41 results about "Cattle herd" patented technology

The invention discloses a cattle oestrus monitoring method based on multi-camera data fusion, and relates to the technical field of cattle oestrus monitoring. According to the cattle oestrus monitoring method based on multi-camera data fusion, cattle behaviors under complete shielding are visualized through multi-camera dynamic cooperation and skeleton movement deduction; a space-time slice attention mechanism is combined with joint dynamics calculation, target actions are reversely deduced from adjacent cattle body movements, the problem of climbing misjudgment caused by shielding of a traditional visual system is thoroughly solved, and the monitoring reliability in dense cattle herds is remarkably improved; the method comprises the following steps: establishing a differential filtering model of limp gaits and oestrus behaviors; based on acceleration signal time-frequency characteristics and duration determination, a false alarm signal is dynamically suppressed, the inherent defects of a contact sensor in a special physiological state are overcome, and the manual re-checking cost is reduced.

The invention provides an ecological pasture cow breeding and health management method and system based on meteorological data, and relates to the technical field of cow breeding and health management. The method comprises the steps of collecting meteorological data and dairy cow breeding basic data in real time; key meteorological factors influencing cow breeding are extracted; defining a meteorological appropriate threshold value corresponding to each breeding stage; cattle herd screening adaptive to weather is realized; selecting a time period with proper temperature to carry out hybridization operation; formulating meteorological guarantee schemes according to pregnancy stages; and quantifying the influence of meteorological conditions on the breeding core indexes, and optimizing a breeding plan and resource allocation based on an evaluation result. Through a perfect dairy cow breeding management process, the breeding resource utilization efficiency is improved, a traceable and iterable analysis framework is provided for breeding plan optimization, the precise and efficient development of ecological pasture breeding management is promoted, the cost of a pasture is reduced, and meanwhile, the population quality and economic benefits are improved.

The invention belongs to the technical field of image processing, and provides a cattle herd density abnormity early warning method based on multi-camera data fusion, which utilizes a multi-camera data fusion technology to convert cattle herd pixel data into two-dimensional coordinates in a unified coordinate system to form a beef cattle global movement track, and meanwhile, divides a field area according to grid units to form a beef cattle density abnormity early warning system. Calculating the density of each grid, delimiting the degree of congestion according to the density, calculating the characteristic values of the grids, introducing a logistic regression model, monitoring and labeling the grids with high characteristic values, combining the grid data and the global motion track, calculating the risk probability, making early warning arrangement according to the level of the risk probability, and counting the false alarm and missing alarm conditions. And retraining the model to complete closed-loop iteration.

The invention relates to the technical field of multi-modal interaction, in particular to a vision, language and behavior linkage grassland robot inspection and driving integrated method, which comprises the following steps: acquiring a cattle herd image annotation coordinate attitude acquisition distance density input model to generate an inspection target sequence, driving a robot to acquire physical signs, judging abnormity and generating an abnormal individual identifier set, screening a monitoring target, inputting visual questions and answers, outputting a next inspection position coordinate, calculating an inspection path motion instruction set by fusing topographic semantics, completing detection and calculation of deviation, adjusting acousto-optic output, and generating a driving control instruction set. A semantic inspection sequence is generated in combination with text vision, target adaptive updating and real-time response are achieved, sign abnormity is recognized through infrared depth comparison, paths are optimized through image semantic matching, obstacle avoidance and adaptability are improved by fusing terrain semantics and motion parameters, and task continuity and recognition reliability are ensured.

The invention relates to the technical field of forage grass drying evaluation, and particularly discloses a forage grass drying evaluation method and system based on multi-modal fusion. According to the method, a quality file containing indexes such as the moisture content, the dry matter content and the dryness grade is established for each batch of forage grass, the feed intake, the feed intake behavior and the production performance data of cattle herds are synchronously collected in the feeding period, and a multi-modal association data set is established with batch identifiers and timestamps as association keys; and obtaining response results of the ingestion performance and the production performance corresponding to different drying characteristics by adopting correlation analysis, regression analysis and statistical significance test, and dynamically correcting the forage grass drying evaluation standard according to the response results of the ingestion performance and the production performance. According to the method, the forage grass drying characteristic data is associated with the cattle herd feed intake, the feed intake behavior and the production performance data according to batches and time, the multi-mode associated data set is constructed, and the direct corresponding relation between the forage grass drying state and the actual feed intake performance and the growth effect of the cattle herd is achieved.

