A method, system, apparatus, and storage medium for optimizing feed formulation.

Abstract

This invention discloses a method, system, apparatus, and storage medium for feed formulation optimization. The method includes: determining feed formulation ingredients and corresponding constraints; constructing a feed formulation optimization problem model and determining the optimization objective of the model; processing the feed formulation optimization problem model using a distribution estimation algorithm with a construction method and a construction information update strategy to obtain the final feed formulation; wherein, the distribution estimation algorithm with a construction method, while satisfying various constraints, obtains multiple feed formulations with the highest nutritional value and the lowest raw material cost; the construction information update strategy can use feed formulation information from the parent population to update the construction information, guiding the constructive feed formulation generation method to generate feed formulations with higher nutritional value, lower raw material cost, and satisfying various constraints. This invention can efficiently solve feed formulation optimization problems with complex constraints and can be widely applied in the field of feed formulation.

Description

Technical Field

[0001] This invention relates to two major fields: feed formulation and evolutionary computation, and in particular to a method, system, device, and storage medium for feed formulation optimization. Background Technology

[0002] With the continuous improvement of people's living standards and the ongoing urbanization process in my country, animal products, represented by meat, eggs, and milk, have seen rapid growth as a rigid daily demand for the general public. Consequently, my country's livestock industry has also developed rapidly, gradually transforming towards standardized, large-scale, and mechanized farming. As a supporting industry for livestock farming, the feed industry is experiencing continuous market expansion and has significant development potential.

[0003] Feed refers to all artificially produced food for artificially raised animals, generally referring to artificially produced food used in agriculture and animal husbandry to feed fish, poultry, and livestock. Feed is a key factor affecting the output and feeding costs of farmed animals, and is related to the economic benefits of farmers. On the one hand, using high-quality feed rich in nutrients required by animals can effectively increase output, such as increasing egg production and meat yield, thereby improving economic benefits. On the other hand, high-quality feed is more expensive, which can reduce economic benefits. Therefore, how to control feed costs while meeting the nutritional needs of animals has become a priority problem for the feed industry.

[0004] Currently, there is no effective technical solution to address the above problems. Summary of the Invention

[0005] In order to at least partially solve one of the technical problems existing in the prior art, the present invention aims to provide a method, system, device and storage medium for optimizing feed formulation.

[0006] The technical solution adopted in this invention is:

[0007] A method for optimizing feed formulation includes the following steps:

[0008] Determine the feed formulation ingredients and the corresponding constraints for those ingredients;

[0009] Construct a model for feed formulation optimization and determine the optimization objective of the model;

[0010] A distribution estimation algorithm with a construction method and a construction information update strategy are used to process the feed formulation optimization problem model to obtain the final feed formulation.

[0011] The distribution estimation algorithm with construction method obtains multiple feed formulas with the highest nutritional value and the lowest raw material cost while satisfying various constraints. The construction information update strategy can use the feed formula information in the parent population to update the construction information, guiding the construction feed formula generation method to generate feed formulas with higher nutritional value, lower raw material cost, and satisfying various constraints.

[0012] Furthermore, the determination of feed formulation ingredients and the corresponding constraints for the ingredients includes:

[0013] Suppose there are N types of raw materials that can be used in the feed formulation, numbered {1,2,3,...,N}, and the raw materials are divided into Z categories, also numbered {1,2,3,...,Z}; the nutritional value of the i-th raw material is v. i The cost is c i It belongs to the category g i ∈{1,2,3,...,Z};

[0014] Single raw material ratio constraint: the usage ratio w of the i-th raw material i The constraint condition lb needs to be satisfied. i ≤w i ≤ub i (0≤lb i <ub i ≤1), where lb i ub is the lower limit of the proportion of raw material i. i This represents the upper limit of the proportion of raw material i;

[0015] Raw material category proportion constraint: materials belonging to the same raw material category G i Total proportion of raw materials used Constraints must be met Where lg i For raw material category G i The lower limit of the ratio, ug i For raw material category G i The upper limit of the proportion;

[0016] Raw material type constraint: The number of raw materials used in the feed formulation is K, where 1≤K≤N.

[0017] Furthermore, the construction of a feed formulation optimization problem model and the determination of the model's optimization objective include:

[0018] The problem of feed formulation optimization is to select K kinds of raw materials and determine the usage ratio of each raw material to form a feed formulation.

[0019] The optimization objectives are: to maximize the nutritional value V of the formula and minimize the raw material cost C of the formula, while satisfying the single raw material ratio constraint BC, the raw material category ratio constraint GC, and the constraint K on the number of raw materials used.

