Intelligent feed rationing management system based on big data

Abstract

This invention relates to the field of feed formulation technology, specifically to a big data-based intelligent quantitative feed formulation management system. The system includes: a data acquisition unit for determining the current total nutritional requirements of the target animal based on real-time growth data and growth stage information; a formula retrieval unit for searching a candidate formula library with feature dimensions no lower than the feature vector within the formula management system, selecting candidate raw material combinations whose total nutrient content matches the total nutritional requirements and whose raw material types and quantities meet the requirements; and determining whether there are reserve formulas with a matching degree index of 0 in each candidate formula library. Through screening the candidate formula library and judging reserve formulas, this invention allows the system to continuously optimize formulas, ensuring the optimal and economical use of feed components. If a new formula does not meet preset standards, the system automatically calculates and resets the formula, thereby ensuring continuous improvement in feed quality.

Description

Technical Field

[0001] This invention relates to the field of feed formulation technology, specifically to a big data-based intelligent quantitative feed formulation management system. Background Technology

[0002] Currently, traditional systems often rely on static data and historical records, making it impossible to monitor the growth status and nutritional needs of farmed animals in real time. This results in feed formulations not being adjusted in a timely manner, potentially leading to malnutrition or feed waste. Furthermore, in traditional systems, feeding plans are usually fixed and cannot be dynamically adjusted according to environmental changes such as temperature, humidity, and equipment status. This rigid feeding strategy may lead to overfeeding or underfeeding, affecting farming results. In addition, traditional methods are usually based on experience or fixed formulas for feed management, lacking systematic screening and optimization of candidate formulas. Therefore, the selection of feed components may not be scientific enough and may be difficult to adapt to different growth stages and specific needs of farmed animals.

[0003] Furthermore, traditional systems often lead to resource waste due to the lack of precise matching between the nutritional needs of the farmed animals and the feed composition; failure to make reasonable use of the characteristics of various raw materials may increase production costs and reduce the overall feed utilization rate; once problems occur, such as animal health risks or sudden environmental changes, traditional systems often react slowly and cannot make adjustments quickly; this delay may lead to greater losses and affect the overall farming efficiency. Summary of the Invention

[0004] To achieve the above objectives, the present invention provides the following technical solution: a big data-based intelligent quantitative feed formulation management system, comprising: The data acquisition unit is used to acquire real-time growth data, growth stage information, and aquaculture environment parameters of the target aquaculture object; determine the current total nutritional requirement of the target aquaculture object based on the real-time growth data and growth stage information; generate a feature vector based on the mapping relationship between the current total nutritional requirement and the nutritional database; and determine the types and quantities of raw materials required to meet the total nutritional requirement. The formula retrieval unit is used in the formula management system to search the candidate formula library with feature dimensions no less than feature vectors, and to select candidate raw material combinations whose total nutrient content matches the total nutrient requirement and whose types and quantities of raw materials meet the requirements; and to determine whether there are reserve formulas with a matching degree index of 0 in each candidate formula library. The first judgment unit is used to allocate new formulas to the candidate formula library with feature dimensions not lower than the feature vector in order of feature dimension from low to high if there are reserve formulas, and determine the optimal ratio of each new formula by combining the candidate raw material combination. The second judgment unit is used to set the matching degree index of the formula in the benchmark formula library with feature dimension as feature vector to 0 if there is no reserve formula. Then, in order of feature dimension from low to high, new formulas are assigned to the candidate formula library with feature dimension not lower than feature vector. The optimal ratio of each new formula is determined by combining the candidate raw material combination. The initial feeding unit is used to generate initial feeding plans for different feeding equipment based on the optimal ratio and total nutrient requirements; determine the number of continuous feeding periods required to execute the initial feeding plan; and determine whether there are available idle feeding windows that meet the requirements during the working time of different feeding equipment in combination with aquaculture environment parameters. The proportioning adjustment unit is used to adjust the feeding plan based on the judgment result of the idle feeding window to obtain the target feeding time period; send the target feeding time period and the corresponding new formula to the corresponding feeding equipment for execution, and update the relevant formula in the formula management system according to the execution result.

[0005] Preferably, the formula management system includes one basic formula library and multiple feature formula libraries; The current total nutritional requirements of the target aquaculture species are determined based on real-time growth data and growth stage information, including: Determine the species identification information of the target aquaculture species; Determine whether the target aquaculture species matches the feature formula library based on the variety identification information; By combining growth stage information and matching results, the required proportions of protein, energy, and fiber for the target cultured organisms during the current growth cycle can be determined. The current total nutritional requirements are calculated based on the protein, energy, and fiber ratios, as well as real-time growth data.