The utility model discloses a cattle driving device and a pneumatic system thereof. The pneumatic system of the cow driving device comprises an air source assembly, an air control assembly, a reversing valve, a three-way air control valve, at least one air cylinder and a pneumatic motor. A first air outlet of the reversing valve is communicated with a forward rotation port of the pneumatic motor; a second air outlet of the reversing valve is communicated with a reversing port of the pneumatic motor; an air outlet of the three-way pneumatic control valve is communicated with a working port of each air cylinder; the pneumatic motor can be in transmission connection with a frame of the cattle driving device; each cylinder can be in transmission connection with an overhead door of the cattle driving device; and the pneumatic control assembly can control the actuating mechanism of the reversing valve and the actuating mechanism of the three-way pneumatic control valve to act so as to open or close the first passage or the second passage of the reversing valve and open or close the first passage or the second passage of the three-way reversing valve. According to the pneumatic system of the cattle driving device, pure pneumatic control of the cattle driving device is achieved, and the cattle herd safety problem caused by power failure, electric leakage and the like can be effectively avoided.

The invention relates to the technical field of intelligent livestock management, in particular to an intelligent monitoring method and system for rumination behaviors of cattle individuals, and solves the technical problem of low monitoring accuracy in the prior art. The method comprises the following steps: acquiring cattle herd monitoring video data, and performing cattle mouth part identification on image frames in the cattle herd monitoring video data through a target identification model to obtain a cattle mouth area image in each image frame; performing shielding detection on the cattle mouth area image in each image frame, and screening out a target cattle mouth area image from the cattle mouth area images in each image frame according to an obtained shielding proportion; processing the screened target cattle mouth area image through an optical flow algorithm, and generating an optical flow field sequence representing cattle mouth movement; and inputting the optical flow field sequence into a preset classifier, judging whether the cattle is in a rumination behavior, and outputting a monitoring result of the rumination behavior.

The invention belongs to the technical field of grazing intensity evaluation, and discloses a household pasture grazing intensity evaluation method and system based on unmanned aerial vehicle aerial photography and the Internet of Things technology, and the method comprises the steps: obtaining information, such as the spatial distribution position of cattle herds, the number and distribution of livestock herds, and the distance to a livestock pen, in a pasture based on the unmanned aerial vehicle aerial photography technology; based on a GPS Internet of Things tracking technology, obtaining a time sequence cattle herd activity track data set; inverting the spatio-temporal dynamic change condition of the pasture grazing intensity by combining the activity rule of the cattle herd in the pasture and the cattle herd activity track obtained by the unmanned aerial vehicle; and based on the time sequence grazing intensity change condition, calculating the pasture accumulative grazing intensity spatial differentiation condition. Results of the method show that the spatio-temporal dynamic change condition of the pasture scale livestock grazing intensity can be effectively monitored by combining the unmanned aerial vehicle aerial photography technology and the GPS Internet of Things tracking technology, and the research result can provide theoretical basis and scientific support for pasture livestock balance and grassland resource sustainable development.

The utility model belongs to the technical field of agricultural machinery, and particularly relates to a cow pregnancy period feed feeding amount adjusting device which comprises a feeding box. A conveying belt is mounted at the bottom of the feeding box; a feeding opening is formed in the side wall of the feeding box; a rotating rod is rotationally connected to the middle part of the feeding opening; the end part of the rotating rod is fixedly connected with a limiting piece; an inner toothed belt is installed on the inner wall of the limiting piece, and the other end of the inner toothed belt is installed on the conveying belt output shaft. The middle part of the rotating rod is fixedly connected with a plurality of groups of connecting pieces, and the plurality of groups of connecting pieces are positioned in the feeding opening; by arranging the inner toothed belt on the conveying belt, the working synchronism of the feeding port and the conveying belt can be improved, and the problem that the blanking amount of different cattle herds is difficult to flexibly change due to the fact that the size of a hole of an original feeding port is manually adjusted is solved; and the distribution uniformity of the feed transferred on the conveying belt can be improved.

This invention belongs to the field of animal breeding technology, specifically relating to a molecular marker, primer pair, detection reagent, and application closely linked to the beef meat color brightness trait. The nucleotide sequence of the molecular marker is shown in SEQ ID NO.1, and the nucleotide sequence shown in SEQ ID NO.1 exhibits a G / T polymorphism at position 217 bp. This molecular marker can be used to select cattle herds with high meat color brightness traits.