[0020] Let the vector formed by the selected K raw materials be: (a1, a2, a3, ..., a K )a i ∈N, the vector formed by the proportion of each raw material is: (w1,w2,w3,...,w K The combination of two vectors represents the use of ingredient a in feed formulation. i The ratio is w i Raw material a i The nutritional value is v i Its cost is c i ;

[0021] The problem of feed formulation optimization is represented as:

[0022]

[0023]

[0024]

[0025] Furthermore, the method employing a distribution estimation algorithm with a construction method and a construction information update strategy to process the feed formulation optimization problem model includes:

[0026] The feed formulation optimization problem is formulated as a cardinality-constrained combinatorial optimization problem with constraints on the proportion of a single raw material, the proportion of raw material categories, and the number of raw materials used. A constructive feed formulation generation method and a constructive proportion allocation method are used to generate feed formulations that satisfy all constraints. A population of distribution estimation algorithms is constructed, and non-dominated sorting is used to select the feed formulation that maximizes nutritional value and minimizes raw material costs.

[0027] Furthermore, the distribution estimation algorithm containing the construction method optimizes feed formulation in the following way:

[0028] A1. Initialize parameters; the initialization parameters include: single raw material ratio constraint BC, raw material category ratio constraint GC, number of raw material types constraint K, pheromone matrix P, ratio mean vector U, ratio standard deviation vector D, population size Size, and iteration number Gen.

[0029] A2. Generate the initial parent population: Based on the constructive feed formulation generation method, a valid solution is randomly generated to form the parent population S. The criteria for a valid solution are that it satisfies the single raw material ratio constraint, the raw material category ratio constraint, and the constraint of the number of raw materials used.

[0030] A3. Generating offspring population: Randomly generate valid solutions to construct the offspring population Q based on the constructive feed formulation generation method;

[0031] A4. Calculate the nutritional value V and raw material cost C of each solution representing the feed formula in the current parent and offspring populations.

[0032] A5. Merge the current parent population and offspring population into a set R, and perform non-dominated sorting on set R based on nutritional value and raw material cost to determine the non-dominated level of each solution. Select solutions in order of increasing non-dominated level to form a new parent population.

[0033] A6. Based on the raw material composition and proportion information of the solution in the new parent population, update the pheromone matrix, the proportion mean vector, and the proportion standard deviation vector.

[0034] A7. If the termination condition is met, the optimization program ends and outputs the feed formula represented by the solution in the current parent population; otherwise, return to step A3.

[0035] Furthermore, the pheromone matrix P is used to characterize the correlation strength among the N raw materials, and is expressed as:

[0036]

[0037] The initial value of the pheromone on the off-diagonal lines of the matrix is ​​1, i.e. The pheromone on the diagonal is equal to the mean of the L largest pheromones in that row, i.e.

[0038] The proportional mean vector U is used to represent the mean of the usage ratio of each raw material. The initial mean of each raw material is the median of the upper and lower bounds of the usage ratio of each raw material, i.e.

[0039] The proportional standard deviation vector D is used to characterize the distribution of the usage proportion of each raw material. The initial standard deviation of each raw material is half the length of the usage proportion interval for that raw material.

[0040] Furthermore, updating the pheromone matrix, the proportion mean vector, and the proportion standard deviation vector based on the raw material composition and proportion information of the solutions in the new parent population includes:

[0041] The pheromone matrix P is updated in the following way:

[0042]

[0043]

[0044] In the formula, ρ is the volatility coefficient, and τ min=0 is the lower bound of pheromones, τ max =1 is the upper bound of pheromone. The updated associated pheromone size needs to be between the upper and lower bounds, while the diagonal pheromone is equal to the average of the L largest pheromones in the row.

[0045] The update method for the proportional mean vector U and the proportional standard deviation vector D is as follows:

[0046]

[0047]

[0048] In the formula, η is the smoothing coefficient, and H is the number of times the raw material appears in the current parent population S. If a certain raw material is not selected in any solution, i.e., H = 0, then the mean converges to the lower bound of the single raw material ratio constraint of that raw material, and the standard deviation converges to the length of the single raw material ratio constraint interval of that raw material. If the number of times a certain raw material appears in the current parent population S is greater than or equal to 1, i.e., H ≥ 1, then the mean and standard deviation of all usage ratios of that raw material are calculated, and ε is a sufficiently small constant to prevent the standard deviation from converging to 0.

[0049] Another technical solution adopted in this invention is:

[0050] A feed formulation optimization system, comprising:

[0051] The constraint determination module is used to determine the feed formulation ingredients and the corresponding constraints for those ingredients.

[0052] The optimization problem building module is used to construct a feed formulation optimization problem model and determine the optimization objective of the model;

[0053] The feed formulation optimization module is used to process the feed formulation optimization problem model using a distribution estimation algorithm with a construction method and a construction information update strategy to obtain the final feed formulation.