[0006] Preferably, the method of determining whether the target aquaculture species matches the feature formula library based on the variety identification information includes: For each feature formula library in the formula management system, the feature value of the feature formula library is determined based on the raw material component values, processing technology, and storage period of the feature formula library; Based on the species identification information, growth stage information, and feature values ​​of each feature formula library of the target aquaculture species, it is determined whether the target aquaculture species matches the feature formula library in the formula management system.

[0007] Preferably, a feature vector is generated based on the mapping relationship between the current total nutritional requirements and the nutritional database to determine the types and quantities of raw materials required to meet the total nutritional requirements, including: The proportions of protein, energy, and fiber in the current total nutritional requirements are encoded and converted to obtain the encoded sequence; The encoded sequence is used as a feature vector and mapped and matched with the raw material nutrient data in the nutrient database. Based on the mapping and matching results, determine the types and quantities of raw materials that can meet the total nutritional requirements.

[0008] Preferably, in the formula management system, a search is performed on a candidate formula library with a feature dimension no lower than that of the feature vector, and candidate ingredient combinations that match the total nutritional requirements and have the required number and variety of ingredients are selected, including: Retrieve all candidate recipes from the recipe management system whose feature dimension is no less than that of the feature vector; Extract raw material combination information from each candidate formula library and screen out alternative raw material combinations whose number of raw material types matches the determined number of raw material types. Calculate the total nutritional content of each candidate ingredient combination and screen out the candidate ingredient combinations whose total nutritional content matches the current total nutritional requirements.

[0009] Preferably, the number of continuous feeding periods required to execute the initial feeding plan is determined, and the availability of sufficient idle feeding windows within the working hours of different feeding equipment is assessed in conjunction with aquaculture environment parameters, including: For each feeding device, the number of continuous feeding periods required for the feeding device to complete a single feeding is determined based on the feeding speed of the feeding device and the single feeding amount in the initial feeding plan. Based on the temperature and humidity data in the aquaculture environment parameters, determine the appropriate working time for each feeding device; Within the suitable working period of each feeding device, retrieve the target cycle of the number of consecutive feeding periods starting from the current moment, and determine whether there is an idle feeding window with a duration that meets the feeding period requirements for each target cycle.

[0010] Preferably, the feeding plan is adjusted based on the determination of the idle feeding window to obtain the target feeding time period, including: If there are available idle feeding windows that meet the requirements in the continuous target cycle of different feeding equipment, select the number of idle feeding windows in the same continuous feeding period from the idle feeding windows in different target cycles as the target feeding period; If the number of consecutive idle feeding windows in any target period is less than the number of feeding time periods, the feeding plan for the non-idle feeding windows allocated in different target periods is adjusted so that there are target feeding time periods in different target periods.

[0011] Preferably, the target feeding time period and the corresponding new formula are sent to the corresponding feeding equipment for execution, and the relevant formulas in the formula management system are updated according to the execution results, including: The target feeding period and the corresponding new formula and optimal ratio are integrated into a target feeding plan and sent to the corresponding feeding equipment for execution. Collect execution data from the feeding equipment and actual feeding data from the target aquaculture species; The suitability of the new formula is determined based on the execution data and actual feeding data. If the compatibility meets the preset standards, the corresponding candidate formula library will be updated, and the matching index and nutritional balance in the formula will be adjusted. If the compatibility does not meet the preset standard, the number of times the new formula is allocated is counted. When the number of allocations reaches the preset number, the formulas corresponding to all feature formula libraries in the formula management system are reset.

[0012] Compared with the prior art, the beneficial effects of the present invention are: This invention enables the system to assess the current nutritional needs of the target aquaculture species in real time by using real-time growth data and growth stage information, thereby accurately formulating feed formulas that meet their growth requirements. Furthermore, by adjusting the feeding plan based on environmental parameters and equipment status, the system can ensure that feeding is carried out at the optimal time, which also greatly improves the efficiency of resource utilization in the aquaculture process and reduces the negative impact on the aquaculture environment. This invention allows the system to continuously optimize formulas by screening candidate formula libraries and judging reserve formulas, ensuring that the feed ingredients used are optimal and economical. If a new formula does not meet the preset standards, the system will automatically count and reset the formula, thereby ensuring continuous improvement in feed quality. Furthermore, by matching the nutritional components of different raw materials with the needs of the target livestock species, resource waste is avoided, which not only helps to reduce production costs but also improves the overall utilization rate of feed, thereby enhancing livestock farming efficiency. Attached Figure Description

[0013] Figure 1 This is a schematic diagram of the overall system architecture in one embodiment of the present invention.

