CN116738185B - An AI algorithm construction method for smart farming - Google Patents

Abstract

The application discloses an AI algorithm construction method for intelligent cultivation, which comprises the following steps: determining a first cultivation parameter and a first AI algorithm corresponding to the first cultivation parameter; determining a second cultivation parameter and a second AI algorithm corresponding to the second cultivation parameter; determining a third cultivation parameter and a third AI algorithm corresponding to the third cultivation parameter; the third AI algorithm is configured to determine the third cultivation parameter based on the first cultivation parameter and/or the second cultivation parameter, the third cultivation parameter being a cultivation parameter based on the environmental dimension and/or the artificial dimension; and constructing a target AI algorithm based on the first AI algorithm, the second AI algorithm, the third AI algorithm, the target culture variety and the target parameter relationship. The method can improve the adaptation degree of the AI algorithm and intelligent cultivation, and improve the accuracy and applicability of the AI algorithm.

Description

An AI algorithm construction method for smart farming

Technical field

This application is about the field of smart farming technology, especially about an AI algorithm construction method for smart farming.

Background technique

With the development of smart farming, more and more farming industries have integrated smart farming. In the breeding industry, smart breeding can use AI (Artificial Intelligence, artificial intelligence) algorithms to analyze breeding data; or predict breeding data; or predict breeding parameters, etc., to save a lot of labor costs.

At present, in the intelligent breeding technology involving AI algorithms, special AI algorithms are usually developed for a breeding species or a specific breeding environment; that is, an AI algorithm is targeted at a breeding scene, and the AI algorithm provides a large number of The algorithm chosen among the AI algorithms. In this way, the adaptability of the AI algorithm to smart farming scenarios is not very high, resulting in poor accuracy and applicability of the AI algorithm.

Contents of the invention

The purpose of this application is to provide an AI algorithm construction method for smart farming, which can improve the adaptability of the AI algorithm to smart farming, as well as improve the accuracy and applicability of the AI algorithm.

To achieve the above purpose, embodiments of the present application provide an AI algorithm construction method for smart farming, which includes: determining a first farming parameter and a first AI algorithm corresponding to the first farming parameter; the first AI The algorithm is used to determine the first breeding parameter based on the target breeding data, and the first breeding parameter is the breeding parameter of the environmental dimension; determine the second breeding parameter and the second AI algorithm corresponding to the second breeding parameter; the third The second AI algorithm is used to determine the second breeding parameter based on the target breeding data, and the second breeding parameter is an artificial dimension breeding parameter; determine the third breeding parameter and the third AI algorithm corresponding to the third breeding parameter ; The third AI algorithm is used to determine the third breeding parameter based on the first breeding parameter and/or the second breeding parameter, and the third breeding parameter is based on the environmental dimension and/or the Artificial dimension breeding parameters; based on the first AI algorithm, the second AI algorithm, the third AI algorithm, the target breeding species and the target parameter relationship, a target AI algorithm is constructed; the target AI algorithm is used to determine the The breeding parameters of the target breeding species are described, and the relationship between the target parameters is expressed as: ;Among them,/> Represents the first breeding parameter,/> Represents the second breeding parameter,/> Represents the third breeding parameter,/> Represents the first weight,/> represents the second weight, Represents the first influence value,/> Represents the second influence value,/> Represents the third influence value,/> Represents the default influence value.

In a possible implementation, the determination of the first breeding parameter and the first AI algorithm corresponding to the first breeding parameter includes: obtaining a plurality of first preset breeding parameters; the plurality of first preset The breeding parameters are all breeding parameters of the environmental dimension; determine the direct impact value of the plurality of first preset breeding parameters on the target breeding species; determine the impact of the plurality of first preset breeding parameters on the target breeding species. Indirect influence value; the indirect influence value is determined based on the direct influence value of the plurality of first preset breeding parameters on the first associated breeding species of the target breeding species; the relationship between the breeding environment of the first associated breeding species and the The similarity between the breeding environments of the target cultured species is greater than the first preset similarity; based on the direct impact value on the target cultured species and the indirect impact value on the target cultured species, determine the plurality of first The integrated influence value of the preset breeding parameters; based on the integrated influence value, the first breeding parameter is determined from the plurality of first preset breeding parameters.

In a possible implementation, determining the first breeding parameter and the first AI algorithm corresponding to the first breeding parameter includes: obtaining a plurality of first preset AI algorithms; the plurality of first preset AI algorithms AI algorithms respectively correspond to different breeding species; the first AI algorithm is determined from the plurality of first preset AI algorithms; wherein the first AI algorithm is among the plurality of first preset AI algorithms. , the number of corresponding breeding species is the highest, and the breeding environment of the corresponding breeding species meets the preset breeding environment conditions.

In a possible implementation, the determination of the second breeding parameter and the second AI algorithm corresponding to the second breeding parameter includes: obtaining a plurality of second preset breeding parameters; the plurality of second preset The breeding parameters are all artificial dimension breeding parameters; determine the direct impact value of the plurality of second preset breeding parameters on the target breeding species; determine the impact of the plurality of second preset breeding parameters on the target breeding species. Indirect influence value; the indirect influence value is determined based on the direct influence value of the plurality of second preset breeding parameters on the second associated breeding species of the target breeding species; the artificial breeding conditions of the second associated breeding species are The similarity between the artificial breeding conditions of the target cultured species is greater than the second preset similarity; based on the direct impact value on the target cultured species and the indirect impact value on the target cultured species, the plurality of The integrated influence value of the second preset breeding parameter; based on the integrated influence value, the second breeding parameter is determined from the plurality of second preset breeding parameters.

In a possible implementation, determining the second breeding parameter and the second AI algorithm corresponding to the second breeding parameter includes: obtaining a plurality of second preset AI algorithms; the plurality of second preset AI algorithms AI algorithms respectively correspond to different breeding species; the second AI algorithm is determined from the plurality of second preset AI algorithms; wherein the second AI algorithm is among the plurality of second preset AI algorithms. , the number of the corresponding breeding species is the highest, and the artificial breeding conditions of the corresponding breeding species meet the preset artificial breeding conditions.