[0054] The distribution estimation algorithm with construction method obtains multiple feed formulas with the highest nutritional value and the lowest raw material cost while satisfying various constraints. The construction information update strategy can use the feed formula information in the parent population to update the construction information, guiding the construction feed formula generation method to generate feed formulas with higher nutritional value, lower raw material cost, and satisfying various constraints.

[0055] Another technical solution adopted in this invention is:

[0056] A feed formulation optimization device, comprising:

[0057] At least one processor;

[0058] At least one memory for storing at least one program;

[0059] When the at least one program is executed by the at least one processor, the at least one processor implements the method described above.

[0060] Another technical solution adopted in this invention is:

[0061] A computer-readable storage medium storing a processor-executable program, which, when executed by a processor, performs the method described above.

[0062] The beneficial effects of this invention are: this invention designs a constructive feed formulation generation method to handle complex constraints, thereby ensuring the generation of legal solutions that satisfy the constraints; at the same time, it optimizes two objectives, namely nutritional value and raw material cost, thereby providing feasible solutions for feed mills to choose from with different combinations of nutritional value and raw material cost, which can efficiently solve the problem of feed formulation optimization with complex constraints. Attached Figure Description

[0063] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following description is provided with accompanying drawings of the relevant technical solutions in the embodiments of the present invention or the prior art. It should be understood that the accompanying drawings described below are only for the purpose of clearly illustrating some embodiments of the technical solutions of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0064] Figure 1 This is a flowchart illustrating the steps of a feed formulation optimization method according to an embodiment of the present invention;

[0065] Figure 2 This is a schematic diagram illustrating the feed formulation optimization problem in an embodiment of the present invention;

[0066] Figure 3 This is a flowchart of the constructive distribution estimation algorithm in an embodiment of the present invention. Detailed Implementation

[0067] The embodiments of the present invention are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention. The step numbers in the following embodiments are set only for ease of explanation, and there is no limitation on the order between the steps. The execution order of each step in the embodiments can be adaptively adjusted according to the understanding of those skilled in the art.

[0068] In the description of this invention, it should be understood that the orientation descriptions, such as up, down, front, back, left, right, etc., are based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limiting this invention.

[0069] In the description of this invention, "several" means one or more, "more than" means two or more, "greater than," "less than," and "exceeding" are understood to exclude the stated number, while "above," "below," and "within" are understood to include the stated number. The use of "first" and "second" in the description is merely for distinguishing technical features and should not be construed as indicating or implying relative importance, or implicitly indicating the number of indicated technical features, or implicitly indicating the order of the indicated technical features.

[0070] In the description of this invention, unless otherwise explicitly defined, terms such as "set up," "install," and "connect" should be interpreted broadly, and those skilled in the art can reasonably determine the specific meaning of the above terms in this invention in conjunction with the specific content of the technical solution.

[0071] The nutritional composition of feed formulations needs to be adjusted according to different animal breeds, growth stages, and seasons. However, the basic components of feed formulations are the same, divided into five categories: energy, protein, minerals, vitamins, and additives. Each category has a wide range of raw materials to choose from: energy sources include corn, barley, wheat bran, and oil; protein sources include fishmeal, soybean meal, peanut meal, and yeast powder; mineral sources include calcium carbonate, salt, potassium iodide, and dicalcium phosphate; vitamins include vitamins A, D, E, and B; and additives include lysine, methionine, medications, and preservatives. Faced with a multitude of feed ingredients, feed mills cannot use every single one. A feasible approach is to comprehensively consider nutritional value and raw material costs, selecting a limited number of ingredients to formulate the feed for production.

[0072] Feed formulation optimization can be structured as a combinatorial optimization problem with cardinality constraints, a type of NP-hard problem. It involves selecting a finite number of raw materials and determining the proportion of each material to maximize the nutritional value of the feed while minimizing the cost of the raw materials. These problems are difficult to solve using traditional methods, but evolutionary computation methods are suitable for solving NP-hard problems. With the help of approximate search and positive / negative feedback mechanisms, a good solution can be found within an acceptable timeframe. Furthermore, feed formulation optimization faces complex constraints. The proportion of each major category of raw materials must fall within a given range, for example, energy-rich materials account for 50-60% in pig feed. Simultaneously, the proportion of each individual raw material must also fall within a given range, for example, corn accounts for 0-50% in pig feed. These constraints on the proportion of individual raw materials and the proportion of major categories of raw materials further increase the difficulty of solving the feed formulation optimization problem, requiring the design of appropriate constraint handling mechanisms to ensure that the feed formulation meets the constraints.

[0073] like Figure 1 As shown, this embodiment provides a method for optimizing feed formulation, including the following steps:

[0074] S1. Determine the feed formulation ingredients and the corresponding constraints for the ingredients.

[0075] S2. Construct a model for feed formulation optimization and determine the optimization objective of the model.

