Rice variety comprehensive evaluation method based on CRITIC-TOPSIS

Abstract

The invention provides a comprehensive evaluation method for a rice variety based on CRITIC-TOPSIS (Clustered Regularly Interspaced Interspaced apart Clusters-TOPSIS Comprising the following steps: step 1, obtaining related data of a tested variety; step 2, carrying out normalization processing on the data; step 3, calculating an index fluctuation degree and a conflict degree; step 4, calculating an index information bearing capacity; 8, calculating the Euclidean distance between each variety and the positive and negative ideal solutions; and 9, calculating the relative pasting degree of each variety and the ideal solutions. According to the method, human interference factors are eliminated, the index weight which contributes little to the evaluation result is reduced, and the objectivity and accuracy of evaluation are improved.

Description

Technical Field

[0001] The invention belongs to the technical field of rice variety evaluation, and in particular relates to a rice variety evaluation method based on CRITIC-TOPSIS. Background Art

[0002] Since rice planting layout is highly regional, accurate, scientific and comprehensive analysis and evaluation of rice variety characteristics based on the ecological and environmental characteristics of different regions is of great significance for variety selection, management and application. The main task of regional trials is to evaluate the high yield, stable yield, stress resistance and high quality characteristics of rice varieties. However, multi-index evaluation of varieties is a key and complex challenge that breeders and variety managers need to face.

[0003] The CRITIC weighting method is based on the two core concepts of data volatility and conflict, and determines the weight distribution by analyzing the differences in the values ​​of each indicator. The TOPSIS method is to non-dimensionalize the trait indicators of multiple varieties so that they can be compared with each other. By comparing the differences between various indicators, the fuzzy indicator of the comprehensive trait of the variety is converted into the degree of closeness to the ideal solution, thereby obtaining the superiority and inferiority ranking of each variety.

[0004] The traditional TOPSIS method often adopts experience-based weighting or entropy weighting. The former is greatly influenced by subjective factors, while the latter only considers the dispersion of indicator data and does not consider the correlation between indicators, which may lead to the evaluation results being not objective and accurate enough.

[0005] Prior art 1

[0006] The application number is CN202010107617.9, and the name of the invention is a comprehensive evaluation method for rice varieties. It uses an analysis method that combines entropy weight and efficacy to screen and evaluate the excellence of varieties, solving the subjective, one-sided and reliability problems of rice variety evaluation methods, and the difficulty of comprehensively and objectively analyzing and distinguishing varieties that are "good to grow", "good to taste" and "good to process" to match the needs of yield, quality, stress resistance and other multi-dimensional traits. The implementation of this plan mainly includes the following steps:

[0007] (1) Establish an evaluation system; (2) Obtain indicator data; (3) Determine indicator weights; (4) Calculate the comprehensive index; (5) Evaluate excellent varieties.

[0008] The disadvantage is that the invention is based on the principle of information entropy and determines the weight of each attribute by calculating the difference of the attributes. It assumes that the greater the difference, the more important the attribute is to the decision. However, this method is highly dependent on the data, especially when there are many data outliers or the data fluctuates greatly, the entropy weight method may be misleading. When calculating the weight, the entropy weight method is mainly based on the data distribution difference of a single attribute, without considering the correlation between different attributes. If there is a strong correlation between the attributes, the entropy weight method may overestimate the importance of certain attributes, resulting in unreasonable weight distribution. The entropy weight method is more sensitive to noise in the data. Because it depends on the discreteness of the attribute value, noise may cause the weights of some attributes to be overestimated or underestimated, thereby affecting the accuracy of the decision result. Summary of the invention

[0009] The purpose of the present invention is to solve the defects of the above-mentioned prior art and provide a comprehensive evaluation method of rice varieties based on CRITIC-TOPSIS, which eliminates human interference factors, reduces the weight of indicators that contribute less to the evaluation results, and improves the objectivity and accuracy of the evaluation.