In a possible implementation, the determination of the third breeding parameter and the third AI algorithm corresponding to the third breeding parameter includes: obtaining a plurality of third preset breeding parameters; the plurality of third preset breeding parameters The breeding parameters are breeding parameters determined based on the first preset breeding parameters and/or the second preset breeding parameters; determining the direct impact value of the plurality of third preset breeding parameters on the target breeding species; Determine the indirect influence value of the plurality of third preset breeding parameters on the target breeding species; the indirect influence value is based on the third associated breeding species of the target breeding species on the plurality of first preset breeding parameters. The direct influence value is determined; the similarity between the breeding environment of the third related breeding species and the breeding environment of the target breeding species is greater than the first preset similarity, and/or the artificial breeding of the third related species The similarity between the conditions and the artificial breeding conditions of the target cultured species is greater than the second preset similarity; based on the direct impact value on the target cultured species and the indirect impact value on the target cultured species, the determination of the The integrated influence value of a plurality of third preset breeding parameters; based on the integrated influence value, the third breeding parameter is determined from the plurality of third preset breeding parameters.

In a possible implementation, determining the third breeding parameter and the third AI algorithm corresponding to the third breeding parameter includes: obtaining a plurality of third preset AI algorithms; the plurality of third preset AI algorithms AI algorithms respectively correspond to different breeding species; the third AI algorithm is determined from the plurality of third preset AI algorithms; wherein the third AI algorithm is among the plurality of third preset AI algorithms. , the number of the corresponding breeding species is the highest, the breeding environment of the corresponding breeding species meets the preset breeding environment conditions, and/or the artificial breeding conditions of the corresponding breeding species meet the preset artificial breeding conditions.

In a possible implementation, constructing a target AI algorithm based on the first AI algorithm, the second AI algorithm, the third AI algorithm, target breeding species, and target parameter relationships includes: based on the The target parameter relationship is used to determine the connection relationship between the first AI algorithm, the second AI algorithm and the third AI algorithm; the connection relationship is used to indicate the relationship between the input and output of each AI algorithm; based on the According to the connection relationship described above, the first AI algorithm, the second AI algorithm and the third AI algorithm are connected to construct an initial target AI algorithm; based on the preset AI algorithm corresponding to the target breeding species and the Initial target AI algorithm, determine the target AI algorithm.

In a possible implementation, determining the target AI algorithm based on the preset AI algorithm corresponding to the target breeding species and the initial target AI algorithm includes: determining the preset AI algorithm corresponding to Whether the input parameters and output parameters include at least one of the first breeding parameter, the second breeding parameter and the third breeding parameter; if so, determine whether the preset AI algorithm is consistent with the third breeding parameter. There is an association relationship between at least one AI algorithm among an AI algorithm, the second AI algorithm and the third AI algorithm; if yes, determine the target AI algorithm based on the initial target AI algorithm; if not, The initial target AI algorithm is adjusted based on the preset AI algorithm, and the target AI algorithm is determined based on the adjusted AI algorithm.

In a possible implementation, the AI algorithm construction method for smart farming further includes: obtaining sample farming data; determining first sample farming parameters based on the sample farming data and the target AI algorithm; The sample farming data and the preset smart farming model determine the second sample farming parameters; the model algorithm corresponding to the preset smart farming model is different from the target AI algorithm; based on the first sample farming parameters, The second sample breeding parameters and the real breeding parameters corresponding to the sample breeding parameters optimize the target AI algorithm and the preset smart breeding model.

Compared with the existing technology, the AI algorithm construction method for smart farming provided by the embodiments of this application can, on the one hand, determine corresponding AI algorithms for different types of farming parameters, which can improve the diversity of AI algorithms. On the basis of higher diversity of algorithms, both the accuracy of the AI algorithm and its adaptability to smart scenarios have been improved accordingly. On the other hand, based on multiple AI algorithms, combined with the relationship between the input objects corresponding to each AI algorithm, the target AI algorithm is integrated and constructed, so that the fitness of the built target AI algorithm and the target breeding species is improved. Furthermore, on the basis of improved adaptability and accuracy, the applicability of AI algorithms has also been improved accordingly. Therefore, the AI algorithm construction method for smart farming can improve the adaptability of the AI algorithm to smart farming, as well as improve the accuracy and applicability of the AI algorithm.

Description of the drawings

Figure 1 is a flow chart of an AI algorithm construction method for smart farming according to an embodiment of the present application;

Figure 2 is a schematic diagram of the connection relationship of the AI algorithm according to an embodiment of the present application;

Figure 3 is a schematic structural diagram of an AI algorithm construction device for smart farming according to an embodiment of the present application;

Figure 4 is a schematic structural diagram of a terminal device according to an embodiment of the present application.

Detailed ways

The specific embodiments of the present application will be described in detail below with reference to the accompanying drawings, but it should be understood that the protection scope of the present application is not limited by the specific embodiments.

Unless expressly stated otherwise, throughout the specification and claims, the term "comprises" or its variations such as "comprises" or "comprising" will be understood to include the stated elements or components, and to Other elements or other components are not excluded.

The technical solutions provided by the embodiments of this application can be applied to smart farming scenarios. In these smart farming scenarios, AI (Artificial Intelligence, artificial intelligence) algorithms are used to analyze farming data; or to predict farming data; or to predict farming parameters, etc. , to save a lot of labor costs.

And, in different smart farming scenarios, the smart farming items involved may change accordingly due to differences in farming environments, farming species, etc. Therefore, different AI algorithms can be designed for different smart farming scenarios so that these AI algorithms can be applied to these scenarios.

At present, a large number of AI algorithms have been produced for these smart farming scenarios. These AI algorithms are usually targeted at a farming species; or a special AI algorithm is developed for a specific farming environment; that is, one AI algorithm is targeted at a farming scenario.

As a result, these large numbers of AI algorithms cannot be effectively applied, making the AI algorithm not very suitable for smart farming scenarios, and thus the accuracy and applicability of the AI algorithm are poor.

Based on this, embodiments of this application provide an AI algorithm construction method for smart farming. On the one hand, determining corresponding AI algorithms for different types of farming parameters can improve the diversity of AI algorithms. On the basis of higher diversity, both the accuracy of the AI algorithm and its adaptability to smart scenarios have been improved accordingly. On the other hand, based on multiple AI algorithms, combined with the relationship between the input objects corresponding to each AI algorithm, the target AI algorithm is integrated and constructed, so that the fitness of the built target AI algorithm and the target breeding species is improved.

Next, please refer to Figure 1, which is a flow chart of an AI algorithm construction method for smart farming provided by an embodiment of the present application. The construction method includes:

Step 101: Determine the first breeding parameter and the first AI algorithm corresponding to the first breeding parameter. Among them, the first AI algorithm is used to determine the first breeding parameter based on the target breeding data, and the first breeding parameter is the breeding parameter of the environmental dimension.