[0076] S3. A distribution estimation algorithm with a construction method and a construction information update strategy are used to process the feed formulation optimization problem model to obtain the final feed formulation. The distribution estimation algorithm with a construction method obtains multiple feed formulations with the highest nutritional value and the lowest raw material cost while satisfying various constraints. The construction information update strategy can use the feed formulation information in the parent population to update the construction information, guiding the construction feed formulation generation method to generate feed formulations with higher nutritional value, lower raw material cost, and satisfying various constraints.

[0077] As an optional implementation, the constructive distribution estimation algorithm in this embodiment is a distribution estimation algorithm applicable to combinatorial optimization problems with cardinality constraints. It is also a general framework where the evaluation function can be customized according to the characteristics of the problem. The steps of the algorithm in optimizing feed formulation are as follows:

[0078] (1) Define the problem clearly. The decision variables of the problem include a vector consisting of a set of selected raw materials and a vector consisting of the proportion of each raw material. The constraints are the proportion of a single raw material, the proportion of a major category of raw materials, and the number of types of raw materials used. The optimization objective function is to maximize the nutritional value of the feed and minimize the cost of raw materials.

[0079] (2) Initialize parameters. The parameters that need to be initialized include: single raw material ratio constraint BC, raw material category ratio constraint GC, number of raw material types constraint K, pheromone matrix P, ratio mean vector U, ratio standard deviation vector D, population size Size, and iteration number Gen.

[0080] (3) Generate the initial parent population. Based on the constructive feed formulation generation method, a valid solution is randomly generated to form the parent population S. The criteria for a valid solution are that it satisfies the single raw material ratio constraint, the raw material category ratio constraint, and the constraint on the number of raw materials used.

[0081] (4) Generate offspring population. Based on the constructive feed formulation generation method, legal solutions are randomly generated to form the offspring population Q.

[0082] (5) Evaluation function calculation. Calculate the nutritional value V and raw material cost C of the feed formulation represented by each solution in the current parent and offspring populations.

[0083] (6) Select a new parent population. Merge the current parent population and the offspring population into a set R, and perform non-dominated sorting on set R based on nutritional value and raw material cost to determine the non-dominated level of each solution. Select solutions in order of increasing non-dominated level to form a new parent population.

[0084] (7) Update the construction information. Based on the raw material composition and proportion information of the solution in the new parent population, update the pheromone matrix, the proportion mean vector, and the proportion standard deviation vector.

[0085] (8) Termination condition judgment. If the termination condition is met, the optimization program ends and outputs the feed formula represented by the solution in the current parent population; otherwise, return to step (4).

[0086] The above method will be explained in detail below with reference to the accompanying drawings and specific embodiments.

[0087] Step 1: Identify feed formulation optimization issues

[0088] See Figure 2 , Figure 2 This is a schematic diagram of feed formulation composition. Feed formulation ingredients are divided into five main categories: energy sources, proteins, minerals, vitamins, and additives. Each category contains multiple optional ingredients. The optimization of feed formulation is discussed in detail below:

[0089] (1) There are N types of raw materials that can be used in the feed formulation, corresponding to the numbers {1,2,3,...,N}. These raw materials are divided into Z categories, corresponding to the numbers {1,2,3,...,Z}. The nutritional value of the i-th raw material is v. i The cost is c i It belongs to the category g i∈{1,2,3,...,Z}.

[0090] (2) Single raw material ratio constraint: the usage ratio w of the i-th raw material i The constraint condition lb needs to be satisfied. i ≤w i ≤ub i (0≤lb i <ub i ≤1), where lb i ub is the lower limit of the proportion of raw material i. i This represents the upper limit of the proportion of raw material i.

[0091] (3) Raw material category proportion constraint: materials belonging to the same raw material category G i Total proportion of raw materials used Constraints must be met Where lg i For raw material category G i The lower limit of the ratio, ug i For raw material category G i The upper limit of the proportion.

[0092] (4) Constraint on the number of raw materials used: The number of raw materials used in the feed formula is K, where 1≤K≤N.

[0093] Step 2: Build a problem model for feed formulation optimization

[0094] The feed formulation optimization problem involves selecting K raw materials and determining the proportion of each material used to form a feed formulation. The optimization objective is to maximize the nutritional value (V) of the formulation and minimize the raw material cost (C), while simultaneously satisfying constraints on the proportion of individual raw materials (BC), the proportion of raw material categories (GC), and the number of raw materials used (K). The problem model is described as follows:

[0095] Let the vector formed by the selected K raw materials be: (a1, a2, a3, ..., a K )a i ∈N, the vector formed by the proportion of each raw material is: (w1,w2,w3,...,w K The combination of two vectors represents the use of ingredient a in feed formulation. i The ratio is w i Raw material a i The nutritional value is v i Its cost is c i The feed formulation optimization problem is represented as:

[0096]

[0097]

[0098]

[0099] This problem is a combinatorial optimization problem with cardinality constraints. It is an NP-hard problem and is suitable for solving using evolutionary computation methods.