[0010] The present invention adopts the following technical solution:

[0011] A comprehensive evaluation method for rice varieties based on CRITIC-TOPSIS, including:

[0012] Step 1. Plant the tested varieties and collect data on 16 evaluation indicators, including average yield per mu, whole growth period, rough rice rate, polished rice rate, chalkiness, amylose content, gel consistency, alkali elimination value, transparency, quality grade, yield point rate, fruit setting rate, comprehensive rice blast index, rice blast resistance grade, sheath blight resistance grade and rice false smut resistance grade, as required;

[0013] Step 2. Normalize the data. Normalization maps all indicator data to the interval [0,1] through linear transformation. The indicators are divided into positive indicators, negative indicators and intermediate indicators according to the evaluation objectives:

[0014]

[0015] Where, X ij is the normalized data; i = 1, 2, 3, ..., m; j = 1, 2, 3, ..., n; m and n represent the number of varieties and the number of evaluation indicators respectively; x j is all the original data of the jth indicator; x ij is the original data of the jth indicator of the ith variety; x best is the optimal target value of the jth indicator.

[0016] Step 3. Calculate the volatility σ of the jth indicatorj , that is, the standard deviation:

[0017]

[0018] In the formula, is the mean value of the jth indicator;

[0019] Calculate the correlation coefficient r between each indicator tj , using the Pearson correlation coefficient, the formula is:

[0020]

[0021] In the formula, r tj is the correlation coefficient between the tth indicator and the jth indicator;

[0022] Calculate the conflict degree R of the jth indicator j :

[0023]

[0024] Step 4. Calculate the information carrying capacity C of the jth indicator j :

[0025] C j =σ j ·R j (5)

[0026] Step 5. Calculate the weight W of the jth indicator j :

[0027]

[0028] Step 6. Construct the weighted norm matrix Z ij :

[0029] Z ij =W j ·X ij (7)

[0030] Step 7. Determine the positive ideal solution for each indicator and negative ideal solution

[0031]

[0032] Step 8. Calculate the Euclidean distance between each variety and the positive and negative ideal solutions:

[0033]

[0034] Step 9. Calculate the relative progress T of each variety to the ideal solution i , T iThe larger the value, the better the comprehensive evaluation of the variety.

[0035]

[0036] Beneficial effects of the invention:

[0037] (1) CRITIC-TOPSIS has shown significant advantages in rice variety evaluation and can effectively improve the scientificity and accuracy of new variety breeding and variety management decisions.

[0038] (2) The present invention overcomes the limitations of the traditional TOPSIS method. The present invention combines the CRITIC and TOPSIS methods to eliminate human interference factors, reduce the weight of indicators that contribute less to the evaluation results, and improve the objectivity and accuracy of the evaluation. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 The figure is a flow chart of the method steps of the present invention. DETAILED DESCRIPTION

[0040] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention is described clearly and completely below. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0041] The present invention proposes a comprehensive evaluation method for rice varieties based on CRITIC-TOPSIS. By comprehensively evaluating the late-maturing mid-season rice varieties in the Hubei provincial regional trials in 2023, the present invention provides a more scientific and comprehensive evaluation result, and provides a reference basis for setting variety approval indicators for the breeding goals of new varieties.

[0042] Example

[0043] Materials and Methods

[0044] Trial Overview

[0045] The identification of rice blast, sheath blight and false smut in this experiment was carried out at the Lianghe Base and Hongmiao Base of the Enshi Autonomous Prefecture Academy of Agricultural Sciences. The identification of other index characteristics was completed by the designated unit of the Hubei Provincial Seed Administration Bureau. All experiments were carried out in accordance with the requirements of the "Implementation Plan for Regional Experiments of Rice Varieties in Hubei Province". A total of 51 varieties were tested, divided into 4 groups, and all varieties were provided by the Hubei Provincial Seed Administration Bureau.

[0046] Among them, the "Implementation Plan for Regional Trials of Rice Varieties in Hubei Province" comes from the Hubei Provincial Neutron Management Bureau, 2023 Hubei Province Rice Variety Regional Trial Implementation Plan: https: / / nyt.hube i.gov.cn / bmdt / yw / ywtz / szzg l ij_9002 / 202303 / t20230303_4569107.shtm l.

[0047] Data Source

[0048] like Figure 1 As shown, the four indicator data of rice blast comprehensive index, rice blast resistance level, sheath blight resistance level, and rice false smut resistance level in this experiment were collected and sorted by the Plant Protection and Soil and Fertilizer Institute of Enshi Autonomous Prefecture Academy of Agricultural Sciences, and the other indicator data were derived from the public data of Hubei Provincial Department of Agriculture and Rural Affairs.