In some embodiments, the target breeding data can be regarded as the input data of the first AI algorithm, and the specific values of the first breeding parameters can be regarded as the output data of the first AI algorithm. Therefore, the first AI algorithm can be based on the target breeding data. Determine the value of the first breeding parameter.

In some embodiments, the first breeding parameter is a breeding parameter in an environmental dimension, and the environmental dimension can be understood as the influence of environmental factors. For example: breeding temperature, breeding humidity, etc. can be regarded as breeding parameters in the environmental dimension.

In some embodiments, the target breeding data may be the breeding data of the expected target breeding species, such as: breeding quantity, breeding cycle, etc.

As an optional implementation manner, step 101 includes: obtaining a plurality of first preset breeding parameters; the plurality of first preset breeding parameters are all breeding parameters of the environmental dimension; determining the relationship between the plurality of first preset breeding parameters and the target The direct impact value of the cultured species; determine the indirect impact value of multiple first preset culture parameters on the target culture species; the indirect impact value is based on the direct impact of the multiple first preset culture parameters on the first associated culture species of the target culture species. The value is determined; the similarity between the breeding environment of the first associated breeding species and the breeding environment of the target breeding species is greater than the first preset similarity; based on the direct impact value on the target breeding species and the indirect impact value on the target breeding species, Determine the integrated influence value of the plurality of first preset breeding parameters; determine the first breeding parameter from the plurality of first preset breeding parameters based on the integrated influence value.

In some embodiments, determining the first breeding parameter can be understood as determining the breeding parameter item corresponding to the target breeding species from the breeding parameter items in multiple optional environmental dimensions.

In some embodiments, the plurality of first preset breeding parameters may be multiple breeding parameters under preset environmental dimensions. These multiple breeding parameters may or may not have a relationship with the target breeding species, so further steps are required. Determine whether it can be applied to the target cultured species.

In some embodiments, the direct impact value of the plurality of first preset breeding parameters on the target breeding species can be determined based on the historical breeding parameters corresponding to the target breeding species. If the historical breeding parameters corresponding to the target breeding species include a certain first preset breeding parameter, then the first preset breeding parameter can have an impact value (for example, 90, up to 100) on the target breeding species. If the historical breeding parameters corresponding to the target breeding species include related breeding parameters of a certain first preset breeding parameter, then the first preset breeding parameter can have an impact value on the target breeding species (for example, 70, up to 100). . If none of the above conditions apply, the corresponding impact value can be between 0 and 50, with the highest value also being 100.

In some embodiments, the similarity between the breeding environment of the first associated cultured species and the culture environment of the target cultured species is greater than the first preset similarity; wherein, the first preset similarity can be set according to different application scenarios. , for example, it can be ninety percent. For example, the similarity between fish pond culture environment and freshwater culture environment can reach 80%.

In some embodiments, the similarity between breeding environments can be determined based on data such as humidity, temperature, and main substances of the breeding environment.

In some embodiments, based on the direct impact value and indirect impact value on the target culture species, weighted integration can be performed according to the preset weight value, and the determined impact value is the integrated impact value.

Further, based on the integrated influence value, the first preset breeding parameter whose integrated influence value is greater than the preset influence value can be determined as the first breeding parameter. Among them, the preset influence value can be set according to different application scenarios, for example, it can be 85 and the highest is 100.

As an optional implementation, step 101 also includes: obtaining a plurality of first preset AI algorithms; the plurality of first preset AI algorithms respectively corresponding to different breeding species; and determining from the plurality of first preset AI algorithms. A first AI algorithm is generated; among the plurality of first preset AI algorithms, the first AI algorithm corresponds to the highest number of breeding species, and the breeding environment of the corresponding breeding species meets the preset breeding environment conditions.

In some embodiments, the plurality of first preset AI algorithms can be understood as algorithms corresponding to different breeding species and applicable to the determination of the first breeding parameters, which are algorithms that have been configured in advance.

In some embodiments, each first preset AI algorithm may correspond to one breeding species or may correspond to multiple breeding species, which is not limited here.

Further, a first AI algorithm is determined from a plurality of first preset AI algorithms. In some embodiments, the AI algorithm with the highest number of corresponding breeding species can be determined first, and then the AI algorithm whose breeding environment of the breeding species meets the preset breeding environment conditions can be determined from these AI algorithms.

It can be understood that the preset breeding environment conditions can be the breeding environment conditions corresponding to the target breeding species, and the relevant data of the minimum required breeding environment required by the target breeding species can be defined in the conditions.

Step 102: Determine the second breeding parameter and the second AI algorithm corresponding to the second breeding parameter. Among them, the second AI algorithm is used to determine the second breeding parameter based on the target breeding data, and the second breeding parameter is an artificial dimension breeding parameter.

In some embodiments, the target breeding data can be regarded as the input data of the second AI algorithm, and the specific values of the second breeding parameters can be regarded as the output data of the second AI algorithm. Therefore, the second AI algorithm can be based on the target breeding data. Determine the value of the first breeding parameter.

In some embodiments, the second breeding parameter is a breeding parameter with an artificial dimension, and the artificial dimension can be understood as the influence of artificial factors. For example: artificial feeding time, number of artificial fishing, number of artificial fishing, etc. can all be regarded as breeding parameters in the artificial dimension.

As an optional implementation, step 102 includes: obtaining a plurality of second preset breeding parameters; the plurality of second preset breeding parameters are all artificial-dimensional breeding parameters; and determining the relationship between the plurality of second preset breeding parameters and the target The direct impact value of the cultured species; determine the indirect impact value of multiple second preset culture parameters on the target culture species; the indirect impact value is based on the direct impact of the multiple second preset culture parameters on the second associated culture species of the target culture species. The value is determined; the similarity between the artificial breeding conditions of the second associated breeding species and the artificial breeding conditions of the target breeding species is greater than the second preset similarity; based on the direct impact value on the target breeding species and the indirect impact on the target breeding species value to determine the integrated influence value of the plurality of second preset breeding parameters; based on the integrated influence value, determine the second breeding parameter from the plurality of second preset breeding parameters.

In some embodiments, determining the second breeding parameter can be understood as determining the breeding parameter item corresponding to the target breeding species from the breeding parameter items under multiple optional artificial dimensions.

In some embodiments, the plurality of second preset breeding parameters may be multiple breeding parameters under preset artificial dimensions. These multiple breeding parameters may or may not have a relationship with the target breeding species, so further steps are required. Determine whether it can be applied to the target cultured species.