The invention discloses a bezoar quality improving method based on natural pasture environment optimization. According to the method, the internal quality of the calculus bovis is remarkably improved through multi-dimensional cooperative regulation and control. A natural feed basis with balanced nutrition is provided for cattle from soil environment baseline investigation to precise pasture proportioning, and by combining directional cultivation and implantation of genetic engineering strains, the metabolic activity of rumen microorganisms is effectively enhanced, and accumulation of key components for synthesis of calculus bovis is promoted. The linkage application of the intelligent monitoring system and the environment regulation and control device optimizes the living environment of cattle in real time, and cooperates with innovative intervention means such as lavender essential oil and music therapy, so that the stress reaction of cattle herds is remarkably reduced, and the efficient operation of liver metabolic pathways is guaranteed. The directional feeding of the functional lick bricks further enhances the conversion efficiency of the calculus bovis precursor substance, the whole-process data tracing ensures the controllability of the quality formation process, and finally the improvement of the content of effective active ingredients in the calculus bovis and the enhancement of the quality stability are realized.

This invention belongs to the field of animal breeding technology, specifically relating to an SNP molecular marker, primer pair, kit, and application closely linked to the drip loss trait in beef. The nucleotide sequence of the molecular marker is shown in SEQ ID NO.1, where a mutation occurs at position 76 from base C to base T. This molecular marker can be used to select cattle herds with low drip loss traits.

The utility model belongs to the technical field of animal husbandry tool cars, and relates to an animal feeding scattering car which comprises a car body, a baffle, a pulley, a lever and a pull rope, a car hopper is arranged on the car body, a discharge port is formed in the car hopper, the baffle is installed at the discharge port in a lifting opening and closing mode, and the pull rope is connected with the pulley. The pulley is rotationally installed at the upper end of the baffle, a fulcrum of the lever is arranged on the car hopper, one end of the pull rope is connected with the car hopper, and the other end of the pull rope bypasses the pulley and then is connected with the lever. According to the baffle opening and closing mechanism, complex electrical elements such as a hydraulic cylinder and an air cylinder are abandoned, the manufacturing cost is reduced, the reliability and durability of equipment are enhanced, the maintenance cost and the downtime caused by faults of the electrical elements are reduced, and a more economical, efficient and stable feed scattering solution is provided for cattle raising work.

The utility model relates to the technical field of cattle farm environment control equipment, and discloses a cattle farm ventilation cooling device which comprises a shell, an air outlet is formed in one side of the shell, a plurality of rotating shafts are rotatably connected to the inner wall, located at the air outlet, of the shell, and air deflectors are fixedly connected to the outer walls of the rotating shafts; the outer wall of the top of the shell is provided with a wind direction adjusting mechanism for driving the multiple wind deflectors to synchronously rotate, the wind direction adjusting mechanism can drive the multiple wind deflectors to synchronously swing in a reciprocating mode, the swing of the multiple wind deflectors can change the wind inlet direction and angle, air can be more evenly distributed in the cowshed, and the wind inlet efficiency is improved. The problems of unsmooth ventilation and high temperature of partial areas are effectively solved, and a comprehensive and uniform ventilation environment is provided for cattle herds.

The invention discloses a cattle herd raw material ratio measuring and calculating method and platform based on big data, and relates to the technical field of big data analysis. According to the method, firstly, physiological, production and ingestion data, raw material information and environmental parameters of cattle are collected through the Internet of Things, multi-dimensional analysis is carried out after standardization processing, a matching optimization model is constructed in combination with an MEMP algorithm, and a preliminary scheme giving consideration to nutrition balance and productivity is output; then inputting real-time data of the cattle herds, calculating to obtain initial proportioning parameters, and synchronously simulating proportioning combinations and effects under different targets; finally, consistency verification is conducted according to the goodness of fit between the productivity predicted value of each scheme and the productivity of the same type of historical cattle herds, the optimal scheme is screened, collaborative optimization of precise measurement and calculation, dynamic adjustment and production benefit improvement of the cattle herd raw material ratio is achieved, and then the problems that in the prior art, dynamic fluctuation of raw material nutritional ingredients is difficult to accurately capture, and the production efficiency is improved are effectively solved. Therefore, the problem of insufficient adaptability between the matching scheme and the actual production scene is solved.