[0049] Data analysis

[0050] Data normalization. When conducting evaluation analysis, since each evaluation indicator has different dimensions, normalization must be performed first to eliminate the dimension effect and ensure the consistency and comparability of the data. Normalization maps all indicator data to the [0,1] interval through linear transformation. Among them, indicators are divided into positive indicators, negative indicators and intermediate indicators according to the evaluation objectives.

[0051]

[0052] In the formula, x j is all the original data of the jth indicator; x ij is the original data of the jth indicator of the ith variety; x best is the optimal target value of the jth indicator. ij is the normalized data; i = 1, 2, 3, ..., m; j = 1, 2, 3, ..., n; m and n represent the number of varieties and the number of evaluation indicators respectively.

[0053] Determine the indicator weights using the CRITIC method: The core of the CRITIC method is to evaluate the volatility and conflict of each indicator. Specifically, the larger the standard deviation, that is, the greater the volatility, the more information the indicator reflects, and therefore a larger weight should be given. On the contrary, the higher the correlation coefficient between indicators, the less conflict, and the smaller the corresponding weight should be. The specific steps are as follows:

[0054] Calculate the volatility σ of the jth indicator j , that is, the standard deviation:

[0055]

[0056] In the formula, is the mean value of the jth indicator.

[0057] Calculate the correlation coefficient r between each indicator tj , using the Pearson correlation coefficient, the formula is:

[0058]

[0059] In the formula, r tj is the correlation coefficient between the tth indicator and the jth indicator.

[0060] Calculate the conflict degree R of the jth indicator j :

[0061]

[0062] Calculate the information carrying capacity C of the jth indicator j :

[0063] C j =σ j ·R j (5)

[0064] Calculate the weight W of the jth indicator j :

[0065]

[0066] Comprehensive evaluation by TOPSIS method: The TOPSIS method is a decision support method for evaluating and determining the relative merits of different objects. This method achieves a systematic evaluation of the merits of varieties by constructing a weighted normative matrix, determining the ideal and negative ideal solutions for each indicator, and calculating the distance between each variety and these ideal solutions. The core of this method is to compare the similarity between each variety and the ideal optimal state (ideal solution) and the least ideal state (negative ideal solution) to determine the relative ranking of each variety. The specific steps are as follows:

[0067] Construct the weighted norm matrix Z ij :

[0068] Z ij =W j ·X ij (7)

[0069] Determine the positive ideal solution for each indicator and negative ideal solution

[0070]

[0071] Calculate the Euclidean distance between each variety and the positive and negative ideal solutions:

[0072]

[0073] Calculate the relative progress T of each variety to the ideal solution i , T i The larger the value, the better the comprehensive evaluation of the variety.

[0074]

[0075] Results and Discussion

[0076] Build a comparison matrix

[0077] The 16 index data of 51 late-maturing mid-season rice varieties in Hubei Province in 2023 were constructed into a comparison matrix. The 16 evaluation indicators are average yield per mu, full growth period, rough rice rate, polished rice rate, chalkiness, amylose content, gel consistency, alkali elimination value, transparency, quality grade, yield point rate, fruit setting rate, comprehensive index of rice blast, resistance to rice blast, resistance to sheath blight and resistance to false smut. The index data of some varieties are listed in Table 1.

[0078] Table 1 Index data of some late-maturing mid-season rice varieties in Hubei Province regional trials in 2023

[0079]

[0080] Constructing the weighted normative matrix

[0081] According to actual production needs, average yield per mu, roughness rate, polished rice rate, gel consistency, alkali elimination value, yield point rate, and fruiting rate are used as positive indicators. The whole growth period, chalkiness, transparency, comprehensive index of rice blast, rice blast resistance, sheath blight resistance, and rice false smut resistance are used as negative indicators. The value of amylose content is converted into a negative indicator with reference to the agricultural industry standard, that is, greater than or equal to 13% and less than or equal to 18% is recorded as 1, greater than 18% and less than or equal to 20% is recorded as 2, greater than 20% and less than or equal to 22% is recorded as 3, and others are recorded as 4. The quality grades "Excellent 1", "Excellent 2", "Excellent 3", and "Ordinary" are recorded as 1, 2, 3, and 4 respectively, as negative indicators. According to formula (1), the normalized standard matrix X is obtained ij , as shown in Table 2.

