Search

CN-115809556-B - Feed formula optimization method, system, device and storage medium

CN115809556BCN 115809556 BCN115809556 BCN 115809556BCN-115809556-B

Abstract

The invention discloses a feed formula optimization method, a system, a device and a storage medium, wherein the method comprises the steps of determining feed formula raw materials and constraint conditions corresponding to the raw materials, constructing a feed formula optimization problem model, determining an optimization target of the model, processing the feed formula optimization problem model by adopting a distribution estimation algorithm containing a construction method and a construction information updating strategy, and obtaining a final feed formula, wherein the distribution estimation algorithm containing the construction method obtains a plurality of feed formulas with highest nutritive value and lowest raw material cost on the basis of meeting constraint conditions, and the construction information updating strategy can update construction information by utilizing feed formula information in a father population, so that the construction feed formula generation method is guided to generate feed formulas with higher nutritive value and lower raw material cost and meets all constraint conditions. The invention can efficiently solve the problem of optimizing the feed formula with complex constraint conditions, and can be widely applied to the field of feed formulas.

Inventors

  • CHEN WEINENG
  • ZHANG ZHIXUAN
  • ZHAN ZHIHUI

Assignees

  • 华南理工大学

Dates

Publication Date
20260505
Application Date
20221206

Claims (7)