This invention discloses a method and system for cattle detection and counting based on unmanned aerial vehicles (UAVs), belonging to the field of smart ranch monitoring technology. The core of this invention lies in constructing a lightweight C2Net deep neural network. This network uses an improved DLA-34 as its backbone and embeds two types of receptive field reconstruction convolutional modules guided by fixed priors to enhance cattle contours and suppress background noise, respectively. During the feature upsampling stage, a Converse2D deconvolution operator based on closed-form solutions is introduced to improve reconstruction quality. Furthermore, it innovatively uses a focal inverse distance transform map as a supervision signal, transforming sparse point annotations into response maps with strictly non-overlapping peaks for training. This invention can achieve high-precision detection and stable counting of dense cattle herds while significantly reducing annotation costs, and has the advantages of strong robustness, high computational efficiency, and ease of deployment.

The present invention provides an intelligent control method and system for an electric cattle rumen brush, which relates to the field of agricultural breeding technology. The method comprises deploying an electric rumen brush at a target vibration position inside the cattle rumen, wherein a pH sensor is integrated in the electric rumen brush; collecting historical cattle physiological data, constructing an offset prediction model, and determining a pH target offset; constructing a personalized activation threshold prediction model, and determining a pH personalized activation threshold based on the pH target offset; obtaining a pH measurement value of the pH sensor, constructing a rumen brush vibration frequency adjustment model, and determining a target rumen brush vibration frequency based on the pH personalized activation threshold and the pH measurement value; and remotely controlling the operating state of the electric rumen brush based on the target rumen brush vibration frequency, thereby accurately improving the digestion and absorption rate of the cattle herd, reducing the disease incidence rate of the cattle herd, reducing the frequency of antibiotic use, and improving the operating efficiency and economic benefits of the ranch.

The invention relates to the technical field of animal husbandry digital management, and discloses a digital cattle budget management system and an application method, and the system comprises a data integration module, a budget model construction module for obtaining and integrating data of multiple systems of a pasture, a budget execution processing module, and a result output application module. Multi-source pasture information systems such as a dairy cow yield monitoring system and an oestrus monitoring system are integrated, a data layer, a model layer and an application layer architecture are constructed, automatic access and linkage of basic data, business data and historical operation data of cows are achieved, and data islands are eliminated; a multi-dimensional adjustable budgeting model containing nine core parameters is constructed in combination with the life cycle of the dairy cow, a data consistency verification mechanism is matched, a traditional spreadsheet management mode is replaced, manual operation and data errors are reduced, the budgeting efficiency and the result accuracy are remarkably improved, and the data dimension and granularity requirements of fine management are met.

The invention discloses precise batching and automatic feeding equipment suitable for cattle herds, and relates to the technical field of breeding, the precise batching and automatic feeding equipment comprises a movable bottom frame, a supporting bracket is clamped to the top end of the movable bottom frame, a batching metering and mixing mechanism is arranged at the top end of the supporting bracket, and the batching metering and mixing mechanism comprises a mixing and stirring barrel; the device is scientific and reasonable in structure and safe and convenient to use, an ingredient metering and mixing mechanism is arranged, the bottom of the weighing cylinder is sealed through cooperation of a sealing baffle and an electronic weighing instrument, the weighing cylinder is arranged, the weighing cylinder is arranged in the weighing cylinder, the weighing cylinder is arranged in the weighing cylinder, and the mixing and stirring efficiency is improved. Meanwhile, ingredients entering the weighing cylinder can be weighed, the weight of the ingredients is recorded in real time, the accuracy of the ingredients is improved, errors are reduced, and through cooperation of an electric telescopic rod and a traction pull plate, a sealing baffle and an electronic weighing instrument can be conveniently driven to slide, and sealing of the weighing cylinder is relieved.

The invention discloses a cattle herd behavior recognition method based on a graph neural network. The method comprises the following steps: S1, collecting information of cattle only at different time steps and performing preprocessing; s2, on the basis of cattle time sequence behavior data, dynamically and locally dividing cattle herds at each time step, and constructing a local dynamic graph on the basis of each local sub-group; s3, modeling the dynamic relationship between the cows by using a dynamic graph neural network; s4, performing pooling operation on the local dynamic features to obtain local sub-group features; s5, taking all the local sub-groups as super nodes, and constructing a global dynamic graph; s6, applying a dynamic graph neural network to obtain global dynamic characteristics of each sub-group; s7, fusing the local dynamic features and the global dynamic features through an interlayer information interaction mechanism; and S8, identifying the behavior type of each cattle, and outputting a cattle group behavior identification result and abnormal early warning information. According to the invention, the hierarchical dynamic graph neural network is adopted to realize cattle behavior intelligent identification and abnormity early warning.