[0082] Table 2 Normalized standard values ​​of some varieties of late-maturing mid-season rice in the regional trial in Hubei Province in 2023

[0083]

[0084] According to formulas (2) to (6), the volatility index, conflict index, information carrying capacity and corresponding weights were calculated. The specific data are shown in Table 3. The results show that the weights of each index are ranked from high to low as follows: rice false smut resistance grade, alkali elimination value grade, quality grade, rice blast resistance grade, transparency, head rice rate, gel consistency, and amylose content. Among them, the total weight of the seven indexes related to rice quality (head rice rate, chalkiness, amylose content, gel consistency, alkali elimination value grade, transparency, and quality grade) is 0.437, which reflects that rice quality indexes occupy a significant position in the rice evaluation system, highlighting that in the process of modern rice breeding and variety improvement, quality factors have a higher priority than other characteristics. The total weight of the four indexes related to disease resistance (rice blast comprehensive index, rice blast resistance grade, sheath blight resistance grade, and rice false smut resistance grade) is 0.282, highlighting the core role of disease resistance in maintaining rice production safety. The combined weight of the three yield-related indicators (average yield per mu, yield increase point rate, and fruiting rate) is 0.172, indicating that yield is still an important consideration in the overall evaluation. In addition, the weight of the roughness rate is 0.048, and the weight of the whole growth period is 0.056, which shows that in the breeding and variety selection process, although these indicators have a certain impact on the performance of the overall rice variety, they have a lower priority than quality and disease resistance. The roughness rate mainly affects the processing quality and market value of rice, while the whole growth period is related to the crop's growth cycle and seasonal adaptability, which may become important under specific environments or market demands. However, in the broad range of variety improvement and evaluation criteria, improving disease resistance and quality characteristics are usually more directly able to improve the commercial value of crops and the sustainability of production.

[0085] Table 3 Weights and rankings of different indicators

[0086]

[0087] After the CRITIC method determines the indicator weights, formula (7) is introduced to construct the weighted normative matrix, and the positive ideal solution and negative ideal solution of each indicator are calculated according to formulas (8) and (9). Then, the Euclidean distance between each variety and the positive ideal solution and the negative ideal solution is calculated according to formulas (10) and (11). The comprehensive score T of each variety is calculated based on the distance. i , as shown in Table 4. T i The larger the value, the closer the comprehensive characteristics of the variety are to the ideal breeding goal. i The top five varieties are: Yangxian 9A / HD377, Zhenliangyou Eyingsi Miao, Hanliangyou 6510, EK8S / HD227, and Hualiangyou 1834. iThe value is 0.611. Yield alone has obvious limitations when evaluating varieties based on yield. Judging the quality of varieties based on yield alone often ignores other key agronomic traits, such as disease resistance and stress tolerance. This bias may lead to the fact that disease-resistant and high-quality varieties are not valued on the one hand, and the promoted high-yield varieties perform poorly under specific conditions on the other hand. For example, "Kai 68S / ZR985" ranks first in yield, but T i The value is relatively low; the T i The value ranking is 6th, and the output ranking is 9th. i The value ranking is 8th and the yield ranking is 6th, indicating that this type of variety has a good balance performance in high yield and multiple agronomic traits, and has the potential to be a high-yield and high-quality rice variety.

[0088] Table 4 Comprehensive evaluation results of late-maturing mid-season rice regional trial varieties in Hubei Province in 2023 using TOPSIS method

[0089]

[0090] The test results show:

[0091] The present invention selected 16 characteristic indicators of 51 varieties of late-maturing mid-season rice in the 2023 Hubei Province regional test as test data, and used the CRITIC method to assign weights to the 16 indicators. The results showed that the total weight of the 7 indicators related to rice quality was 0.437, indicating that quality factors have a higher priority in modern rice breeding and variety improvement. The total weight of the 4 indicators related to disease resistance is 0.282, indicating the core role of disease resistance in maintaining rice production safety. Although yield-related indicators are important, their total weight is 0.172, indicating that their relative importance in the comprehensive variety improvement strategy is relatively low.