[0014] In the diagram: 1. Data acquisition unit; 2. Formula retrieval unit; 3. First judgment unit; 4. Second judgment unit; 5. Initial feeding unit; 6. Proportion adjustment unit. Detailed Implementation

[0015] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0016] Example 1, please refer to Figure 1 This invention provides a technical solution: a big data-based intelligent quantitative feed formulation management system, comprising: Data acquisition unit 1 is used to acquire real-time growth data, growth stage information and aquaculture environment parameters of the target aquaculture object, determine the current total nutritional requirements of the target aquaculture object based on the real-time growth data and growth stage information, generate feature vectors based on the mapping relationship between the current total nutritional requirements and the nutritional database, and determine the types and quantities of raw materials required to meet the total nutritional requirements. Formula retrieval unit 2 is used to search the candidate formula library with feature dimensions not lower than feature vector in the formula management system, filter out candidate raw material combinations whose total nutritional components match the total nutritional requirements and whose raw material types and quantities meet the requirements; and determine whether there are reserve formulas with a matching degree index of 0 in each candidate formula library. The first judgment unit 3 is used to allocate new formulas to the candidate formula library with feature dimensions not lower than the feature vector in order of feature dimension from low to high if there are reserve formulas, and determine the optimal ratio of each new formula by combining the candidate raw material combination. The second judgment unit 4 is used to set the matching degree index of the formula corresponding to the benchmark formula library with feature dimension as feature vector to 0 if there is no reserve formula. Then, in order of feature dimension from low to high, new formulas are assigned to the candidate formula library with feature dimension not lower than feature vector, and the optimal ratio of each new formula is determined by combining the candidate raw material combination. The initial feeding unit 5 is used to generate initial feeding plans for different feeding equipment based on the optimal ratio and total nutrient requirements; determine the number of continuous feeding periods required to execute the initial feeding plan; and determine whether there are any available feeding windows that meet the requirements during the working time of different feeding equipment in combination with aquaculture environment parameters. The proportioning adjustment unit 6 is used to adjust the feeding plan based on the judgment result of the idle feeding window to obtain the target feeding time period; send the target feeding time period and the corresponding new formula to the corresponding feeding equipment for execution, and update the relevant formula in the formula management system according to the execution result.

[0017] It should be noted that the role of the data acquisition unit is to monitor the growth of farmed organisms (such as fish, shrimp or other aquatic products) and their environmental conditions in real time. For example, when environmental parameters such as water temperature and dissolved oxygen are collected, the system will analyze the data to determine the current growth stage of the farmed organisms and how much nutrition they need to grow healthily. The formula retrieval unit uses the obtained nutritional requirement information to search for suitable feed formulas in a database containing multiple formulas. Assuming the system finds multiple potential feed combinations based on feature vectors, it then checks whether these combinations can fully meet the nutritional requirements of the farmed animals. If a formula is found to have a matching degree of zero, it means that the formula is not suitable for the current nutritional requirements. The task of the first judgment unit is to process the reserve formulas with a matching degree of zero. If such a formula exists, the system will gradually adjust the formula according to the complexity of the features until the optimal feed ratio is found so that the new formula can better meet the needs of the farmed species. For example, if a fish needs a specific ratio of protein and fat, and the existing reserve formula fails to meet this ratio, the system will try to replace it with a formula with a higher feature dimension. If there is no suitable reserve formula, the second judgment unit will set the matching degree of the benchmark formula to zero and continue to search for a formula that meets the requirements. In this process, the system will also consider the complexity of the feature dimensions to ensure that the final selected formula can accurately meet the nutritional needs of the aquaculture species. The initial feeding unit is responsible for formulating a specific feeding plan. It calculates the amount of each raw material to be fed based on the optimal ratio and total nutritional requirements, and determines when to feed these raw materials. If the system calculates that a certain amount of feed needs to be fed within a certain period of time, and finds a suitable time window based on the equipment's operating status, then the feeding can be scheduled for that period of time. The proportioning adjustment unit will fine-tune the feeding plan according to the actual situation to ensure smooth execution in idle feeding equipment; for example, if there are two feeding devices, but only one can be used within the scheduled time, the system will adjust the feeding plan so that all nutrients are satisfied within the appropriate time.

