A preparation and metering method and system for extracting and preparing macadamia nut oil

Abstract

The present invention relates to the technical field of macadamia nut preparation and processing, and in particular to a method and system for dispensing and metering macadamia nut oil nutrient extraction and preparation. The ideal dispensing ratio of each nutrient component that meets the state characteristics is determined by the deployment of a nutrient ratio interactive network to output; the process state variables are globally optimized through each ideal dispensing ratio, and the feedback gain matrix of the current process state control is analyzed based on the expected process state variables to calibrate the current process; the actual ratio of each nutrient component in the macadamia nut oil sample and the corresponding concentration peak area are obtained based on accurate dispensing process control parameters, and the actual ratio and the corresponding concentration peak area are hashed and cumulatively measured and analyzed with the ideal dispensing ratio as the bubble volume benchmark to obtain the metering result. The present invention can accurately dispense the nutrient ratio and reliably measure the components for the nutrient extraction and preparation of macadamia nut oil, ensuring the nutritional balance and practical value of macadamia nut oil products.

Description

Technical Field

[0001] The present invention relates to the technical field of macadamia nut preparation and processing, and in particular to a mixing and metering method and system for the nutritional extraction and preparation of macadamia nut oil. Background Art

[0002] Macadamia oil is a vegetable oil extracted from the fruit of macadamia nuts. It is widely used for its rich nutrients and multiple health benefits. Its unique nutrients and high-quality fatty acid structure make it a healthy edible oil and an ingredient in beauty and skin care products. Therefore, it is crucial to ensure the rationality of the nutrient blending and concentration measurement during the extraction and preparation of macadamia oil. However, the traditional methods of nutrient extraction, blending and concentration measurement of macadamia oil usually require manual calculation and determination with the help of specific tools, which requires the collaborative execution of multiple people, is time-consuming and labor-intensive, and increases cost output; and the final blending and measurement results have large manual measurement errors, which is not conducive to the high-quality extraction and preparation of macadamia oil, and greatly reduces the nutritional value and quality of macadamia oil. At the same time, it is difficult for traditional methods to calibrate and optimize the control of the extraction and preparation process according to the ideal blending of nutrients, so that the final extracted and prepared macadamia oil cannot reach the ideal blending ratio and concentration. Summary of the invention

[0003] The present invention overcomes the shortcomings of the prior art and provides a mixing and metering method and system for extracting and preparing macadamia nut oil nutrients.

[0004] To achieve the above object, the technical solution adopted by the present invention is:

[0005] The first aspect of the present invention provides a method for preparing and dispensing macadamia nut oil by nutrient extraction, comprising the following steps:

[0006] S102: Obtaining the application scenario and nutritional preparation requirements of macadamia nut oil, analyzing and calculating several groups of historical preparation interaction indicator variables of various nutritional components through the application scenario and nutritional preparation requirements, so as to determine the interaction mechanism index between the various nutritional components, constraining the nutritional ratio value according to the interaction mechanism index, and obtaining the nutritional ratio value range of various nutritional components;

[0007] S104: obtaining one or more target state characteristics of macadamia nut oil, adaptively learning and constructing a nutrient ratio interaction network in which each nutrient ratio value in the nutrient ratio value range meets the state characteristics through a mean square error regression equation, and outputting an ideal blending ratio of each nutrient component meeting the state characteristics through the deployment of the nutrient ratio interaction network;

[0008] S106: Obtaining the process state variables of the macadamia nut oil nutrient extraction preparation process, globally optimizing the process state variables through each of the ideal blending ratios to obtain the expected process state variables, analyzing the feedback gain matrix of the current process state control based on the expected process state variables, calibrating the current process using the feedback gain matrix, and obtaining accurate blending process control parameters;

[0009] S108: Based on the accurate blending process control parameters, the actual ratio of each nutrient component in the macadamia oil sample and the corresponding concentration peak area are obtained, and the actual ratio and the corresponding concentration peak area are hashed and cumulatively measured and analyzed with the ideal blending ratio as the bubble volume benchmark to obtain the measurement result.