[0092] The TOPSIS method was used to further determine the comprehensive score T of each variety. i Value, where T i The varieties with the top five values ​​showed comprehensive characteristics that were closer to the ideal breeding goal. In particular, varieties such as "Yangxian 9A / HD377" and "Zhenliangyou Eyingsi Miao" not only had good T i The value of "Kai 68S / ZR985" is ranked high, and the importance of key agronomic traits such as disease resistance and stress tolerance is highlighted. At the same time, the present invention also reveals the limitations of relying solely on yield as an evaluation criterion. For example, although "Kai 68S / ZR985" has a high yield, its performance in the comprehensive evaluation is not ideal.

[0093] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A comprehensive evaluation method for rice varieties based on CRITIC-TOPSIS, characterized in that: include: Step 1. Plant the tested varieties and collect data on 16 evaluation indicators, including average yield per mu, whole growth period, rough rice rate, polished rice rate, chalkiness, amylose content, gel consistency, alkali elimination value, transparency, quality grade, yield point rate, fruit setting rate, comprehensive rice blast index, rice blast resistance grade, sheath blight resistance grade and rice false smut resistance grade, as required; Step 2. Normalize the data. Normalization maps all indicator data to the interval [0,1] through linear transformation. The indicators are divided into positive indicators, negative indicators and intermediate indicators according to the evaluation objectives: X ij is the normalized data; i = 1, 2, 3, ..., m; j = 1, 2, 3, ..., n; m and n represent the number of varieties and the number of evaluation indicators respectively; x j is all the original data of the jth indicator; x ij is the original data of the jth indicator of the ith variety; x best is the optimal target value of the jth indicator; Step 3. Calculate the volatility σ of the jth indicator j , that is, the standard deviation: In the formula, is the mean value of the jth indicator; Calculate the correlation coefficient r between each indicator tj , using the Pearson correlation coefficient, the formula is: In the formula, r tj is the correlation coefficient between the tth indicator and the jth indicator; Calculate the conflict degree R of the jth indicator j : Step 4. Calculate the information carrying capacity C of the jth indicator j : C j =s j ·R j (5) Step 5. Calculate the weight W of the jth indicator j : Step 6. Construct the weighted norm matrix Z ij : Z ij =W j ·X ij (7) Step 7. Determine the positive ideal solution for each indicator and negative ideal solution Step 8. Calculate the Euclidean distance between each variety and the positive and negative ideal solutions: Step 9. Calculate the relative progress T of each variety to the ideal solution i : T i The larger the value, the better the comprehensive evaluation of the variety.

2. The method according to claim 1, characterized in that The requirements for collecting data in step 1 are specifically the Implementation Plan for Regional Trials of Rice Varieties in Hubei Province.

3. The method according to claim 1, characterized in that In step 2, the average yield per mu, rough rice rate, whole polished rice rate, gel consistency, alkali elimination value, yield increase point rate, and fruit setting rate are used as positive indicators, and the whole growth period, chalkiness, transparency, comprehensive rice blast index, rice blast resistance level, sheath blight resistance level, and rice false smut resistance level are used as negative indicators.

4. The method according to claim 1, characterized in that: In step 2, the value of amylose content is converted into a negative indicator with reference to the agricultural industry standard, that is, greater than or equal to 13% and less than or equal to 18% is recorded as 1, greater than 18% and less than or equal to 20% is recorded as 2, greater than 20% and less than or equal to 22% is recorded as 3, and the rest is recorded as 4.

5. The method according to claim 1, characterized in that In step 2, the quality level is converted into a negative indicator, that is, "Excellent 1", "Excellent 2", "Excellent 3", and "Ordinary" are recorded as 1, 2, 3, and 4 respectively.

6. The method according to claim 1, characterized in that Step 3 includes that the larger the standard deviation, the more information the indicator reflects, and therefore a larger weight should be given. On the contrary, the higher the correlation coefficient between indicators, the smaller the conflict, and the smaller the corresponding weight should be.

Images