.687694	4
Strain 8	0.600968	7
			Strain 9	0.640903	6
Strain 10	0.686003	5

The invention have the benefit that according to gray system theory, it is provided that a kind of based on optimum resolution ratio objective All character of educated tobacco bred, before quality of tobacco overall merit, are carried out Comprehensive Assessment and quantitatively marking by evaluation methodology, The field growing situation of institute's selection-breeding tobacco bred can be judged, quickly select preferable flue-cured tobacco cultivars.

The above, the only detailed description of the invention of the present invention, but protection scope of the present invention is not limited thereto, and any The change expected without creative work or replacement, all should contain within protection scope of the present invention, therefore, and the present invention's Protection domain should be as the criterion with the protection domain that claims are limited.

Claims (6)

1. the method for quantitatively evaluating of a tobacco bred selection-breeding, it is characterised in that: comprise the following steps:

1) biological character of educated tobacco bred, Economic Characters and disease index index are obtained；

2) calculate the meansigma methods of each index, and carry out equalization process；

3) calculate tobacco bred character and all refer to the optimal resolution ratio of target；

${\mathrm{\Δ}}_{\stackrel{\‾}{X}}=\frac{1}{n\·m}\underset{i=1}{\overset{m}{\Σ}}\underset{k=1}{\overset{n}{\Σ}}|x_0\left(k\right)-x_i\left(k\right)|$

$\mathrm{\ρ}=1.5\frac{{\mathrm{\Δ}}_{\stackrel{\‾}{X}}}{{\mathrm{\Δ}}_{max}}$

Wherein, x₀(k)、x_iK () is index x₀、x_iFirst value sequence, k=1,2 ..., n, represent kth index；I=1, 2 ..., m；Represent i-th kind；Δ_XIt it is the average of all absolute difference；Δ_maxAbsolute value for maximum difference；ρ is optimal Resolution ratio；

4) coefficient of association of each index of different cultivars is calculated respectively；

${\mathrm{\ξ}}_{i\left(k\right)}=\frac{{\min }_i\min k|r_{0\left(k\right)}-r_{i\left(k\right)}|+{\mathrm{\ρmax}}_i{\max }_k|r_{0\left(k\right)}-r_{i\left(k\right)}|}{|r_{0\left(k\right)}-r_{i\left(k\right)}|+{\mathrm{\ρmax}}_i{\max }_k|r_{0\left(k\right)}-r_{i\left(k\right)}|}$

Wherein, ξ_i(k)It it is the coefficient of association of i-th kind kth index；

5) weight of each index is calculated；

6) calculate the comprehensive score of educated tobacco bred, and be ranked up.

The method for quantitatively evaluating of tobacco bred selection-breeding the most according to claim 1, it is characterised in that: described biological character Including plant height, plant height of pinching, the actual number of blade, the number of blade of pinching, effective blade number, stem girth, pitch, maximum leaf length and Great Ye width.

The method for quantitatively evaluating of tobacco bred selection-breeding the most according to claim 2, it is characterised in that: described Economic Characters Including yield, the output value, average price and better-than-average cigarette ratio.

The method for quantitatively evaluating of tobacco bred selection-breeding the most according to claim 3, it is characterised in that: described disease index bag Include climacteric spot disease, anthrax, rust, powdery mildew, tobacco mosaic virus (tmv), balck shank and root black rot index.

5. according to the method for quantitatively evaluating of the tobacco bred selection-breeding according to any one of claim 1-4, it is characterised in that: in step Rapid 5), in, the weighing computation method of each index is as follows:

Data principal component analysis after equalization is carried out factorial analysis, analyzes characteristic root and the side of each main constituent of main constituent Journey contribution rate, extracts main constituent, obtains being rotated into sub matrix；By factor loading divided by the characteristic root of main constituent, obtain each index Coefficient in main constituent linear combination, is normalized, and calculates the weight of each index；

Z_j=β_1j*X₁+β_2j*X₂+…+β_nj*X_k

$Q_i=\frac{(m*\underset{i=1}{\overset{n}{\Σ}}{X}_i){\mathrm{\β}}_{ij}*{\mathrm{\γ}}_j}{(n*\underset{i=1}{\overset{n}{\Σ}}{X}_i)(m*\underset{i=1}{\overset{n}{\Σ}}{X}_i){\mathrm{\β}}_{ij}*{\mathrm{\γ}}_j}$

Wherein, Z_jFor main constituent linear combination (j=1,2 ..., p), X₁、X₂、…、X_nFor each index, β_1j、β_2j、β_3j、……、 β_njFor the principal component scores of each index, γ_jRepresent Z_jEquation contribution rate, Q_iIt it is index X_iWeight.

6. according to the method for quantitatively evaluating of the tobacco bred selection-breeding according to any one of claim 1-4, it is characterised in that: in step Rapid 6), in, the comprehensive score computation method of tobacco bred is as follows；

D_i=ξ_i(1)*Q_i+ξ_i(2)*Q_i+…+ξ_i(k)*Q_i

Wherein, D_iComprehensive score by educated tobacco bred.