In some embodiments, the direct impact value of the plurality of second preset breeding parameters on the target breeding species can be determined based on the historical breeding parameters corresponding to the target breeding species. If the historical breeding parameters corresponding to the target breeding species include a certain second preset breeding parameter, then the second preset breeding parameter can have an impact value (for example, 90, up to 100) on the target breeding species. If the historical breeding parameters corresponding to the target breeding species include related breeding parameters of a second preset breeding parameter, then the second preset breeding parameter can have an impact value on the target breeding species (for example, 70, up to 100). . If none of the above conditions apply, the corresponding impact value can be between 0 and 50, with the highest value also being 100.

In some embodiments, the similarity between the artificial breeding conditions of the second associated breeding species and the artificial breeding conditions of the target breeding species is greater than the second preset similarity; wherein the second preset similarity can be based on different application scenarios. Make a setting, maybe ninety percent, for example.

In some embodiments, the similarity between artificial breeding conditions can be determined based on various data involved in the artificial breeding conditions.

In some embodiments, based on the direct impact value and indirect impact value on the target culture species, weighted integration can be performed according to the preset weight value, and the determined impact value is the integrated impact value.

Further, based on the integrated influence value, the second preset breeding parameter whose integrated influence value is greater than the preset influence value can be determined as the second breeding parameter. Among them, the preset influence value can be set according to different application scenarios, for example, it can be 85 and the highest is 100.

As an optional implementation, the steps further include: obtaining a plurality of second preset AI algorithms; the plurality of second preset AI algorithms respectively corresponding to different breeding species; and determining from the plurality of second preset AI algorithms The second AI algorithm; among the plurality of second preset AI algorithms, the second AI algorithm has the highest number of corresponding breeding species, and the artificial breeding conditions of the corresponding breeding species meet the preset artificial breeding conditions.

In some embodiments, multiple second preset AI algorithms can be understood as algorithms that correspond to different breeding species and are applicable to the determination of second breeding parameters, and are algorithms that have been configured in advance.

In some embodiments, each second preset AI algorithm may correspond to one breeding species or may correspond to multiple breeding species, which is not limited here.

Further, a second AI algorithm is determined from a plurality of second preset AI algorithms. In some embodiments, the AI algorithm with the highest number of corresponding cultured species can be first determined, and then the AI algorithm whose artificial breeding conditions of the cultured species meet the preset artificial breeding conditions can be determined from these AI algorithms.

It can be understood that the preset artificial breeding conditions can be artificial breeding conditions corresponding to the target breeding species, and the relevant data of the minimum artificial breeding conditions required for the target breeding species can be defined in the conditions.

Step 103: Determine the third breeding parameter and the third AI algorithm corresponding to the third breeding parameter. Wherein, the third AI algorithm is used to determine the third breeding parameter based on the first breeding parameter and/or the second breeding parameter, and the third breeding parameter is a breeding parameter based on the environmental dimension and/or the artificial dimension.

In some embodiments, the first breeding parameter and/or the second breeding parameter can be regarded as the input data of the third AI algorithm, and the specific value of the third breeding parameter can be regarded as the output data of the third AI algorithm, so that the third AI algorithm The three AI algorithms may determine the value of the third breeding parameter based on the second breeding parameter and/or the second breeding parameter.

In some embodiments, the third breeding parameter is a breeding parameter based on environmental dimensions and/or artificial dimensions. Based on environmental dimensions and/or artificial dimensions, it can be understood that it is affected by two dimensions; or is affected by only one of the dimensions. For example: the breeding cycle is a breeding parameter based on the environmental dimension and the artificial dimension.

Furthermore, the third breeding parameter may need to be determined using the first breeding parameter and the second breeding parameter.

As an optional implementation, step 103 includes: obtaining a plurality of third preset breeding parameters; the plurality of third preset breeding parameters are determined based on the first preset breeding parameters and/or the second preset breeding parameters. Breeding parameters; determine the direct impact value of multiple third preset breeding parameters on the target breeding species; determine the indirect impact value of multiple third preset breeding parameters on the target breeding species; the indirect impact value is based on multiple first preset breeding The direct impact value of the parameters on the third associated breeding species of the target breeding species is determined; the similarity between the breeding environment of the third associated breeding species and the breeding environment of the target breeding species is greater than the first preset similarity, and/or the third The similarity between the artificial breeding conditions of the related species and the artificial breeding conditions of the target breeding species is greater than the second preset similarity; based on the direct impact value on the target breeding species and the indirect impact value on the target breeding species, multiple third The integrated influence value of the three preset breeding parameters; based on the integrated influence value, the third breeding parameter is determined from a plurality of third preset breeding parameters.

In some embodiments, determining the third breeding parameter can be understood as determining the breeding parameter item corresponding to the target breeding species from a plurality of optional breeding parameter items based on environmental dimensions and/or artificial dimensions.

In some embodiments, the plurality of third preset breeding parameters are breeding parameters determined based on the first preset breeding parameters and/or the second preset breeding parameters; the environmental dimensions corresponding to the first preset breeding parameters and the second preset breeding parameters may be used. The dimensional correlation between the artificial dimensions corresponding to the preset breeding parameters determines a plurality of third preset breeding parameters.

In some embodiments, the direct impact value of the plurality of third preset breeding parameters on the target breeding species can be determined based on the historical breeding parameters corresponding to the target breeding species. If the historical breeding parameters corresponding to the target breeding species include a third preset breeding parameter, then the third preset breeding parameter can have an impact value (for example, 90, up to 100) on the target breeding species. If the historical breeding parameters corresponding to the target breeding species include related breeding parameters of a third preset breeding parameter, then the third preset breeding parameter can have an impact value on the target breeding species (for example, 70, up to 100). . If none of the above conditions apply, the corresponding impact value can be between 0 and 50, with the highest value also being 100.

In some embodiments, the similarity between the breeding environment of the third associated breeding species and the breeding environment of the target breeding species is greater than the first preset similarity, and/or the artificial breeding conditions of the third associated species are similar to those of the target breeding species. The similarity between artificial breeding conditions is greater than the second preset similarity; wherein, the first preset similarity and the second preset similarity can be introduced with reference to the foregoing embodiments.

In some embodiments, based on the direct impact value and indirect impact value on the target culture species, weighted integration can be performed according to the preset weight value, and the determined impact value is the integrated impact value.

Further, based on the integrated influence value, the third preset breeding parameter whose integrated influence value is greater than the preset influence value can be determined as the third breeding parameter. Among them, the preset influence value can be set according to different application scenarios, for example, it can be 85 and the highest is 100.