[0018] In one optional embodiment, the formula management system includes a basic formula library and multiple feature formula libraries; The current total nutritional requirements of the target aquaculture species are determined based on real-time growth data and growth stage information, including: Determine the species identification information of the target aquaculture species; Determine whether the target aquaculture species matches the feature formula library based on the variety identification information; By combining growth stage information and matching results, the required proportions of protein, energy, and fiber for the target cultured organisms during the current growth cycle can be determined. The current total nutritional requirements are calculated based on the protein, energy, and fiber ratios, as well as real-time growth data.

[0019] It should be noted that determining the species identification information of the target aquaculture object means that the system needs to identify the specific species of the aquaculture object; this information is crucial for subsequent calculation of nutritional requirements, because different species of aquaculture objects may have very different nutritional requirements during their growth process. The system determines whether the target species matches the feature formula library based on the species identification information. This means that the system will check whether there is a suitable feed formula for the species. If the species does not match the formula in the existing feature formula library, the system may not be able to provide appropriate nutritional support. The system will determine the proportions of protein, energy, and fiber required by the cultured organism during its current growth cycle; for example, a rapidly growing fish may require a higher proportion of protein at a certain growth stage to support muscle growth, while its energy and fiber requirements may be adjusted as it enters maturity. Based on the protein, energy, and fiber ratios, as well as real-time growth data, the system calculates the total current nutritional requirements of the farmed organisms. This means that the system can extrapolate the total amount of specific nutrients that need to be provided based on known ratios and the latest growth data (such as weight and body length). For example, if the system determines that a certain fish needs 40% protein, 30% energy, and 30% fiber at its current growth stage, and measurements show that the fish currently weighs 1 kilogram, then the system will calculate the specific grams of each nutrient that needs to be provided.

[0020] In an optional embodiment, determining whether a target aquaculture species matches a feature formula library based on species identification information includes: For each feature formula library in the formula management system, the feature value of the feature formula library is determined based on the raw material component values, processing technology, and storage period of the feature formula library; Based on the species identification information, growth stage information, and feature values ​​of each feature formula library of the target aquaculture species, it is determined whether the target aquaculture species matches the feature formula library in the formula management system.

[0021] It should be noted that the ingredient composition values ​​include the types and proportions of ingredients used in the formula; for example, the formula may contain fish meal, soybean meal, vitamins and minerals, etc. The specific composition and content of each ingredient will affect the final nutritional characteristics. Different processing methods (such as cooking, drying, grinding, etc.) will affect the digestibility and nutritional value of feed; for example, formulas that are processed at high temperatures may be more easily digested by livestock, thereby improving their absorption efficiency. The storage conditions and time of feed also affect its nutritional value; over time, some nutrients may degrade, so freshness is also part of the characteristic value. Taking into account the above factors, the system will generate a comprehensive feature value for each feature formula. This value reflects the overall performance and adaptability of the formula in actual use. The system will determine whether the formula matches the formula in the feature formula library based on the specific circumstances of the target aquaculture species; this process includes the following aspects: The system will first identify the species of the aquaculture object; different species have significantly different nutritional requirements. The growth stage of the farmed organisms (such as the seedling stage, growth stage, or maturity stage) is also crucial, because the nutritional needs of farmed organisms at different stages will be different; for example, in the early stages of growth, they may need more protein, while in the maturity stage, their need for energy and fiber increases. The system compares the species and growth stage of the target aquaculture object with the feature values ​​in the feature formula library to determine the matching degree; if the feature value of a certain formula can meet the current nutritional needs of the aquaculture object, then the formula is considered suitable.

[0022] In an optional embodiment, a feature vector is generated based on the mapping relationship between the current total nutritional requirement and the nutritional database to determine the types and quantities of raw materials required to meet the total nutritional requirement, including: The proportions of protein, energy, and fiber in the current total nutritional requirements are encoded and converted to obtain the encoded sequence; The encoded sequence is used as a feature vector and mapped and matched with the raw material nutrient data in the nutrient database. Based on the mapping and matching results, determine the types and quantities of raw materials that can meet the total nutritional requirements.