[0100] Step 3: Constructed Feed Formulation Method

[0101] (1) Selection of raw materials for structural design

[0102] The construction formula first randomly selects two raw materials i and j from N raw materials, and compares the magnitude of their diagonal pheromone values. If τ ii ≥τ jj If τ , then raw material i is selected as raw material a1 and raw material i is set as selected; otherwise, raw material j is selected as raw material a1 and raw material j is set as selected. Then, two more raw materials i and j are randomly selected from the unselected raw materials, and their correlation pheromone levels with raw material a1 are compared. If τ a1i ≥τ a1j If material i is selected as material a2, then material i is set as selected; otherwise, material j is selected as material a2 and material j is set as selected. Repeat the above steps, randomly selecting two more materials from the unselected materials, comparing their pheromone levels with the previously selected material, and selecting the material with the higher pheromone level as the new selected material, until K materials have been selected, forming a material set vector (a1, a2, a3, ..., a...). K To ensure that the constructed raw material set meets the constraints, it is necessary to check the single raw material ratio constraint and the raw material category ratio constraint. If the check passes and there is a feasible ratio allocation space, proceed to the next step of constructive ratio allocation; otherwise, reconstruct the raw material set.

[0103] (2) Constructive proportional distribution

[0104] The raw material set vector (a1, a2, a3, ..., a) is obtained by constructively selecting raw materials. K After that, the usage ratio of each raw material is determined sequentially. Before allocating a ratio to the i-th raw material, the remaining Ki raw materials and their constraints are considered as a smaller-scale raw material allocation problem containing only Ki raw materials. Then, the upper limit Up and lower limit Down of the allocable ratio of the i-th raw material are calculated, assuming a feasible allocation space exists for this smaller-scale raw material allocation problem. Subsequently, the average usage ratio of the i-th raw material is used as the basis for the allocation. with standard deviation Generate a normally distributed random number W between the interval [0,1]. If W is greater than the upper limit Up, set W = Up; if W is less than the lower limit Down, set W = Down. Then, use W as the usage ratio w for the i-th raw material.i Repeat the above steps to allocate the usage ratios of the first K-1 raw materials, and then allocate the usage ratio w of the last raw material K. K Set as By constructing a proportional allocation formula, the raw material proportion vector (w1, w2, w3, ..., w) can be obtained. K ).

[0105] The constructive feed formulation generation method obtains a vector of feed formulation raw material sets (a1, a2, a3, ..., a) through constructive raw material selection and constructive proportion allocation. K The ratio vector of the raw materials is (w1, w2, w3, ..., w K It can satisfy the single raw material ratio constraint BC, the raw material category ratio constraint GC, and the constraint K on the number of raw materials used.

[0106] Step 4: Based on the constructive distribution estimation algorithm

[0107] See Figure 3 , Figure 3 This is a schematic diagram of a constructive distribution estimation algorithm. Details are as follows:

[0108] (1) Initialize parameters

[0109] The problem parameters that need to be initialized include: single raw material ratio constraint BC, raw material category ratio constraint GC, number of raw material types constraint K, population size Size, and number of iterations Gen.

[0110] The pheromone matrix P is used to characterize the correlation strength between N raw materials, and is expressed as:

[0111]

[0112] The initial value of the pheromone on the off-diagonal lines of the matrix is ​​1, i.e. The pheromone on the diagonal is equal to the mean of the L largest pheromones in that row, i.e.

[0113] The proportional mean vector U is used to represent the mean of the usage proportion of each raw material. The initial mean of each raw material is the median of the upper and lower bounds of the usage proportion of each raw material, i.e. The proportional standard deviation vector D is used to characterize the distribution of the usage proportion of each raw material. The initial standard deviation of each raw material is half the length of the usage proportion interval for that raw material, i.e.

[0114]

[0115] (2) Generate the initial parent population

[0116] Using the constructive feed formulation generation method, we randomly generate Size feed formulations that meet the constraints using the initial pheromone matrix P, the initial proportion mean vector U, and the initial proportion standard deviation vector D to form the parent population S.

[0117] (3) Generate offspring population

[0118] Using a constructive feed formulation generation method, Size feed formulations that meet the constraints are randomly generated using the latest pheromone matrix P, the latest proportion mean vector U, and the latest proportion standard deviation vector D to form the offspring population Q.

[0119] (4) Evaluation function calculation

[0120] Calculate the nutritional value and raw material cost of each feed formulation in the current parent population S and offspring population Q.

[0121] (5) Select a new parent population. Merge the current parent population and the offspring population into a set R, and sort the set R based on the nutritional value V and the raw material cost C. Determine the non-dominated rank of each solution, and select Size solutions in order of increasing non-dominated rank to form a new parent population S.

