CN-122022155-A - Bee seed preservation effect evaluation method and device, electronic equipment and storage medium

CN122022155ACN 122022155 ACN122022155 ACN 122022155ACN-122022155-A

Abstract

The application provides a method, a device, electronic equipment and a storage medium for evaluating bee species retention effect, wherein the method comprises the steps of obtaining a plurality of original morphological indexes of bees to be tested of target varieties, inputting the plurality of original morphological indexes of the bees to be tested into a XGBoost model obtained through training in advance for learning analysis, outputting a plurality of target morphological indexes corresponding to the target varieties through the XGBoost model, and evaluating the varieties of the bees to be tested according to the target morphological indexes to obtain evaluation results of the bee species retention effect to be tested. The key morphological index intelligent screening and assessment flow standardization is realized by means of XGBoost model, the rapid detection is realized, and the efficiency and scientificity of seed preservation work are greatly improved.

Inventors

DENG XIAOYIN
XU KAI
DU YALI
HE JINMING
JIANG HAIBIN
WU JIAXU
NIU QINGSHENG
LIU YULING

Assignees

吉林省养蜂科学研究所（吉林省蜂产品质量管理监督站、吉林省蜜蜂遗传资源基因保护中心）

Dates

Publication Date: 20260512
Application Date: 20260128

Claims (10)

1. A method for evaluating bee seed preservation effect, the method comprising: acquiring a plurality of original morphological indexes of bees to be detected of a target variety; Inputting a plurality of original morphological indexes of the bees to be tested into a XGBoost model obtained through training in advance for learning analysis, and outputting a plurality of target morphological indexes corresponding to the target variety by using the XGBoost model; And evaluating the varieties of the bees to be tested according to the target morphological indexes to obtain an evaluation result of the seed preservation effect of the bees to be tested.
2. The method for evaluating bee seed-keeping effect according to claim 1, wherein the plurality of target morphological indexes corresponding to the target variety include kissing length, sixth web width, sixth web length, front fin angle D7, front fin angle A4, front fin angle N23, front fin angle 026, third web wax lens pitch, elbow length, front fin angle L13.
3. The method for evaluating the bee seed preservation effect according to claim 1, wherein the parameters adopted by the XGBoost model in training at least comprise 100 weak learners, 0.3 learning rate, 0 minimum loss reduction required by tree splitting, 15 maximum depth of tree, 1 minimum value of hessian matrix sum in child nodes, 1 sample ratio used in training each tree, 0.8 characteristic ratio used in constructing each tree, two-class logistic regression as objective function, and logarithmic loss as evaluation standard.
4. The method for evaluating bee species retention according to claim 1, wherein evaluating the variety of the bee to be tested according to each target morphological index to obtain a target evaluation result comprises: inputting each target morphological index into a pre-constructed classification and identification model, and determining a seed preservation effect evaluation value of the bees to be tested according to each target morphological index by the classification and identification model to obtain the target evaluation result.
5. The method for evaluating bee seed retention according to claim 1, wherein the training process of XGBoost model is as follows: collecting a plurality of sample data, wherein the sample data comprise sample data of a target variety and sample data of a comparison bee species, and each sample data comprise all original morphological indexes obtained by pre-measurement; Performing complement processing on the missing values in the plurality of sample data to obtain a plurality of processed sample data, taking the plurality of processed sample data with a first proportion as a training set, and taking the plurality of processed sample data with a second proportion as a test set; And inputting the training set into an initial XGBoost model for training, and verifying the trained XGBoost model according to the test set to obtain the XGBoost model.
6. The method for evaluating bee seed preservation effect according to claim 1, wherein the outputting, by the pre-trained XGBoost model, at least one target morphological index corresponding to the bee to be tested includes: outputting scores of all original morphological indexes according to all original morphological indexes by the pre-trained XGBoost model; sorting the scores of the morphological indexes to obtain sorted original morphological indexes; And determining and outputting the target morphological indexes according to the ordered original morphological indexes.
7. The method for evaluating bee seed retention according to claim 6, wherein determining and outputting the plurality of target morphology metrics based on the ranked raw morphology metrics comprises: and selecting a preset number of original morphology indexes which are sequenced in front from the sequenced original morphology indexes as the target morphology indexes.
8. An evaluation device for bee seed preservation effect, which is characterized by comprising: the acquisition module is used for acquiring a plurality of original morphological indexes of the bees to be detected of the target variety; The learning module is used for inputting a plurality of original morphological indexes of the bees to be tested into a XGBoost model which is obtained through training in advance for learning analysis, and outputting a plurality of target morphological indexes corresponding to the target variety through the XGBoost model; The determining module is used for evaluating the varieties of the bees to be tested according to the target morphological indexes to obtain the evaluation result of the seed preservation effect of the bees to be tested.
9. An electronic device comprising a memory and a processor, the memory storing a computer program executable by the processor, the processor executing the computer program to perform the steps of the method of assessing bee seed retention as claimed in any one of claims 1 to 7.
10. A computer-readable storage medium, characterized in that the computer-readable storage medium has stored thereon a computer program which, when executed by a processor, performs the steps of the method for evaluating bee seed retention as claimed in any one of claims 1 to 7.