[0023] It should be noted that the proportions of protein, energy, and fiber in the current total nutritional requirements are encoded and converted; this means that the system will convert the proportions of these three nutrients into a coded form that is easy to calculate and process; for example, simple numbers or letters can be used to represent different nutrients. For example, if protein accounts for 50%, energy for 30%, and fiber for 20%, the system may convert it into a coding sequence, such as P50E30F20; here P represents protein, E represents energy, and F represents fiber; such coding makes the expression of nutritional needs more concise and clear. The system uses this encoded sequence as a feature vector and maps it to the nutritional data of each raw material in the nutrition database. The nutrition database contains detailed nutritional information of various raw materials (such as fish meal, soybean meal, grains, etc.), such as the specific content of protein, energy and fiber in each raw material. The system analyzes the nutritional requirements represented by the feature vectors and compares them with the raw material data in the database. For example, if a certain raw material has a high protein content and its energy and fiber content also meet the requirements, it can be considered that the raw material is suitable for meeting the nutritional requirements of the target aquaculture species. Based on the mapping and matching results, the system will determine the types and quantities of raw materials that can meet the current total nutritional requirements; this means that the system will count the number of raw materials that meet the nutritional requirements and provide farmers with options.

[0024] In an optional embodiment, the formula management system searches a candidate formula library with a feature dimension no less than the feature vector, and selects candidate ingredient combinations whose total nutrient content matches the total nutrient requirement and whose ingredient types and quantities meet the requirements, including: Retrieve all candidate recipes from the recipe management system whose feature dimension is no less than that of the feature vector; Extract raw material combination information from each candidate formula library and screen out alternative raw material combinations whose number of raw material types matches the determined number of raw material types. Calculate the total nutritional content of each candidate ingredient combination and screen out the candidate ingredient combinations whose total nutritional content matches the current total nutritional requirements.

[0025] It should be noted that the system will search from all candidate recipes with a feature dimension no lower than that feature vector, according to the definition of the feature vector. For example, if the nutritional requirement represented by the feature vector is P40E35F25, then the system will search for all recipes that meet or exceed this standard in the feature dimension (such as protein, energy, fiber, etc.). This retrieval ensures that the selected candidate formula library has sufficient nutrients to meet the basic needs of the target aquaculture species; therefore, only those formulas that meet or exceed the feature vector requirements in terms of major nutrients can be listed as candidate formulas. After finding the candidate recipe library, the system will extract the raw material combination information of each candidate recipe; the purpose of this step is to understand which specific raw materials are contained in each candidate recipe and their quantities; for example, one candidate recipe may contain fish meal, soybean meal and corn, while another candidate recipe may contain shrimp meal and grains; The system will filter out candidate raw material combinations whose number of raw material types matches the previously determined number of raw material types; if the goal is to find combinations of three raw materials, then only those candidate formulas that contain exactly three raw materials will be included in the candidate list. For the selected candidate ingredient combinations, the system will calculate their total nutritional composition. This means that the system will sum up the nutritional composition of each ingredient in a specific formula, such as protein, energy, and fiber, to obtain a comprehensive nutritional composition data. For example, if a formula consists of 50% fishmeal and 50% soybean meal, the system will calculate the protein, energy, and fiber content of these two ingredients separately and add them together to form the total nutritional composition of the formula. The system will compare the total nutritional content of these candidate ingredient combinations with the current total nutritional requirements to screen out candidate ingredient combinations that can be perfectly matched; only those combinations that fully meet the needs of the target aquaculture species in terms of protein, energy, and fiber will be considered suitable.

[0026] In an optional embodiment, determining the number of continuous feeding periods required to execute the initial feeding plan, and judging whether there are sufficient idle feeding windows in the working time of different feeding devices based on aquaculture environment parameters, includes: For each feeding device, the number of continuous feeding periods required for the feeding device to complete a single feeding is determined based on the feeding speed of the feeding device and the single feeding amount in the initial feeding plan. Based on the temperature and humidity data in the aquaculture environment parameters, determine the appropriate working time for each feeding device; Within the suitable working period of each feeding device, retrieve the target cycle of the number of consecutive feeding periods starting from the current moment, and determine whether there is an idle feeding window with a duration that meets the feeding period requirements for each target cycle.