As an optional implementation, step 103 also includes: obtaining a plurality of third preset AI algorithms; the plurality of third preset AI algorithms respectively corresponding to different breeding species; and determining from the plurality of third preset AI algorithms. A third AI algorithm is generated; among the plurality of third preset AI algorithms, the third AI algorithm has the highest number of corresponding breeding species, and the breeding environment of the corresponding breeding species meets the preset breeding environment conditions, and/or The artificial breeding conditions of the corresponding breeding species meet the preset artificial breeding conditions.

In some embodiments, the plurality of third preset AI algorithms can be understood as algorithms that correspond to different breeding species and are applicable to the determination of the third breeding parameters, and are algorithms that have been configured in advance.

In some embodiments, each third preset AI algorithm may correspond to one breeding species or may correspond to multiple breeding species, which is not limited here.

Further, a third AI algorithm is determined from a plurality of third preset AI algorithms. In some embodiments, the AI algorithm with the highest number of corresponding breeding species can be determined first, and then the AI algorithm whose artificial breeding conditions of the breeding species meet the preset artificial breeding conditions can be determined from these AI algorithms, and finally from these AI algorithms The AI algorithm determines that the breeding environment of the breeding species meets the preset breeding environment conditions.

In the embodiments of this application, the first preset AI algorithm, the second preset AI algorithm, and the third preset AI algorithm can be algorithms selected from mature algorithms in the field; they can also be based on specific application scenarios. , the artificial intelligence algorithm designed.

For example, for the first preset AI algorithm, AI algorithm training, parameter adjustment, optimization, etc. can be performed separately for multiple breeding species to obtain the first preset AI algorithm that meets the conditions. That is to say, the above three algorithms, their forms and acquisition methods are not limited in the embodiments of this application.

Step 104: Construct a target AI algorithm based on the relationship between the first AI algorithm, the second AI algorithm, the third AI algorithm, the target breeding species and the target parameters. Among them, the target parameter relationship is the relationship between the first breeding parameter, the second breeding parameter and the third breeding parameter, and the target AI algorithm is used to determine the breeding parameters of the target breeding species.

In some embodiments, the target parameter relationship is expressed as: ;Among them,/> Represents the first breeding parameter,/> Represents the second breeding parameter,/> Represents the third breeding parameter,/> Represents the first weight,/> Represents the second weight,/> Represents the first influence value,/> Represents the second influence value,/> Represents the third influence value,/> Represents the default influence value.

In some embodiments, the first weight and the second weight can be configured in combination with different application scenarios, and represent the correlation between the first breeding parameter and the second breeding parameter and the third breeding parameter respectively. The greater the correlation. The stronger, the higher the corresponding weight value.

In some embodiments, the first influence value, the second influence value and the third influence value respectively represent the influence of each breeding parameter on the target breeding species. The higher the influence, the greater the influence value of the response. The influence value It can be determined based on historical data and is not limited here.

In some embodiments, the preset influence value can be configured in combination with different application scenarios, representing an influence value that has a greater influence on the target breeding species.

Therefore, based on the relationship between the three breeding parameters, the corresponding connection relationship can be determined. For example, if the third breeding parameter is determined based on the first breeding parameter and the second breeding parameter, then the third AI algorithm has a connection relationship with the first AI algorithm and the second AI algorithm; and, between the input and the output time, and also add corresponding weight value restrictions.

If the third breeding parameter is determined based on the first breeding parameter or the second breeding parameter, then the third AI algorithm has a connection relationship with the first AI algorithm or the second AI algorithm; and between the input and the output, there is also The corresponding weight value is limited.

Therefore, as an optional implementation, step 104 includes: determining the connection relationship of the first AI algorithm, the second AI algorithm and the third AI algorithm based on the target parameter relationship; the connection relationship is used to indicate the input of each AI algorithm. and the relationship between the output; based on the connection relationship, connect the first AI algorithm, the second AI algorithm and the third AI algorithm to construct the initial target AI algorithm; based on the preset AI algorithm corresponding to the target breeding species and the initial target AI algorithm, determine the target AI algorithm.

As shown in Figure 2, an algorithm connection relationship is provided in the embodiment of the present application. In this connection relationship, the input data of the first AI algorithm and the second AI algorithm are the same, and the input data of the first AI algorithm and the second AI algorithm are the same. The outputs are connected to the input of the third AI algorithm.

It can be understood that in different application scenarios, based on different target parameter relationships, other algorithm connection relationships can also be used, which are not limited here.

Therefore, by connecting the three AI algorithms based on this connection relationship, the initial target AI algorithm can be constructed.

Further, determining the target AI algorithm based on the preset AI algorithm corresponding to the target breeding species and the initial target AI algorithm includes: determining whether the input parameters and output parameters corresponding to the preset AI algorithm include the first breeding parameter, the first breeding parameter, and the first breeding parameter. At least one of the two breeding parameters and the third breeding parameter; if so, determine whether the preset AI algorithm is related to at least one of the first AI algorithm, the second AI algorithm, and the third AI algorithm. relationship; if yes, determine the target AI algorithm based on the initial target AI algorithm; if not, adjust the initial target AI algorithm based on the preset AI algorithm, and determine the target AI algorithm based on the adjusted AI algorithm.

In some embodiments, the correlation between algorithms may be that the complexity of the AI algorithms is similar, the types of the AI algorithms are the same or similar, etc.

In some embodiments, determining the target AI algorithm based on the initial target AI algorithm includes: improving the initial target AI algorithm, such as adding some basic data preprocessing algorithms, etc., and the improved algorithm is determined as the target AI algorithm.

In some embodiments, the preset AI algorithm may be a preset AI algorithm that matches the target breeding species.

In some embodiments, adjusting the initial target AI algorithm based on the preset AI algorithm includes: adding the preset AI algorithm to the initial target AI algorithm, and the preset AI algorithm is in a parallel relationship with other AI algorithms. Therefore, in the final target AI algorithm, after the output results of the preset AI algorithm and the third AI algorithm are integrated, the final algorithm result can be output.

In some embodiments, the construction method also includes: obtaining sample breeding data; determining the first sample breeding parameters based on the sample breeding data and the target AI algorithm; determining the second sample based on the sample breeding data and the preset smart breeding model Breeding parameters; the model algorithm corresponding to the preset smart breeding model is different from the target AI algorithm; based on the first sample breeding parameters, the second sample breeding parameters and the real breeding parameters corresponding to the sample breeding parameters, the target AI algorithm and the preset Optimize the designed smart breeding model.

In some embodiments, the sample breeding data is input into the target AI algorithm, and the target AI algorithm outputs the first sample breeding parameters. And, input the sample breeding data into the smart breeding model and output the second sample breeding parameters.