[0122] (6) Update construction information

[0123] Based on the raw material composition and proportion information of the solutions in the new parent population S, update the information group matrix, the proportion mean vector, and the proportion standard deviation vector.

[0124] The pheromone matrix P is updated by selecting solutions from the new parent population S whose non-dominant level is less than or equal to the threshold Th, and calculating the increment of the associated pheromone between each pair of raw materials based on the types and proportions of raw materials selected in these solutions. Where ξ is the incremental intensity, and the specific calculation method is as follows:

[0125]

[0126] Then, each associated pheromone in the pheromone matrix P is updated, and the specific calculation method is as follows:

[0127]

[0128]

[0129] Where ρ is the volatility coefficient, τ min =0 is the lower bound of pheromones, τ max =1 is the upper bound of pheromones. The updated associated pheromone size needs to be between the upper and lower bounds, while the diagonal pheromone is still equal to the average of the L largest pheromones in the row.

[0130] The update method for the proportional mean vector U and the proportional standard deviation vector D is as follows:

[0131]

[0132]

[0133] Where η is the smoothing coefficient, and H is the number of times a raw material appears in the current parent population S. If a certain raw material is not selected in any solution, i.e., H = 0, then the mean converges to the lower bound of the single raw material proportion constraint, and the standard deviation converges to the length of the single raw material proportion constraint interval. If a certain raw material appears more than or equal to 1 time in the current parent population S, i.e., H ≥ 1, then the mean and standard deviation of all usage proportions of that raw material are calculated, and ε is a sufficiently small constant to prevent the standard deviation from converging to 0.

[0134] (7) Termination condition judgment

[0135] Determine whether the algorithm iteration count Gen has been reached. If it has, the optimization program ends and the feed formula represented by the solution in the current parent population is output. Otherwise, return to step (3).

[0136] As shown above, feed formulation optimization is an NP-hard problem with complex constraints. Traditional evolutionary computation methods often use repair or penalty mechanisms to handle these constraints. However, when the constraints are complex, these mechanisms suffer from low efficiency or fail to produce valid solutions. This invention aims to solve feed formulation optimization problems with constraints such as single ingredient ratios, ingredient category ratios, and the number of ingredient types used. Existing algorithms urgently need improvement to adapt to these problem characteristics. Therefore, this invention designs a constructive feed formulation generation method to handle complex constraints, thereby ensuring the generation of valid solutions that satisfy the constraints. Subsequently, a multi-objective optimization algorithm framework is used to simultaneously optimize both nutritional value and ingredient cost, providing feasible solutions with different combinations of nutritional value and ingredient cost for feed mills to choose from, thus efficiently solving feed formulation optimization problems with complex constraints.

[0137] The constructive distribution estimation algorithm in this invention is an effective solution method for feed formulation optimization problems. Using this method, multiple feed formulations with the highest nutritional value and lowest raw material cost can be obtained within an acceptable time range while satisfying various constraints, allowing feed mills to choose the formulations for production, thereby maximizing economic benefits. It has a very wide range of application prospects.

[0138] This embodiment also provides a feed formulation optimization system, including:

[0139] The constraint determination module is used to determine the feed formulation ingredients and the corresponding constraints for those ingredients.

[0140] The optimization problem building module is used to construct a feed formulation optimization problem model and determine the optimization objective of the model;

[0141] The feed formulation optimization module is used to process the feed formulation optimization problem model using a distribution estimation algorithm with a construction method and a construction information update strategy to obtain the final feed formulation.

[0142] The distribution estimation algorithm with construction method obtains multiple feed formulas with the highest nutritional value and the lowest raw material cost while satisfying various constraints. The construction information update strategy can use the feed formula information in the parent population to update the construction information, guiding the construction feed formula generation method to generate feed formulas with higher nutritional value, lower raw material cost, and satisfying various constraints.

[0143] This embodiment of the feed formulation optimization system can execute the feed formulation optimization method provided in the method embodiment of the present invention, and can execute any combination of the implementation steps of the method embodiment, and has the corresponding functions and beneficial effects of the method.

[0144] This embodiment also provides a feed formulation optimization device, including:

[0145] At least one processor;

[0146] At least one memory for storing at least one program;

[0147] When the at least one program is executed by the at least one processor, the at least one processor implements the method described above.

[0148] This embodiment of the feed formulation optimization system can execute the feed formulation optimization method provided in the method embodiment of the present invention, and can execute any combination of the implementation steps of the method embodiment, and has the corresponding functions and beneficial effects of the method.

[0149] This application also discloses a computer program product or computer program, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device can read the computer instructions from the computer-readable storage medium and execute the computer instructions, causing the computer device to perform... Figure 1 The method shown.