Description

Bee seed preservation effect evaluation method and device, electronic equipment and storage medium Technical Field The application relates to the technical field of bee raising, in particular to a method and a device for evaluating bee seed preservation effect, electronic equipment and a storage medium. Background Bees, one of the most important pollinating insects worldwide, play an irreplaceable role in maintaining the balance of the ecosystem and guaranteeing agricultural production. The western bees are widely applied to modern agriculture due to high pollination capability and good breeding adaptability, while the indigenous bees species such as Hunchun black bees have excellent cold resistance, disease resistance and high adaptability to local environment, and are precious germplasm resources. Along with the changes of ecological environment, the introduction of external bee species and the intensive development of bee-keeping activities, the indigenous bee species face serious threats such as genetic erosion, hybridization pollution, population atrophy and the like, so that the development of scientific and effective bee species conservation work has become urgent demands for biodiversity protection and agricultural sustainable development. At present, the evaluation of bee seed preservation effect mainly depends on two main technologies, namely a traditional molecular marking method and a conventional morphological determination method. However, both methods have significant limitations in practical application, and it is difficult to meet the requirements of high-efficiency, low-cost and high-precision seed conservation monitoring. Disclosure of Invention The application aims to provide a method, a device, electronic equipment and a storage medium for evaluating the bee seed-keeping effect, aiming at the defects in the prior art, and improving the accuracy of evaluating the bee seed-keeping effect. In order to achieve the above purpose, the technical scheme adopted by the embodiment of the application is as follows: in a first aspect, an embodiment of the present application provides a method for evaluating a bee seed-keeping effect, where the method includes: acquiring a plurality of original morphological indexes of bees to be detected of a target variety; Inputting a plurality of original morphological indexes of the bees to be tested into a XGBoost model obtained through training in advance for learning analysis, and outputting a plurality of target morphological indexes corresponding to the target variety by using the XGBoost model; And evaluating the varieties of the bees to be tested according to the target morphological indexes to obtain an evaluation result of the seed preservation effect of the bees to be tested. Optionally, the multiple target morphological indexes corresponding to the target variety include kissing length, sixth web width, sixth web length, front fin angle D7, front fin angle A4, front fin angle N23, front fin angle 026, third web wax lens spacing, elbow length, front fin angle L13. Optionally, parameters adopted by the XGBoost model in training at least comprise 100 weak learners, 0.3 learning rate, 0 minimum loss reduction required by tree splitting, 15 maximum depth of the tree, 1 minimum value of hessian matrix sum in the child nodes, 1 sample ratio used in training each tree, 0.8 characteristic ratio used in constructing each tree, two-class logistic regression as an objective function, and logarithmic loss as an evaluation standard. Optionally, the evaluating the variety of the bee to be tested according to each target morphological index to obtain a target evaluation result includes: inputting each target morphological index into a pre-constructed classification and identification model, and determining a seed preservation effect evaluation value of the bees to be tested according to each target morphological index by the classification and identification model to obtain the target evaluation result. Optionally, the training process of XGBoost model is as follows: collecting a plurality of sample data, wherein the sample data comprise sample data of a target variety and sample data of a comparison bee species, and each sample data comprise all original morphological indexes obtained by pre-measurement; Performing complement processing on the missing values in the plurality of sample data to obtain a plurality of processed sample data, taking the plurality of processed sample data with a first proportion as a training set, and taking the plurality of processed sample data with a second proportion as a test set; And inputting the training set into an initial XGBoost model for training, and verifying the trained XGBoost model according to the test set to obtain the XGBoost model. Optionally, the outputting, by the pre-trained XGBoost model, at least one target morphological index corresponding to the bee to be tested includes: outputting scores of all original morph