Among them, the smart breeding model can be an artificial intelligence model obtained in advance through model training, etc., which can be a neural network model, a random forest model, etc.

Further, the sample breeding parameter can correspond to a real breeding parameter, that is, the sample breeding data has been implemented and the corresponding breeding parameter data is obtained.

In some embodiments, the first sample breeding parameters are compared with the real breeding parameters. If the matching degree is high, for example, the parameter values are similar, there is no need to optimize the target AI algorithm.

If the matching degree is low, for example, the parameter values are far apart, the target AI algorithm needs to be optimized. Optimization methods include: optimizing the functions, parameters, algorithm cycle methods, etc.

In some embodiments, the second sample breeding parameters are compared with the real breeding parameters. If the matching degree is high, for example, the parameter values are similar, there is no need to optimize the smart breeding model.

If the matching degree is low, for example, the parameter values are far apart, the smart breeding model needs to be optimized. Optimization methods include: training again using a new training data set, etc.

It can be seen from the introduction of the embodiments of this application that on the one hand, determining the corresponding AI algorithms for different types of breeding parameters can improve the diversity of the AI algorithms. On the basis of higher diversity of the AI algorithms, regardless of Whether it is the accuracy of the AI algorithm or its adaptability to smart scenarios, both have improved accordingly. On the other hand, based on multiple AI algorithms, combined with the relationship between the input objects corresponding to each AI algorithm, the target AI algorithm is integrated and constructed, so that the fitness of the built target AI algorithm and the target breeding species is improved. Furthermore, on the basis of improved adaptability and accuracy, the applicability of AI algorithms has also been improved accordingly. Therefore, the AI algorithm construction method for smart farming can improve the adaptability of the AI algorithm to smart farming, as well as improve the accuracy and applicability of the AI algorithm.

Please refer to Figure 3, which is a schematic structural diagram of an AI algorithm construction device for smart farming provided by an embodiment of the present application, including:

The determination unit 301 is used to determine the first breeding parameter and the first AI algorithm corresponding to the first breeding parameter; the first AI algorithm is used to determine the first breeding parameter based on the target breeding data. The parameters are the breeding parameters of the environmental dimension; the second breeding parameters and the second AI algorithm corresponding to the second breeding parameters are determined; the second AI algorithm is used to determine the second breeding parameters based on the target breeding data, so The second breeding parameter is an artificial dimension breeding parameter; the third breeding parameter and the third AI algorithm corresponding to the third breeding parameter are determined; the third AI algorithm is used based on the first breeding parameter and/or the The second breeding parameter determines the third breeding parameter, and the third breeding parameter is a breeding parameter based on the environmental dimension and/or the artificial dimension; the construction unit 302 is configured to based on the first AI algorithm, The second AI algorithm, the third AI algorithm, the target breeding species and the target parameter relationship construct a target AI algorithm; the target AI algorithm is used to determine the breeding parameters of the target breeding species, and the target parameter relationship represents for: ;Among them,/> Represents the first breeding parameter,/> Represents the second breeding parameter,/> Represents the third breeding parameter,/> Represents the first weight,/> represents the second weight, Represents the first influence value,/> Represents the second influence value,/> Represents the third influence value,/> Represents the default influence value.

In some embodiments, the determining unit 301 is further configured to: obtain a plurality of first preset breeding parameters; the plurality of first preset breeding parameters are all breeding parameters of an environmental dimension; determine the plurality of first preset breeding parameters. The direct influence value of the breeding parameters on the target breeding species; the indirect influence value of the plurality of first preset breeding parameters on the target breeding species is determined; the indirect influence value is based on the plurality of first preset breeding The direct impact value of the parameter on the first associated cultured species of the target cultured species is determined; the similarity between the cultured environment of the first associated cultured species and the cultured environment of the target cultured species is greater than the first preset similarity ; Based on the direct impact value on the target cultured species and the indirect impact value on the target cultured species, determine the integrated impact value of the plurality of first preset breeding parameters; based on the integrated impact value, from the The first breeding parameter is determined among a plurality of first preset breeding parameters.

In some embodiments, the determining unit 301 is further configured to: obtain a plurality of first preset AI algorithms; the plurality of first preset AI algorithms respectively correspond to different breeding species; and obtain a plurality of first preset AI algorithms from the plurality of first preset AI algorithms. The first AI algorithm is determined in the algorithm; wherein the first AI algorithm has the highest number of corresponding breeding species among the plurality of first preset AI algorithms, and the breeding environment of the corresponding breeding species Comply with the preset breeding environment conditions.

In some embodiments, the determining unit 301 is further configured to: obtain a plurality of second preset breeding parameters; the plurality of second preset breeding parameters are all artificial dimension breeding parameters; determine the plurality of second preset breeding parameters. The direct influence value of the breeding parameters on the target breeding species; the indirect influence value of the plurality of second preset breeding parameters on the target breeding species is determined; the indirect influence value is based on the plurality of second preset breeding The direct influence value of the parameter on the second associated cultured species of the target cultured species is determined; the similarity between the artificial breeding conditions of the second associated cultured species and the artificial breeding conditions of the target cultured species is greater than the second preset Similarity; based on the direct impact value on the target cultured species and the indirect impact value on the target cultured species, determine the integrated impact value of the plurality of second preset breeding parameters; based on the integrated impact value, from Among the plurality of second preset breeding parameters, the second breeding parameter is determined.

In some embodiments, the determining unit 301 is further configured to: obtain a plurality of second preset AI algorithms; the plurality of second preset AI algorithms respectively correspond to different breeding species; and obtain a plurality of second preset AI algorithms from the plurality of second preset AI algorithms. The second AI algorithm is determined in the algorithm; wherein the second AI algorithm has the highest number of corresponding breeding species among the plurality of second preset AI algorithms, and the artificial breeding of the corresponding breeding species The conditions meet the preset artificial breeding conditions.

In some embodiments, the determining unit 301 is further configured to: obtain a plurality of third preset breeding parameters; the plurality of third preset breeding parameters are based on the first preset breeding parameters and/or the second The breeding parameters determined by the preset breeding parameters; determining the direct influence value of the plurality of third preset breeding parameters on the target breeding species; determining the indirect influence of the plurality of third preset breeding parameters on the target breeding species. The influence value; the indirect influence value is determined based on the direct influence value of the plurality of first preset breeding parameters on the third related breeding species of the target breeding species; the breeding environment of the third related breeding species is different from the The similarity between the breeding environments of the target breeding species is greater than the first preset similarity, and/or the similarity between the artificial breeding conditions of the third associated species and the artificial breeding conditions of the target breeding species is greater than the second Preset similarity; determine the integrated impact value of the plurality of third preset farming parameters based on the direct impact value on the target cultured species and the indirect impact value on the target cultured species; based on the integrated impact value , determine the third breeding parameter from the plurality of third preset breeding parameters.