[0027] It should be noted that for each feeding device, it is necessary to understand the feeding speed of the device and the single feeding amount in the initial feeding plan; the feeding speed refers to the amount of material that the device can feed per unit time, while the single feeding amount is the actual weight or volume of material fed each time. These two parameters can be used to calculate the number of consecutive feeding periods required to complete one feeding cycle. For example, if a feeding device has a feeding speed of 10 kg per minute and a single feeding amount of 50 kg, it will take 5 minutes to complete one feeding cycle, which means that the number of consecutive feeding periods is 5. The system determines the appropriate operating time for each feeding device based on temperature and humidity data in the aquaculture environment parameters; this step is very important because different aquaculture environment conditions will affect the operating efficiency of the equipment and the quality of the materials. For example, a certain feeding device may not be suitable for operation in conditions of excessively high temperature or humidity. Therefore, the system will set a suitable operating time range, such as from 7:00 AM to 7:00 PM. During this time period, the device can operate normally and will not be affected by environmental factors. During the appropriate working period of each feeding device, the system will retrieve the number of consecutive target cycles starting from the current moment; these target cycles refer to a series of time periods planned by the system within the appropriate working period for arranging future feeding operations. For each target period, the system will determine whether there is an idle feeding window that meets the feeding time period; in other words, the system will check whether there is a sufficient continuous idle time period within the set target period that can be used for feeding operations. For example, assuming the suitable working hours are from 7:00 AM to 7:00 PM, and the current time is 9:00 AM; if the next planned feeding period requires 5 minutes, the system will check the next target cycle starting from 9:00 AM, for example: Target period 1: From 9:00 to 9:30; Target period 2: From 9:30 to 10:00; Target period 3: From 10:00 to 10:30; The system will check these time periods one by one to confirm whether there is at least 5 minutes of continuous idle time in each cycle to complete the feeding operation.

[0028] In an optional embodiment, adjusting the feeding plan based on the determination result of the idle feeding window to obtain the target feeding time period includes: If there are available idle feeding windows that meet the requirements in the continuous target cycle of different feeding equipment, select the number of idle feeding windows in the same continuous feeding period from the idle feeding windows in different target cycles as the target feeding period; If the number of consecutive idle feeding windows in any target period is less than the number of feeding time periods, the feeding plan for the non-idle feeding windows allocated in different target periods is adjusted so that there are target feeding time periods in different target periods.

[0029] It should be noted that during the continuous target cycles of multiple feeding devices, the system will check whether there are any idle feeding windows that meet the requirements in each target cycle. If all target cycles meet this condition, the system will select a specified number of time periods from these idle windows as the target feeding time periods. The key to this step is to ensure that there are enough idle time periods in the target cycles of all devices to complete the predetermined feeding task. For example, if the system is set to take 4 minutes for each feeding, and it finds that there are 4-minute idle time periods available in each of the different target cycles, then the system will select these time periods as the target feeding time periods; If, during a target period, the number of available continuous idle feeding windows is found to be less than the required number of feeding time slots, then the system needs to adjust the allocated non-idle feeding windows; this means that the system may need to reschedule some feeding operations to ensure that there is sufficient time for feeding in all target periods. For example, suppose the system needs to schedule 5 feedings within three target cycles, with each feeding taking 4 minutes; after retrieving the target cycles, the system finds: Within target period 1, there are 3 suitable idle windows; In target period 2, there are 4 suitable idle windows; In target period 3, there is only one suitable idle window; In this situation, since the number of idle windows in target cycle 3 is less than 5, the system will try to adjust the feeding plan. For example, it may consider reallocating some feeding operations that have been scheduled in target cycles 1 and 2 to target cycle 3, or moving the feeding operations in target cycle 3 to other time periods with more idle windows. This flexible adjustment can ensure that there is enough time to complete the feeding task in each target cycle, thereby improving the efficiency of overall aquaculture management.

[0030] In an optional embodiment, the target feeding time period and the corresponding new formula are sent to the corresponding feeding equipment for execution, and the relevant formula in the formula management system is updated according to the execution result, including: The target feeding period and the corresponding new formula and optimal ratio are integrated into a target feeding plan and sent to the corresponding feeding equipment for execution. Collect execution data from the feeding equipment and actual feeding data from the target aquaculture species; The suitability of the new formula is determined based on the execution data and actual feeding data. If the compatibility meets the preset standards, the corresponding candidate formula library will be updated, and the matching index and nutritional balance in the formula will be adjusted. If the compatibility does not meet the preset standard, the number of times the new formula is allocated is counted. When the number of allocations reaches the preset number, the formulas corresponding to all feature formula libraries in the formula management system are reset.