The application discloses a kind of equipment and method for identifying cattle herd, it is related to the technical field of identifying cattle herd, including horizontal plate, the upper end of the horizontal plate is fixedly connected with hanging element, the right end of the hanging element is penetrated with threaded hole, the left end of the horizontal plate is fixedly connected with lamp plate, the left end of the lamp plate is fixedly connected with illuminating lamp, the inside of the horizontal plate is provided with sliding slot, and the inside of sliding slot is movably connected with threaded shaft.The height of movable disc, identification block and identification element can be adjusted by the setting of hydraulic cylinder and telescopic rod, so as to adjust according to the height of the animal, so that the application range of the device is wider, and the angle of the movable disc can be adjusted according to the setting of telescopic rod, movable disc, groove, connecting block and annular groove, so that the movable disc can be adjusted without changing the hanging direction of the horizontal plate, which is convenient for identifying the animals passing through the back, and the operation is more convenient.

The utility model provides a cattle raising feed putting mechanism, which relates to the field of feed putting equipment and comprises a feed putting vehicle located on a cattle raising trough, a stirring assembly used for stirring feed is arranged in the feed putting vehicle, a discharge port is formed in the center of the bottom of the feed putting vehicle, and a control door plate is movably mounted in the discharge port. Two opposite moving mechanisms are arranged at the two ends of the bottom of the feed throwing vehicle respectively, an adjusting piece used for adjusting the distance between the two moving mechanisms is further arranged at the bottom of the feed throwing vehicle, sliding rollers are transversely arranged on the two moving mechanisms, and the sliding rollers are attached to the outer side wall of the cattle raising trough; the stirring assembly in the feed throwing vehicle scatters and stirs cattle raising feed, eating and absorption of cattle herds are better facilitated, the control door plate is pulled to enable the discharge port to be in an open state, feed is added along with movement of the feed throwing vehicle, and the feed adding mode is more flexible and convenient; and the labor intensity of workers is reduced to a certain extent.

The invention discloses a dairy cow rumen butyric acid concentration related SNP molecular marker, a detection primer pair and application thereof, and belongs to the technical field of molecular biology and genetic breeding. The SNP site is located in a fourth intron of an SPINK5 gene, the number of the SNP site is Bovine HD0700017637, and the SNP site is subjected to Agt; the site is subjected to basic group mutation of G, and the site has three genotypes, namely AA, AG and GG. When the genotype of the SNP site is AG and GG, the dairy cow individual has higher rumen butyric acid concentration. The SNP molecular marker can be used for screening and identifying dairy cows with high rumen butyric acid level, and then dairy cow individuals with high rumen butyric acid concentration are selected for cattle herd optimization and assistant breeding.

The invention discloses a cattle herd benefit measurement and calculation process evaluation method oriented to a cattle herd raw material matching task, and relates to the technical field of cattle herd benefit measurement and calculation management. The method is realized through the following steps: quantifying the real-time demand condition of cattle feeding data measurement and calculation to reflect the transmission condition of cattle feeding related information, and determining cattle benefit information and an optimal cattle raw material ratio according to the real-time demand of cattle feeding data measurement and calculation.

The application belongs to the technical field of biology and particularly relates to a bovine viral diarrhea virus NS2 truncated protein and application thereof in BVDV antibody detection. The application provides a truncated protein of BVDV NS2 protein. The truncated protein is soluble expressed after being fused with glutathione S transferase (GST) and horseradish peroxidase (HRP) respectively, has good reactivity with BVDV antibody positive serum, does not react with BVDV antibody negative serum, and can be used for BVDV antibody detection. Furthermore, the application constructs a double antigen sandwich ELISA method for BVDV antibody detection by taking the fusion protein of GST and the truncated protein as a coating antigen and taking the fusion protein of HRP and the truncated protein as a detection antigen. The method has the technical advantages of high sensitivity, good specificity, simple operation, easy popularization, low cost and the like, and can be applied to screening of BVDV antibodies in a cattle herd.