In some embodiments, the determining unit 301 is further configured to: obtain a plurality of third preset AI algorithms; the plurality of third preset AI algorithms respectively correspond to different breeding species; and obtain a plurality of third preset AI algorithms from the plurality of third preset AI algorithms. The third AI algorithm is determined in the algorithm; wherein the third AI algorithm has the highest number of corresponding breeding species among the plurality of third preset AI algorithms, and the breeding environment of the corresponding breeding species meets The preset breeding environment conditions, and/or the artificial breeding conditions of the corresponding breeding species meet the preset artificial breeding conditions.

In some embodiments, the building unit 302 is further configured to: determine a connection relationship between the first AI algorithm, the second AI algorithm and the third AI algorithm based on the target parameter relationship; the connection relationship is used to Indicate the relationship between the input and output of each AI algorithm; based on the connection relationship, connect the first AI algorithm, the second AI algorithm and the third AI algorithm to construct an initial target AI algorithm; The target AI algorithm is determined based on the preset AI algorithm corresponding to the target breeding species and the initial target AI algorithm.

In some embodiments, the construction unit 302 is also used to determine whether the input parameters and output parameters corresponding to the preset AI algorithm include the first breeding parameter, the second breeding parameter, and the third breeding parameter. at least one breeding parameter in; if yes, determine whether the preset AI algorithm is associated with at least one AI algorithm among the first AI algorithm, the second AI algorithm and the third AI algorithm. relationship; if yes, determine the target AI algorithm based on the initial target AI algorithm; if not, adjust the initial target AI algorithm based on the preset AI algorithm, and determine the target AI based on the adjusted AI algorithm algorithm.

As shown in Figure 4, the embodiment of the present application also provides a terminal device, including a processor 401 and a memory 402. The processor 401 and the memory 402 are communicatively connected. The terminal device can be used as the aforementioned AI algorithm construction method for smart breeding. the execution subject.

The processor 401 and the memory 402 are electrically connected directly or indirectly to realize data transmission or interaction. For example, these components may be electrically connected through one or more communication buses or signal buses. The aforementioned AI algorithm construction methods for smart farming each include at least one software function module that can be stored in the memory 402 in the form of software or firmware.

The processor 401 may be an integrated circuit chip with signal processing capabilities. The processor 401 can be a general-purpose processor, including a CPU (Central Processing Unit, central processing unit), NP (Network Processor, network processor), etc.; it can also be a digital signal processor, an application-specific integrated circuit, an off-the-shelf programmable gate array, or Other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. It can implement or execute the disclosed methods, steps and logical block diagrams in the embodiments of the present invention. A general-purpose processor may be a microprocessor or the processor may be any conventional processor, etc.

The memory 402 can store various software programs and modules, such as program instructions/modules corresponding to the image processing method and device provided by embodiments of the present invention. The processor 401 executes various functional applications and data processing by running software programs and modules stored in the memory 402, that is, implementing the methods in the embodiments of the present application.

The memory 402 may include but is not limited to RAM (Random Access Memory), ROM (Read Only Memory), PROM (Programmable Read-Only Memory), EPROM (Erasable Programmable Read). -Only Memory, Erasable Read-Only Memory), EEPROM (Electric Erasable Programmable Read-Only Memory, Electrically Erasable Programmable Read-Only Memory), etc.

It can be understood that the structure shown in Figure 4 is only illustrative, and the terminal device may also include more or fewer components than shown in Figure 4, or have a different configuration than that shown in Figure 4.

Those skilled in the art will understand that embodiments of the present application may be provided as methods, systems, or computer program products. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment that combines software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) having computer-usable program code embodied therein.

The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the application. It will be understood that each process and/or block in the flowchart illustrations and/or block diagrams, and combinations of processes and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing device to produce a machine, such that the instructions executed by the processor of the computer or other programmable data processing device produce a use A device for realizing the functions specified in a process or processes in a flowchart and/or a block or blocks in a block diagram.

These computer program instructions may also be stored in a computer-readable memory that causes a computer or other programmable data processing apparatus to operate in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including the instruction means, the instructions The device implements the functions specified in a process or processes in the flowchart and/or in a block or blocks in the block diagram.

These computer program instructions may also be loaded onto a computer or other programmable data processing device, causing a series of operating steps to be performed on the computer or other programmable device to produce computer-implemented processing, thereby executing on the computer or other programmable device. Instructions provide steps for implementing the functions specified in a process or processes of a flowchart diagram and/or a block or blocks of a block diagram.

The foregoing descriptions of specific exemplary embodiments of the present application have been presented for purposes of illustration and illustration. These descriptions are not intended to limit the application to the precise forms disclosed, and obviously many modifications and variations are possible in light of the above teachings. The exemplary embodiments were chosen and described in order to explain certain principles of the invention and their practical applications, thereby enabling others skilled in the art to make and utilize various exemplary embodiments of the invention and various different applications. Choice and change. The scope of the application is intended to be defined by the claims and their equivalents.

Claims (8)