  1. 1. The feed formula optimizing method is characterized by comprising the following steps of: Determining raw materials of a feed formula and constraint conditions corresponding to the raw materials; constructing a feed formula optimization problem model, and determining an optimization target of the model; Adopting a distribution estimation algorithm containing a construction method and a construction information updating strategy to process a feed formula optimization problem model to obtain a final feed formula; the construction information updating strategy can update construction information by utilizing feed formula information in a parent population, and guide a construction feed formula generation method to generate a feed formula with higher nutritive value and lower raw material cost and meeting various constraint conditions; The method for processing the feed formula optimization problem model by adopting a distribution estimation algorithm containing a construction method and a construction information updating strategy comprises the following steps: expressing the feed formula optimization problem as a combined optimization problem with single raw material proportion constraint, raw material large proportion example constraint and base number constraint using raw material seed number constraint, adopting a structural feed formula generation method and structural proportion distribution to generate a feed formula meeting all constraints, constructing a population of a distribution estimation algorithm, and adopting non-dominant sorting to select a feed formula with maximized nutritive value and minimized raw material cost; the distribution estimation algorithm of the building method optimizes the feed formulation by: a1, initializing parameters, wherein the initialized parameters comprise single raw material proportion constraint conditions Example constraint conditions for large scale of raw materials Constraint of the number of raw materials used Pheromone matrix Proportional mean vector Standard deviation vector of proportion Population size Number of iterations ; A2, generating an initial parent population, namely randomly generating legal solutions to form the parent population based on a structural feed formula generation method The legal solution standard is to meet the constraint condition of single raw material proportion, the constraint condition of raw material large class example and the constraint of the number of used raw materials; a3, generating a offspring population, namely randomly generating legal solutions based on a structural feed formula generation method to form the offspring population ; A4, calculating the nutritive value of each solution representing the feed formula in the current parent population and offspring population And raw material cost ; A5, merging the current parent population and the offspring population into a set And pair sets Determining the non-dominant grade of each solution based on the nutrition value and the raw material cost, and sequentially selecting solutions from small to large according to the non-dominant grade to form a new parent population; A6, updating the pheromone matrix, the proportional mean value vector and the proportional standard deviation vector according to the raw material composition and the proportional information of the solution in the new parent population; A7, if the ending condition is reached, ending the optimization program, outputting a feed formula represented by a solution in the current parent population, otherwise, returning to the step A3; The pheromone matrix For characterising The strength of the association between the raw materials is expressed as: the initial value of the pheromone on the non-diagonal line in the matrix is 1, namely The pheromone on the diagonal is equal to the maximum line Mean value of individual pheromones, i.e. ; The proportional mean vector For characterizing the average value of the usage proportion of each raw material, the initial average value of each raw material is the median value of the upper and lower limits of the usage proportion of each raw material, namely ; The proportional standard deviation vector For characterizing the distribution of the proportion of each raw material, the initial standard deviation of each raw material is half of the length of the interval of the proportion of each raw material, i.e 。
  2. 2. The method for optimizing a feed formulation according to claim 1, wherein the determining the raw materials of the feed formulation and the corresponding constraints of the raw materials comprises: The raw materials selected by the feed formulation are shared The seeds are correspondingly numbered The raw materials are divided into Are of large class and are correspondingly numbered First, a third step The nutritive value of the seed raw material is The cost is The general category is ; Constraint of single raw material proportion Seed raw material ratio Needs to satisfy constraint conditions Wherein Is made of raw materials Is defined as the lower limit of the ratio of (c), Is made of raw materials Upper limit of the ratio of (2); the raw material large class example constraint belongs to the same raw material large class Is used in total proportion Needs to satisfy constraint conditions Wherein Is of the general class of raw materials Is defined as the lower limit of the ratio of (c), Is of the general class of raw materials Upper limit of the ratio of (2); the number of raw materials used is constrained by the number of raw materials used in the feed formula Seed of (a) wherein 。
  3. 3. The method of claim 1, wherein constructing a feed formulation optimization problem model and determining an optimization objective for the model comprises: The optimization problem of the feed formula is to select The raw materials are planted, and the use proportion of each raw material is determined to form a feed formula; the optimization goal is to maximize the nutritional value of the formulation Minimizing raw material cost of the formulation At the same time, the constraint condition of single raw material proportion needs to be satisfied Example constraint conditions for large scale of raw materials Constraint of the number of raw materials used ; In the design The vector formed by the raw material sets is as follows: the vector formed by the proportions of each raw material is as follows: the combination of two vectors represents the use of raw materials in a feed formulation The ratio is as follows Raw materials Has a nutritional value of The cost is ; The feed formulation optimization problem is expressed as: 。
  4. 4. The method of claim 1, wherein updating the pheromone matrix, the proportional mean vector and the proportional standard deviation vector based on the raw material composition and the proportional information of the solutions in the new parent population comprises: The pheromone matrix The updating mode of (a) is as follows: In the formula, In order to obtain the volatility coefficient, Is the lower bound of the pheromone, For the upper bound of pheromones, the updated associated pheromone size needs to be located between the upper and lower bounds, while the diagonal pheromone is equal to the maximum line An average value of the individual pheromones; the proportional mean vector And proportional standard deviation vector The updating mode of (a) is as follows: In the formula, In order to smooth the coefficient of the coefficient, Is the current parent population of raw materials If a material is not selected in all solutions, i.e. The average value converges to the lower bound of the single raw material ratio constraint of the raw material, the standard deviation converges to the length of the single raw material ratio constraint interval of the raw material, and if a certain raw material is in the current parent population The number of occurrences is 1 or more, i.e The mean value and standard deviation of all the use ratios of the raw materials are calculated, Is a sufficiently small constant to prevent the standard deviation from converging to 0.
  5. 5. A feed formulation optimisation system for carrying out the method of any one of claims 1 to 4, comprising: The constraint condition determining module is used for determining raw materials of the feed formula and constraint conditions corresponding to the raw materials; the optimization problem building module is used for building a feed formula optimization problem model and determining an optimization target of the model; The feed formula optimization module is used for processing the feed formula optimization problem model by adopting a distribution estimation algorithm containing a construction method and a construction information updating strategy to obtain a final feed formula; The construction information updating strategy can update construction information by utilizing feed formula information in a parent population, and guide a construction feed formula generation method to generate a feed formula with higher nutritive value and lower raw material cost and meeting various constraint conditions.
  6. 6. A feed formula optimizing device, comprising: At least one processor; At least one memory for storing at least one program; The at least one program, when executed by the at least one processor, causes the at least one processor to implement the method of any one of claims 1-4.
  7. 7. A computer readable storage medium, in which a processor executable program is stored, characterized in that the processor executable program is for performing the method according to any of claims 1-4 when being executed by a processor.