[0010] More specifically, the step S102 includes the following steps:

[0011] Obtain the application scenarios and nutritional preparation requirements of macadamia oil, and obtain the various nutrients necessary for macadamia oil in the application scenarios based on the big data network;

[0012] Based on the nutritional preparation requirements, the optimal nutritional supply base corresponding to the application scenario of various nutritional components in macadamia nut oil and the logic of the interaction mechanism between each nutritional component are retrieved in the big data network;

[0013] Obtain the nutritional extraction and preparation logs of macadamia nuts, and extract several groups of historical preparation interaction indicator variables on the optimal nutritional supply base between each nutritional component of macadamia nut oil through the nutritional extraction and preparation logs;

[0014] Determine the dependency relationship between each nutrient component according to the interaction mechanism logic, construct a variable dependency directed graph based on the dependency relationship, import several groups of historically prepared interaction indicator variables into the variable dependency directed graph for fitting definition, and generate a Belief network model;

[0015] The joint probability distribution between each nutrient component is calculated according to the Belief network model analysis, and the marginalized probability summation operation is performed on the joint probability distribution to obtain the marginal probability distribution. The interaction coefficient of each group of historical preparation interaction indicator variables is approximated based on the joint probability distribution and the marginal probability distribution, and the interaction mechanism index between various nutrients is determined according to the interaction coefficient;

[0016] The expected interaction mechanism index is preset according to the nutritional preparation requirements. If the interaction mechanism index exceeds the expected interaction mechanism index, the index difference between the interaction mechanism index and the expected interaction mechanism index is calculated to obtain multiple interaction mechanism index deviation values, and a discrete inclusion constraint matrix is ​​established through the multiple interaction mechanism index deviation values.

[0017] Through the big data network, various nutrients are obtained to achieve multiple nutritional ratios corresponding to the optimal nutritional supply base, and a backtracking algorithm is introduced. In the backtracking algorithm, each nutritional ratio of various nutrients is discretely combined until it approaches the expected interaction mechanism index, and multiple discrete combination sub-matrices are output;

[0018] If the discrete combination submatrix exists in the discrete inclusion constraint matrix, the combined nutritional ratios of various nutrients in the discrete combination submatrix are retained, and finally the nutritional ratio value range of various nutrients is formed.

[0019] More specifically, the step S104 includes the following steps:

[0020] Constructing a sparse feature dictionary, performing sparse feature encoding on the nutritional ratio value range of each nutritional component through the sparse feature dictionary, and generating a nutritional ratio sparse feature matrix;

[0021] Obtain one or more target state characteristics of macadamia oil through nutritional preparation requirements, and set linear weights, global bias terms and latent vectors of the target state characteristics according to the nutritional ratio sparse feature matrix;

[0022] Based on the linear weight, global bias term and latent vector, a state characteristic mean square error regression equation is defined and established, and the first-order moment estimation parameters and the second-order moment estimation parameters of the state characteristic mean square error regression equation are initialized based on the nutrient ratio coefficient characteristic matrix;

[0023] An adaptive gradient algorithm is introduced to calculate an adaptive gradient value between a sparse characteristic matrix of a nutrient ratio and a mean square error of a state characteristic, and a first-order moment estimation parameter and a second-order moment estimation parameter are updated based on the adaptive gradient value to obtain an updated first-order moment estimation parameter and an updated second-order moment estimation parameter;

[0024] Correcting the deviation between the updated first-order moment estimation parameter and the first-order moment estimation parameter to obtain the first-order moment estimation deviation correction parameter, correcting the deviation between the updated second-order moment estimation parameter and the second-order moment estimation parameter to obtain the second-order moment estimation deviation correction parameter;

[0025] Based on the first-order moment estimation deviation correction parameter and the second-order moment estimation deviation correction parameter, the learning rate of each nutrient ratio sparse feature in the nutrient ratio sparse feature matrix for the state characteristic mean square error is solved to obtain an adaptive gradient learning rate for each nutrient ratio sparse feature;

[0026] Designing an interactive term update rule of the sparse features of the nutrient ratio through an adaptive gradient learning rate, updating and iterating the sparse feature matrix of the nutrient ratio based on the interactive term update rule, and generating a nutrient ratio interactive network that conforms to the state characteristics;

[0027] The nutrient ratio interaction network is deployed in the nutrient ratio value domain, and finally outputs an ideal mixing ratio for each nutrient component that meets the state characteristics.