1. An AI algorithm construction method for smart farming, which is characterized by including: Determine the first breeding parameter and the first AI algorithm corresponding to the first breeding parameter; the first AI algorithm is used to determine the first breeding parameter based on the target breeding data, and the first breeding parameter is the environmental dimension of breeding parameter; Determine the second breeding parameter and the second AI algorithm corresponding to the second breeding parameter; the second AI algorithm is used to determine the second breeding parameter based on the target breeding data, and the second breeding parameter is an artificial dimension breeding parameters; Determine a third breeding parameter and a third AI algorithm corresponding to the third breeding parameter; the third AI algorithm is used to determine the third breeding parameter based on the first breeding parameter and/or the second breeding parameter , the third breeding parameter is a breeding parameter based on the environmental dimension and/or the artificial dimension; Based on the first AI algorithm, the second AI algorithm, the third AI algorithm, the target breeding species and the target parameter relationship, a target AI algorithm is constructed; the target AI algorithm is used to determine the breeding of the target breeding species Parameters, the target parameter relationship is expressed as: ;Among them,/> Represents the first breeding parameter/> Represents the second breeding parameter,/> Represents the third breeding parameter,/> Represents the first weight,/> Represents the second weight,/> Represents the first influence value,/> Represents the second influence value,/> Represents the third influence value,/> Represents the default influence value; Wherein, the determination of the first breeding parameter and the first AI algorithm corresponding to the first breeding parameter includes: obtaining a plurality of first preset breeding parameters; the plurality of first preset breeding parameters are all environmental dimensions. Breeding parameters; determine the direct impact value of the plurality of first preset breeding parameters on the target cultured species; determine the indirect impact value of the plurality of first preset cultured parameters on the target cultured species; the indirect The influence value is determined based on the direct influence value of the plurality of first preset breeding parameters on the first associated breeding species of the target breeding species; the breeding environment of the first associated breeding species and the breeding environment of the target breeding species The similarity between them is greater than the first preset similarity; based on the direct impact value on the target culture variety and the indirect impact value on the target culture species, the integrated impact of the multiple first preset culture parameters is determined value; based on the integrated influence value, determine the first breeding parameter from the plurality of first preset breeding parameters; Obtain a plurality of first preset AI algorithms; the plurality of first preset AI algorithms respectively correspond to different breeding species; determine the first AI algorithm from the plurality of first preset AI algorithms; wherein, Among the plurality of first preset AI algorithms, the first AI algorithm corresponds to the highest number of breeding species, and the breeding environment of the corresponding breeding species meets the preset breeding environment conditions. 2. The AI algorithm construction method for smart breeding according to claim 1, characterized in that the determination of the second breeding parameter and the second AI algorithm corresponding to the second breeding parameter includes: Obtain a plurality of second preset breeding parameters; the plurality of second preset breeding parameters are all artificial dimension breeding parameters; Determine the direct impact value of the plurality of second preset breeding parameters on the target breeding species; Determine the indirect influence value of the plurality of second preset breeding parameters on the target breeding species; the indirect influence value is based on the second associated breeding species of the target breeding species on the plurality of second preset breeding parameters. The direct influence value is determined; the similarity between the artificial breeding conditions of the second associated breeding species and the artificial breeding conditions of the target breeding species is greater than the second preset similarity; Determine the integrated impact value of the plurality of second preset breeding parameters based on the direct impact value on the target cultured species and the indirect impact value on the target cultured species; Based on the integrated influence value, the second breeding parameter is determined from the plurality of second preset breeding parameters. 3. The AI algorithm construction method for smart breeding according to claim 1, characterized in that the determination of the second breeding parameter and the second AI algorithm corresponding to the second breeding parameter includes: Obtain a plurality of second preset AI algorithms; the plurality of second preset AI algorithms respectively correspond to different breeding species; The second AI algorithm is determined from the plurality of second preset AI algorithms; wherein the second AI algorithm has the highest number of corresponding breeding species among the plurality of second preset AI algorithms. , and the artificial breeding conditions of the corresponding breeding species meet the preset artificial breeding conditions. 4. The AI algorithm construction method for smart farming according to claim 1, characterized in that the determination of the third farming parameter and the third AI algorithm corresponding to the third farming parameter include: Acquire a plurality of third preset breeding parameters; the plurality of third preset breeding parameters are breeding parameters determined based on the first preset breeding parameters and/or the second preset breeding parameters; Determine the direct impact value of the plurality of third preset breeding parameters on the target breeding species; Determine the indirect influence value of the plurality of third preset breeding parameters on the target breeding species; the indirect influence value is based on the direct influence of the first preset breeding parameters on the third associated breeding species of the target breeding species. The influence value is determined; the similarity between the breeding environment of the third related breeding species and the breeding environment of the target breeding species is greater than the first preset similarity, and/or the artificial breeding conditions of the third related breeding species The similarity with the artificial breeding conditions of the target breeding species is greater than the second preset similarity; Determine the integrated impact value of the plurality of third preset breeding parameters based on the direct impact value on the target cultured species and the indirect impact value on the target cultured species; Based on the integrated influence value, the third breeding parameter is determined from the plurality of third preset breeding parameters. 5. The AI algorithm construction method for smart farming according to claim 1, characterized in that the determination of the third farming parameter and the third AI algorithm corresponding to the third farming parameter include: Obtain a plurality of third preset AI algorithms; the plurality of third preset AI algorithms respectively correspond to different breeding species; The third AI algorithm is determined from the plurality of third preset AI algorithms; wherein the third AI algorithm has the highest number of corresponding breeding species among the plurality of third preset AI algorithms. , the breeding environment of the corresponding breeding species meets the preset breeding environment conditions, and/or the artificial breeding conditions of the corresponding breeding species meet the preset artificial breeding conditions. 6. The AI algorithm construction method for smart breeding according to claim 1, characterized in that the method is based on the first AI algorithm, the second AI algorithm, the third AI algorithm, the target breeding species relationship with the target parameters to construct the target AI algorithm, including: Based on the target parameter relationship, determine the connection relationship between the first AI algorithm, the second AI algorithm and the third AI algorithm; the connection relationship is used to indicate the relationship between the input and output of each AI algorithm; Based on the connection relationship, connect the first AI algorithm, the second AI algorithm and the third AI algorithm to construct an initial target AI algorithm; The target AI algorithm is determined based on the preset AI algorithm corresponding to the target breeding species and the initial target AI algorithm. 7. The AI algorithm construction method for smart breeding according to claim 6, characterized in that the target is determined based on the preset AI algorithm corresponding to the target breeding species and the initial target AI algorithm. AI algorithms, including: Determine whether the input parameters and output parameters corresponding to the preset AI algorithm include at least one of the first breeding parameter, the second breeding parameter and the third breeding parameter; If so, determine whether the preset AI algorithm has an associated relationship with at least one of the first AI algorithm, the second AI algorithm, and the third AI algorithm; If so, determine the target AI algorithm based on the initial target AI algorithm; If not, the initial target AI algorithm is adjusted based on the preset AI algorithm, and the target AI algorithm is determined based on the adjusted AI algorithm. 8. The AI algorithm construction method for smart farming according to claim 1, characterized in that the AI algorithm construction method for smart farming further includes: Obtain sample breeding data; Based on the sample breeding data and the target AI algorithm, determine the first sample breeding parameters; Determine the second sample breeding parameters based on the sample breeding data and the preset smart breeding model; the model algorithm corresponding to the preset smart breeding model is different from the target AI algorithm; Based on the first sample breeding parameters, the second sample breeding parameters and the real breeding parameters corresponding to the sample breeding parameters, the target AI algorithm and the preset smart breeding model are optimized.