[0150] This embodiment also provides a storage medium storing instructions or programs that can execute the feed formulation optimization method provided in the method embodiment of the present invention. When the instructions or programs are run, any combination of implementation steps of the method embodiment can be executed, and the method has the corresponding functions and beneficial effects.

[0151] In some alternative embodiments, the functions / operations mentioned in the block diagrams may not occur in the order shown in the operation diagrams. For example, depending on the functions / operations involved, two consecutively shown blocks may actually be executed substantially simultaneously, or the blocks may sometimes be executed in reverse order. Furthermore, the embodiments presented and described in the flowcharts of this invention are provided by way of example to provide a more comprehensive understanding of the technology. The disclosed methods are not limited to the operations and logic flows presented herein. Alternative embodiments are contemplated in which the order of various operations is altered and sub-operations described as part of a larger operation are executed independently.

[0152] Furthermore, although the invention has been described in the context of functional modules, it should be understood that, unless otherwise stated, one or more of the described functions and / or features may be integrated into a single physical device and / or software module, or one or more functions and / or features may be implemented in a separate physical device or software module. It is also understood that a detailed discussion of the actual implementation of each module is unnecessary for understanding the invention. Rather, given the properties, functions, and internal relationships of the various functional modules in the apparatus disclosed herein, the actual implementation of the module will be understood within the scope of conventional skill of an engineer. Therefore, those skilled in the art can implement the invention as set forth in the claims using ordinary techniques without excessive experimentation. It is also understood that the specific concepts disclosed are merely illustrative and not intended to limit the scope of the invention, which is determined by the full scope of the appended claims and their equivalents.

[0153] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, essentially, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0154] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0155] More specific examples of computer-readable media (a non-exhaustive list) include: electrical connections (electronic devices) having one or more wires, portable computer disk drives (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Furthermore, computer-readable media can even be paper or other suitable media on which the program can be printed, since the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in computer memory.

[0156] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0157] In the foregoing description of this specification, references to terms such as "one embodiment," "another embodiment," or "some embodiments" indicate that a specific feature, structure, material, or characteristic described in connection with an embodiment or example is included in at least one embodiment or example of the present invention. In this specification, illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0158] Although embodiments of the invention have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.

[0159] The above is a detailed description of the preferred embodiments of the present invention. However, the present invention is not limited to the above embodiments. Those skilled in the art can make various equivalent modifications or substitutions without departing from the spirit of the present invention. All such equivalent modifications or substitutions are included within the scope defined by the claims of this application.

Claims

1. A method for optimizing feed formulation, characterized in that, Includes the following steps: Determine the feed formulation ingredients and the corresponding constraints for those ingredients; Construct a model for feed formulation optimization and determine the optimization objective of the model; A distribution estimation algorithm with a construction method and a construction information update strategy are used to process the feed formulation optimization problem model to obtain the final feed formulation. The distribution estimation algorithm with construction method obtains multiple feed formulas with the highest nutritional value and the lowest raw material cost while satisfying various constraints. The construction information update strategy can use the feed formula information in the parent population to update the construction information, guiding the construction feed formula generation method to generate feed formulas with higher nutritional value, lower raw material cost, and satisfying various constraints. The method employs a distribution estimation algorithm with a construction method and a construction information update strategy to process the feed formulation optimization problem model, including: The feed formulation optimization problem is formulated as a combinatorial optimization problem with cardinality constraints, including constraints on the proportion of a single raw material, the proportion of a major category of raw materials, and the number of types of raw materials used. A constructive feed formulation generation method and a constructive proportion allocation method are used to generate feed formulations that satisfy all constraints. A population of distribution estimation algorithms is constructed, and non-dominated sorting is used to select the feed formulation that maximizes nutritional value and minimizes raw material cost. The distribution estimation algorithm containing the construction method optimizes feed formulation in the following ways: A1. Initialize parameters; the parameters to be initialized include: single raw material ratio constraints. Raw material category ratio constraints Constraints on the number of raw material types used pheromone matrix , proportional mean vector , Scale standard deviation vector Population size Number of iterations ; A2. Generating the initial parent population: Based on the constructive feed formulation generation method, legal solutions are randomly generated to form the parent population. The criteria for a valid solution are that it satisfies the single raw material ratio constraint, the raw material category ratio constraint, and the constraint on the number of raw materials used. A3. Generating offspring population: Randomly generating valid solutions to construct the offspring population based on the constructive feed formulation generation method. ; A4. Calculate the nutritional value of the feed formulation represented by each solution in the current parent and offspring populations. With raw material costs ; A5. Merge the current parent population and the child population into a set. and for the set Based on nutritional value and raw material cost, non-dominated sorting is performed to determine the non-dominated level of each solution. Solutions are selected in order of increasing non-dominated level to form a new parent population. A6. Based on the raw material composition and proportion information of the solution in the new parent population, update the pheromone matrix, the proportion mean vector, and the proportion standard deviation vector. A7. If the termination condition is met, the optimization program ends and outputs the feed formula represented by the solution in the current parent population; otherwise, return to step A3. The pheromone matrix Used for characterization The strength of the correlation between the raw materials is expressed as: The initial value of the pheromone on the off-diagonal lines of the matrix is ​​1, i.e. The pheromone on the diagonal is equal to the largest in its row. The mean of each pheromone, i.e. ; The proportional mean vector The mean value is used to characterize the proportion of each raw material used. The initial mean value for each raw material is the median of the upper and lower bounds of the proportion of each raw material used. ; The proportional standard deviation vector The distribution of the usage proportion of each raw material is used to characterize the distribution. The initial standard deviation of each raw material is half the length of the usage proportion interval for that raw material. .