[0028] More specifically, the step S106 includes the following steps:

[0029] According to the ideal blending ratio of each nutrient component in accordance with the state characteristics, the macadamia oil is blended and extracted in the macadamia oil nutrient extraction and preparation process, and the process state variables and process control input parameters are obtained at this time;

[0030] Obtaining the actual blending ratio of each nutrient component in the extracted macadamia nut oil, presetting an allowable blending ratio drift threshold based on nutrient preparation requirements and an optimal nutrient supply base, calculating a blending ratio error between the actual blending ratio value and a corresponding ideal blending ratio, and obtaining a blending ratio deviation value;

[0031] If the blending ratio deviation value is greater than the allowable blending ratio drift threshold, a genetic algorithm is introduced, and based on the fitness of the offspring individuals of the process state variables of the blending ratio deviation value in the genetic algorithm, the parent generation is selected through the offspring individual fitness and a crossover mutation process is performed to obtain a new offspring individual;

[0032] Repeat the above steps to continuously update and iterate the new offspring individuals to replace all the old offspring individuals, forming a new population of process state variables, and extract the process state variable with the maximum offspring individual fitness from the new population of process state variables, which is defined as the expected process state variable;

[0033] A state control model of the macadamia oil nutrient extraction and preparation process is constructed based on the process state variables and process control parameters, a hash state value between the expected process state variables and the process state variables is calculated, and a cost function of the control input for the process state difference is determined based on the hash state value;

[0034] Constructing an expected state control algebraic differential equation of a state control model according to expected process state variables, solving the expected state control algebraic differential equation through a cost function to obtain a feedback gain matrix, and globally calibrating the state control model based on the feedback gain matrix to obtain a calibrated state control model;

[0035] The calibrated state control model is applied to the macadamia oil nutrient extraction and preparation process to obtain accurate formulation process control parameters of the macadamia oil nutrient extraction and preparation process.

[0036] More specifically, the step S108 includes the following steps:

[0037] Obtaining the variety information of the current macadamia nuts, and controlling the macadamia nut oil nutrient extraction and preparation process to extract and prepare the current macadamia nuts according to the accurate blending process control parameters to obtain a macadamia nut oil sample;

[0038] Performing nutritional component testing on the macadamia nut oil sample by liquid chromatography to obtain the actual ratio of each nutritional component in the macadamia nut oil sample and the corresponding concentration peak area;

[0039] Construct a bubble chart, preset a bubble volume limit threshold based on the ideal mixing ratio, define and plan the bubble volume of each nutrient in the bubble chart until the corresponding bubble volume limit threshold is reached, and generate an ideal mixing bubble volume range corresponding to each nutrient;

[0040] Based on the preset initial bubble color cumulative hash value of the ideal blended bubble volume range, the actual ratio of each nutrient component and the corresponding concentration peak area are bundled and converted into a bubble color coding character, and a hash algorithm is introduced to calculate the bubble color hash value of each bubble color coding character;

[0041] By accumulating and splicing each bubble color coding character within the ideal blending bubble volume range of the corresponding nutrient component, during the accumulative splicing process, the initial bubble color cumulative hash value is continuously updated based on the bubble color hash value of each bubble color coding character to generate a hash cumulative table for each nutrient component;

[0042] Traverse the actual ratio and the corresponding concentration peak area in the metering hash accumulation table to output the final bubble color cumulative hash value of the ideal mixing bubble volume range, and determine the final concentration metering value of each nutrient when it reaches the corresponding ideal mixing ratio according to the final bubble color hash accumulation value;

[0043] Based on the rationality of the concentration effect of the final concentration measurement value when each nutrient reaches the corresponding ideal blending ratio, the quality analysis and determination of the macadamia oil product are carried out to obtain the measurement results.

[0044] More specifically, the quality analysis and determination of the macadamia oil product based on the rationality of the concentration effect of the final concentration measurement value when each nutrient component reaches the corresponding ideal blending ratio in step S108 to obtain the measurement result specifically includes the following steps:

[0045] According to the nutritional preparation requirements, the expected effect coefficient of macadamia nut oil in the application scenario is obtained, and based on the expected effect coefficient, the appropriate concentration range of each nutrient component is retrieved in the big data network;

[0046] Constructing a multidimensional binning domain, dividing the multidimensional binning domain into sub-binning domains representing each nutrient component based on multiple nutrients, and obtaining a plurality of gradient critical points of each of the suitable concentration ranges;