Description

Feed formula optimization method, system, device and storage medium Technical Field The invention relates to two large fields of feed formula and evolution calculation, in particular to a feed formula optimization method, a feed formula optimization system, a feed formula optimization device and a feed formula storage medium. Background With the continuous improvement of the living standard of people in China and the continuous promotion of the urban process, the daily rigid demands of animal products represented by meat, eggs and milk as vast residents are rapidly increased, and the aquaculture industry in China is rapidly developed, and gradually changes to standardized, large-scale and mechanized aquaculture industry. The feed industry is used as a supporting industry of the breeding industry, the market scale is continuously enlarged, and the feed has a large development space. The feed refers to the general term of artificial foods for all artificial feeding animals, and generally refers to artificial foods for feeding fish, livestock in the agriculture and animal husbandry. The feed is a key factor influencing the output and the raising cost of the raised animals, and relates to the economic benefit of farmers. On the one hand, the high-quality feed rich in nutrients required by animals can effectively improve the yield, such as the egg yield and the meat yield, so that the economic benefit is improved. On the other hand, the high-quality feed has higher cost and can reduce economic benefit. Therefore, how to control the feed cost on the premise of meeting the nutrient substances required by animals becomes a problem which the feed industry needs to solve preferentially. Aiming at the problems, an effective technical scheme is not available at present. Disclosure of Invention In order to solve at least one of the technical problems existing in the prior art to a certain extent, the invention aims to provide a feed formula optimization method, a feed formula optimization system, a feed formula optimization device and a feed formula storage medium. The technical scheme adopted by the invention is as follows: a feed formulation optimization method comprising the steps of: Determining raw materials of a feed formula and constraint conditions corresponding to the raw materials; constructing a feed formula optimization problem model, and determining an optimization target of the model; Adopting a distribution estimation algorithm containing a construction method and a construction information updating strategy to process a feed formula optimization problem model to obtain a final feed formula; The construction information updating strategy can update construction information by utilizing feed formula information in a parent population, and guide a construction feed formula generation method to generate a feed formula with higher nutritive value and lower raw material cost and meeting various constraint conditions. Further, the determining the raw materials of the feed formula and the constraint conditions corresponding to the raw materials comprises the following steps: Setting N raw materials which can be selected by a feed formula, wherein the raw materials are respectively divided into Z major categories with corresponding numbers of {1,2,3, & gt and Z }, the nutritional value of the ith raw material is v i, the cost is c i, and the major categories are g i E {1,2,3, & gt and Z }; The single raw material proportion constraint is that the use proportion w i of the ith raw material needs to meet constraint condition lb i≤wi≤ubi(0≤lbi<ubi which is less than or equal to 1), wherein lb i is the lower limit of the proportion of the raw material i, and ub i is the upper limit of the proportion of the raw material i; Raw materials major class example constraint, total raw materials usage ratio belonging to the same raw materials major class G iNeeds to satisfy constraint conditionsWherein lg i is the lower limit of the proportion of the raw material class G i, ug i is the upper limit of the proportion of the raw material class G i; the number of the used raw materials is restricted, wherein the number of the used raw materials in the feed formula is K, and K is more than or equal to 1 and less than or equal to N. Further, the constructing the feed formula optimization problem model and determining the optimization target of the model comprises the following steps: the feed formula optimization problem is that K raw materials are selected and the use proportion of each raw material is determined to form a feed formula; The optimization targets are that the nutrition value V of the formula is maximized, the raw material cost C of the formula is minimized, and meanwhile, the single raw material proportion constraint condition BC, the raw material big class constraint condition GC and the raw material seed number constraint K are required to be met; The vector formed by the K raw material sets is (a 1,a2,a3,...,aK)a