[0031] It should be noted that the system will integrate the target feeding time period with the corresponding new formula and its optimal ratio to form a complete target feeding plan; this plan will be sent to the corresponding feeding equipment so that the equipment can feed the materials according to the predetermined time and formula; For example, suppose the system is set to feed materials three times between 9:00 AM and 9:30 AM, with each feeding having a formula of A, B, and C, corresponding to optimal proportions of 40%, 30%, and 30%, respectively; then the system will create a target feeding plan, integrate this information, and send it to the feeding equipment. After the feeding equipment performs the feeding operation, the system will collect the equipment's execution data (such as the actual amount of feed fed, feeding time, etc.) and the actual feeding data of the target livestock (such as the actual amount of feed consumed, feed acceptance, etc.). By analyzing this data, the system will evaluate the suitability of the new formula. For example, suppose that after feeding, the equipment records the amount of feed for each formula, and the feeding data of the livestock shows that when using the new formula A, the feed intake of the target livestock increases significantly, which indicates that formula A is well-suited. If the compatibility of a new formula meets preset standards, such as if the actual feed intake reaches or exceeds expectations, the system will update the formulas in the candidate formula library. This includes adjusting the formula matching index and nutritional balance to make them more in line with actual needs. The purpose of this is to continuously optimize the formula and improve breeding efficiency. If the actual intake of formula A is significantly higher than expected, the system will increase the matching index of the formula and may adjust the ingredients to ensure that the nutritional balance is improved. Conversely, if the suitability of a new formula does not meet the preset standards, such as when the actual feed intake is lower than expected, the system will start counting the number of times the new formula is distributed. When the number of times this new formula is distributed reaches a preset threshold, such as 5 times, the system will reset all formulas in the feature formula library in the formula management system. This means that the system will no longer use this formula and may revert to the previous formula or select other candidate formulas for feeding. For example, if Formula B fails to reach the expected feed intake after 5 feedings, the system will remove Formula B from the formula library and re-examine all feature formulas to find new, more suitable formulas to replace it.

[0032] The embodiments of the present invention have been described in detail above with reference to the accompanying drawings. However, the present invention is not limited thereto. Various changes can be made within the scope of knowledge possessed by those skilled in the art without departing from the spirit of the present invention.

Claims

1. A big data-based intelligent quantitative feed formulation management system, characterized in that, include: The data acquisition unit is used to acquire real-time growth data, growth stage information, and aquaculture environment parameters of the target aquaculture object; determine the current total nutritional requirement of the target aquaculture object based on the real-time growth data and growth stage information; generate a feature vector based on the mapping relationship between the current total nutritional requirement and the nutritional database; and determine the types and quantities of raw materials required to meet the total nutritional requirement. The formula retrieval unit is used in the formula management system to search the candidate formula library with feature dimensions no less than feature vectors, and to select candidate raw material combinations whose total nutrient content matches the total nutrient requirement and whose types and quantities of raw materials meet the requirements; and to determine whether there are reserve formulas with a matching degree index of 0 in each candidate formula library. The first judgment unit is used to allocate new formulas to the candidate formula library with feature dimensions not lower than the feature vector in order of feature dimension from low to high if there are reserve formulas, and determine the optimal ratio of each new formula by combining the candidate raw material combination. The second judgment unit is used to set the matching degree index of the formula in the benchmark formula library with feature dimension as feature vector to 0 if there is no reserve formula. Then, in order of feature dimension from low to high, new formulas are assigned to the candidate formula library with feature dimension not lower than feature vector. The optimal ratio of each new formula is determined by combining the candidate raw material combination. The initial feeding unit is used to generate initial feeding plans for different feeding devices based on the optimal ratio and total nutrient requirements. Determine the number of continuous feeding periods required to execute the initial feeding plan, and determine whether there are any available feeding windows that meet the requirements during the working time of different feeding equipment, based on the aquaculture environment parameters. The proportioning adjustment unit is used to adjust the feeding plan based on the judgment result of the idle feeding window to obtain the target feeding time period; Send the target feeding time period and the corresponding new formula to the corresponding feeding equipment for execution, and update the relevant formula in the formula management system based on the execution results.