[0047] Taking the root node as the starting node, each gradient critical point is used to split the appropriate concentration value between two adjacent gradient critical points in the corresponding appropriate concentration range, and a number of branch nodes are generated. Repeat the above steps to continuously split the appropriate concentration value of each branch node, and finally form a binning decision tree;

[0048] Obtain all leaf node values ​​of the binning decision tree, determine the binning points of each suitable concentration range according to all leaf node values, divide each sub-binning area into N bins based on the binning points, and calculate the upper and lower suitable concentration values ​​at the binning boundaries in each bin;

[0049] If the final concentration measurement value is less than the upper suitable concentration value and greater than the lower suitable concentration value, the final concentration measurement value is embedded in the box compartment corresponding to the upper suitable concentration value and the lower suitable concentration value to obtain a fixed-point box compartment;

[0050] Obtain the fixed-point bins and cells whose final concentration measurement values ​​of all nutrients are in the corresponding sub-binning areas, strip off the multidimensional binning areas formed by all the fixed-point bins and cells, define them as the first multidimensional binning areas, and plan out the expected multidimensional binning areas where all nutrients reach the expected effect coefficients, define them as the second multidimensional binning areas;

[0051] The overlapping area between the first multidimensional binning area and the second multidimensional binning area is eliminated to obtain a non-overlapping multidimensional binning area. If the volume value of the non-overlapping multidimensional binning area is less than the preset volume value, the macadamia oil is calibrated as a qualified product. If the volume value of the non-overlapping multidimensional binning area is greater than the preset volume value, the macadamia oil is calibrated as an unqualified product to obtain a measurement result.

[0052] A second aspect of the present invention provides a mixing and metering system for the preparation of nutritional extraction of macadamia nut oil, the mixing and metering system for the preparation of nutritional extraction of macadamia nut oil comprising a memory and a processor, the memory storing a mixing and metering method program for the preparation of nutritional extraction of macadamia nut oil, when the mixing and metering method program for the preparation of nutritional extraction of macadamia nut oil is executed by the processor, any one of the steps of the mixing and metering method for the preparation of nutritional extraction of macadamia nut oil is implemented.

[0053] The present invention solves the technical defects existing in the background technology, and the beneficial technical effects of the present invention are:

[0054] The interaction mechanism index between various nutrients is determined through the application scenario of macadamia oil and the analysis of nutritional preparation requirements, and the nutrient ratio is constrained according to the interaction mechanism index to obtain the nutrient ratio value range of various nutrients; the nutrient ratio interaction network in which each nutrient ratio value in the nutrient ratio value range meets the state characteristics is adaptively learned through the mean square error regression equation, and the ideal blending ratio of each nutrient component that meets the state characteristics is determined by output through the deployment of the nutrient ratio interaction network; the process state variables are globally optimized through each of the ideal blending ratios to obtain the expected process state variables, the feedback gain matrix of the current process state control is analyzed based on the expected process state variables, and the current process is calibrated using the feedback gain matrix to obtain accurate blending process control parameters; the actual ratio of each nutrient component in the macadamia oil sample and the corresponding concentration peak area are obtained based on the accurate blending process control parameters, and the actual ratio and the corresponding concentration peak area are hashed and cumulatively measured and analyzed with the ideal blending ratio as the bubble volume benchmark to obtain the measurement results. The present invention can accurately and high-quality prepare the nutritional components in proportion and reliably measure the components in the nutritional extraction and preparation of macadamia nut oil, thereby ensuring the nutritional balance of the macadamia nut oil and improving the practical value of the macadamia nut oil product. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, drawings of other embodiments can be obtained based on these drawings without paying creative work.

[0056] Figure 1 A first method flow chart of a method for preparing and dispensing macadamia nut oil nutrient extraction is shown;

[0057] Figure 2 A second method flow chart of a method for preparing nutritional extraction of macadamia nut oil is shown;

[0058] Figure 3 A third method flow chart of a method for preparing and dispensing macadamia nut oil nutrient extraction is shown;

[0059] Figure 4 The system framework diagram of a mixing and metering system for preparing macadamia nut oil nutritional extraction is shown. DETAILED DESCRIPTION

[0060] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0061] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.