2. The feed formulation optimization method according to claim 1, characterized in that, The determination of feed formulation ingredients and the corresponding constraints for those ingredients includes: Assume there are a total of raw materials that can be selected for feed formulation. Species, corresponding number The raw materials are divided into a total of There are 1 major categories, with corresponding numbers as follows: ;No. The nutritional value of the raw materials is The cost is The major category is ; Single raw material ratio constraint: The proportion of raw materials used Constraints must be met ,in as raw materials The lower limit of the proportion, as raw materials The upper limit of the proportion; Raw material category proportion constraint: materials must belong to the same raw material category Total proportion of raw materials used Constraints must be met ,in Raw materials The lower limit of the proportion, Raw materials The upper limit of the proportion; Constraints on the Number of Raw Materials Used: The number of raw materials used in the feed formulation is [number to be specified]. Species, among which .

3. The feed formulation optimization method according to claim 1, characterized in that, The construction of a feed formulation optimization problem model and the determination of the model's optimization objective include: Feed formulation optimization is a matter of selection The feed formula is composed of a variety of raw materials and the proportion of each raw material used. The optimization goal is to maximize the nutritional value of the formula. Minimize the raw material cost of the formula At the same time, it is necessary to meet the single raw material ratio constraint condition. Raw material category ratio constraints Constraints on the number of raw material types used ; Set the selected The vector formed by the set of raw materials is: The vector formed by the proportion of each raw material is: The combination of two vectors represents the raw materials used in feed formulation. The proportion is ,raw material Its nutritional value is Its cost is ; The problem of feed formulation optimization is represented as: 。 4. The feed formulation optimization method according to claim 1, characterized in that, The step of updating the pheromone matrix, the proportion mean vector, and the proportion standard deviation vector based on the raw material composition and proportion information of the solutions in the new parent population includes: The pheromone matrix The update method is as follows: In the formula, The volatility coefficient is... The lower bound of pheromones The upper bound of pheromones is defined as the upper limit, and the updated associated pheromone size must fall between the upper and lower bounds. The diagonal pheromone is equal to the largest pheromone in its row. The mean of each pheromone; The proportional mean vector with the standard deviation vector The update method is as follows: In the formula, For smoothing coefficients, As raw materials in the current parent population The number of times a certain raw material appears in all solutions; if a certain raw material is not selected in any solution, that is... Then the mean converges to the lower bound of the single-material proportion constraint of that raw material, and the standard deviation converges to the length of the single-material proportion constraint interval of that raw material; if a certain raw material in the current parent population The number of times it appears is greater than or equal to 1, that is Then calculate the mean and standard deviation of all usage proportions of this raw material. It is a sufficiently small constant to prevent the standard deviation from converging to 0.

5. A feed formulation optimization system for implementing the method according to any one of claims 1-4, characterized in that, include: The constraint determination module is used to determine the feed formulation ingredients and the corresponding constraints for those ingredients. The optimization problem building module is used to construct a feed formulation optimization problem model and determine the optimization objective of the model; The feed formulation optimization module is used to process the feed formulation optimization problem model using a distribution estimation algorithm with a construction method and a construction information update strategy to obtain the final feed formulation. The distribution estimation algorithm with construction method obtains multiple feed formulas with the highest nutritional value and the lowest raw material cost while satisfying various constraints. The construction information update strategy can use the feed formula information in the parent population to update the construction information, guiding the construction feed formula generation method to generate feed formulas with higher nutritional value, lower raw material cost, and satisfying various constraints.

6. A feed formulation optimization device, characterized in that, include: At least one processor; At least one memory for storing at least one program; When the at least one program is executed by the at least one processor, the at least one processor implements the method of any one of claims 1-4.

7. A computer-readable storage medium storing a processor-executable program, characterized in that, The processor-executable program, when executed by the processor, is used to perform the method as described in any one of claims 1-4.

Images