2. The intelligent feed quantitative proportioning management system based on big data according to claim 1, characterized in that, The formula management system includes one basic formula library and multiple feature formula libraries; The current total nutritional requirements of the target aquaculture species are determined based on real-time growth data and growth stage information, including: Determine the species identification information of the target aquaculture species; Determine whether the target aquaculture species matches the feature formula library based on the variety identification information; By combining growth stage information and matching results, the required proportions of protein, energy, and fiber for the target cultured organisms during the current growth cycle can be determined. The current total nutritional requirements are calculated based on the protein, energy, and fiber ratios, as well as real-time growth data.

3. The intelligent feed quantitative proportioning management system based on big data according to claim 2, characterized in that, Determine whether the target aquaculture species matches the feature formula library based on the species identification information, including: For each feature formula library in the formula management system, the feature value of the feature formula library is determined based on the raw material component values, processing technology, and storage period of the feature formula library; Based on the species identification information, growth stage information, and feature values ​​of each feature formula library of the target aquaculture species, it is determined whether the target aquaculture species matches the feature formula library in the formula management system.

4. The intelligent feed quantitative proportioning management system based on big data according to claim 3, characterized in that, Based on the mapping relationship between the current total nutritional requirements and the nutritional database, a feature vector is generated to determine the types and quantities of raw materials required to meet the total nutritional requirements, including: The proportions of protein, energy, and fiber in the current total nutritional requirements are encoded and converted to obtain the encoded sequence; The encoded sequence is used as a feature vector and mapped and matched with the raw material nutrient data in the nutrient database. Based on the mapping and matching results, determine the types and quantities of raw materials that can meet the total nutritional requirements.

5. The intelligent feed quantitative proportioning management system based on big data according to claim 4, characterized in that, In the formula management system, a search is conducted on the candidate formula library with feature dimensions no lower than the feature vector. Candidate ingredient combinations that match the total nutritional requirements and have the required variety and quantity of ingredients are selected, including: Retrieve all candidate recipes from the recipe management system whose feature dimension is no less than that of the feature vector; Extract raw material combination information from each candidate formula library and screen out alternative raw material combinations whose number of raw material types matches the determined number of raw material types. Calculate the total nutritional content of each candidate ingredient combination and screen out the candidate ingredient combinations whose total nutritional content matches the current total nutritional requirements.

6. The intelligent feed quantitative proportioning management system based on big data according to claim 5, characterized in that, Determine the number of continuous feeding periods required to execute the initial feeding plan, and determine whether there are sufficient idle feeding windows within the working hours of different feeding equipment, based on aquaculture environment parameters, including: For each feeding device, the number of continuous feeding periods required for the feeding device to complete a single feeding is determined based on the feeding speed of the feeding device and the single feeding amount in the initial feeding plan. Based on the temperature and humidity data in the aquaculture environment parameters, determine the appropriate working time for each feeding device; Within the suitable working period of each feeding device, retrieve the target cycle of the number of consecutive feeding periods starting from the current moment, and determine whether there is an idle feeding window with a duration that meets the feeding period requirements for each target cycle.

7. The intelligent feed quantitative proportioning management system based on big data according to claim 6, characterized in that, Adjust the feeding plan based on the results of the idle feeding window assessment to obtain the target feeding time period, including: If there are available idle feeding windows that meet the requirements in the continuous target cycle of different feeding equipment, select the number of idle feeding windows in the same continuous feeding period from the idle feeding windows in different target cycles as the target feeding period; If the number of consecutive idle feeding windows in any target period is less than the number of feeding time periods, the feeding plan for the non-idle feeding windows allocated in different target periods is adjusted so that there are target feeding time periods in different target periods.

8. The intelligent feed quantitative proportioning management system based on big data according to claim 7, characterized in that, Send the target feeding time period and the corresponding new formula to the corresponding feeding equipment for execution, and update the relevant formulas in the formula management system based on the execution results, including: The target feeding period and the corresponding new formula and optimal ratio are integrated into a target feeding plan and sent to the corresponding feeding equipment for execution. Collect execution data from the feeding equipment and actual feeding data from the target aquaculture species; The suitability of the new formula is determined based on the execution data and actual feeding data. If the compatibility meets the preset standards, the corresponding candidate formula library will be updated, and the matching index and nutritional balance in the formula will be adjusted. If the compatibility does not meet the preset standard, the number of times the new formula is allocated is counted. When the number of allocations reaches the preset number, the formulas corresponding to all feature formula libraries in the formula management system are reset.

Images