The invention relates to the technical field of livestock breeding intellectualization, in particular to a cattle body shape automatic scoring method and device based on 3D vision, equipment and a medium, and the method comprises the steps: obtaining cattle body shape data based on the 3D vision; inputting the body type data of the cattle into a deep learning model to obtain key anatomical parts of the cattle, and calculating feature parameters of fat covering thickness, lumbar vertebra corner angle acutance and pelvic depression depth according to the key parts; inputting the characteristic parameters into a random forest regression model, outputting a body condition score value, and obtaining a scoring result of the body condition of the cattle; the grading result is associated with the individual ear tag number and synchronized to a management system, the management system triggers grouping, feeding or veterinary intervention instructions according to the grading result, and precise management of the cattle herd is achieved. Therefore, the problems that in the prior art, subjective errors are large, efficiency is low, data are isolated, the camera coverage range is not comprehensive, the scoring precision is prone to being affected by animal postures and illumination conditions, and it is difficult to adapt to the actual scene of a pasture are solved.

The intelligent yak saddle comprises a drive-by-wire box, the bottom of the drive-by-wire box is fixedly connected with a back clamping plate, the bottom face of the back clamping plate is an arc-shaped face with one high side and one low side, the low arc-shaped face is attached to the front side of the back of a yak and protrudes, and the high arc-shaped face is located at the low-lying position of the rear side of the back of the yak. The two sides of the back clamping plate are movably connected with the outer side clamping plates, the attaching angle can be automatically adjusted according to the body type difference of yaks, the non-intrusive self-adaptive structure effectively disperses the load bearing pressure while guaranteeing the connection stability of the device, the technical problems that animal skin is abraded and the device falls off due to long-term field wearing are solved, and the service life of the device is prolonged. Meanwhile, in the use process of the cattle saddle, a camera, a main controller and other components are matched, long-time real-time remote monitoring on cattle herds is achieved, the cattle herds with intelligent ear tags are matched, the positions of the cattle herds and the physiological state of yaks are monitored in a centralized mode, and the economic losses caused by missing of the cattle herds and diseases are reduced.

The invention belongs to the technical field of intelligent agriculture, and particularly relates to a cattle herd abnormal behavior real-time recognition method and system based on deep learning, and the method comprises the steps: collecting behavior recognition data of a to-be-recognized cattle herd, and carrying out the preprocessing; constructing a recognition model of the abnormal behaviors of the cattle herds on the basis of an improved transformer; and on the basis of the preprocessed behavior recognition data, comparing differences between individuals and groups and comparing daily data and weekly data of the individuals of the to-be-recognized cattle herds by using the recognition model of the abnormal behaviors of the cattle herds, so that real-time recognition of the abnormal behaviors of the cattle herds is realized. According to the method, through Gaussian priori sparse transform + multi-dimensional cross validation, accurate recognition and abnormal early warning of cattle herd behaviors are realized, and a set of complete solution which is low in power consumption, high in precision and easy to land is provided for a smart pasture.

According to the method for detecting the early pregnancy of the cattle by using the metabonomics technology, tiny metabolic changes in bovine serum can be detected through the metabonomics technology, the pregnancy state can be accurately judged even in the early pregnancy stage, and the detection sensitivity is improved. The differential metabolites screened out through multivariate statistical analysis and a machine learning algorithm have high specificity, pregnant cattle and non-pregnant cattle can be effectively distinguished, and misdiagnosis and missed diagnosis are reduced. The sample collection and pretreatment processes are relatively simple, complex equipment and professional skills are not needed, and the method is suitable for rapid detection of large-scale farms. Detection can be carried out 18-25 days after cattle hybridization, the pregnancy state can be judged earlier than that of a traditional method, and more timely information is provided for breeding management of cattle herds.

The invention discloses a cattle herd structure optimization method and system based on multi-source data fusion, and relates to the technical field of intelligent breeding data processing. The cattle herd structure optimization method based on multi-source data fusion comprises the following steps of data preprocessing, cattle herd efficiency evaluation and cattle herd structure optimization. According to the method, data preprocessing is carried out on the obtained cattle herd multi-source data sets of all pasture cattle to improve data quality, and then the cattle herd multi-source data sets after data preprocessing are screened to obtain cattle herd efficiency characteristic data reflecting cattle herd production performance; the cattle herd performance characteristic data is input into a set performance evaluation model to obtain an individual performance score for comprehensively reflecting the performance of the cattle in each pasture, and finally the cattle in the pasture are eliminated based on the individual performance score of the cattle in each pasture to realize the optimization of the cattle herd structure, so that the reliability of the optimization decision of the cattle herd structure is improved. The problem of low reliability of a cattle herd structure optimization decision in the prior art is solved.