[0062] The first aspect of the present invention provides a method for preparing and dispensing macadamia nut oil by nutrient extraction, such as Figure 1 As shown, the following steps are included:

[0063] S102: Obtaining the application scenario and nutritional preparation requirements of macadamia nut oil, analyzing and calculating several groups of historical preparation interaction indicator variables of various nutritional components through the application scenario and nutritional preparation requirements, so as to determine the interaction mechanism index between the various nutritional components, constraining the nutritional ratio value according to the interaction mechanism index, and obtaining the nutritional ratio value range of various nutritional components;

[0064] S104: obtaining one or more target state characteristics of macadamia nut oil, adaptively learning and constructing a nutrient ratio interaction network in which each nutrient ratio value in the nutrient ratio value range meets the state characteristics through a mean square error regression equation, and outputting an ideal blending ratio of each nutrient component meeting the state characteristics through the deployment of the nutrient ratio interaction network;

[0065] S106: Obtaining the process state variables of the macadamia nut oil nutrient extraction preparation process, globally optimizing the process state variables through each of the ideal blending ratios to obtain the expected process state variables, analyzing the feedback gain matrix of the current process state control based on the expected process state variables, calibrating the current process using the feedback gain matrix, and obtaining accurate blending process control parameters;

[0066] S108: Based on the accurate blending process control parameters, the actual ratio of each nutrient component in the macadamia oil sample and the corresponding concentration peak area are obtained, and the actual ratio and the corresponding concentration peak area are hashed and cumulatively measured and analyzed with the ideal blending ratio as the bubble volume benchmark to obtain the measurement result.

[0067] More specifically, the step S102 includes the following steps:

[0068] Obtain the application scenarios and nutritional preparation requirements of macadamia oil, and obtain the various nutrients necessary for macadamia oil in the application scenarios based on the big data network;

[0069] Based on the nutritional preparation requirements, the optimal nutritional supply base corresponding to the application scenario of various nutritional components in macadamia nut oil and the logic of the interaction mechanism between each nutritional component are retrieved in the big data network;

[0070] Obtain the nutritional extraction and preparation logs of macadamia nuts, and extract several groups of historical preparation interaction indicator variables on the optimal nutritional supply base between each nutritional component of macadamia nut oil through the nutritional extraction and preparation logs;

[0071] Determine the dependency relationship between each nutrient component according to the interaction mechanism logic, construct a variable dependency directed graph based on the dependency relationship, import several groups of historically prepared interaction indicator variables into the variable dependency directed graph for fitting definition, and generate a Belief network model;

[0072] The joint probability distribution between each nutrient component is calculated according to the Belief network model analysis, and the marginalized probability summation operation is performed on the joint probability distribution to obtain the marginal probability distribution. The interaction coefficient of each group of historical preparation interaction indicator variables is approximated based on the joint probability distribution and the marginal probability distribution, and the interaction mechanism index between various nutrients is determined according to the interaction coefficient;

[0073] The expected interaction mechanism index is preset according to the nutritional preparation requirements. If the interaction mechanism index exceeds the expected interaction mechanism index, the index difference between the interaction mechanism index and the expected interaction mechanism index is calculated to obtain multiple interaction mechanism index deviation values, and a discrete inclusion constraint matrix is ​​established through the multiple interaction mechanism index deviation values.

[0074] Through the big data network, various nutrients are obtained to achieve multiple nutritional ratios corresponding to the optimal nutritional supply base, and a backtracking algorithm is introduced. In the backtracking algorithm, each nutritional ratio of various nutrients is discretely combined until it approaches the expected interaction mechanism index, and multiple discrete combination sub-matrices are output;

[0075] If the discrete combination submatrix exists in the discrete inclusion constraint matrix, the combined nutritional ratios of various nutrients in the discrete combination submatrix are retained, and finally the nutritional ratio value range of various nutrients is formed.

[0076] It should be noted that the nutrients include monounsaturated fatty acids, saturated fatty acids, antioxidants, minerals and phytosterols. Macadamia oil is rich in a variety of nutrients, for example, macadamia oil is rich in monounsaturated fatty acids, especially oleic acid and palmitoleic acid, which makes it have excellent moisturizing and skin care effects. There are certain nutritional values ​​in these ingredients, which can provide the best nutritional supply according to the application scenario and nutritional needs. If the allocation ratio between the nutrients is not considered during the preparation of macadamia oil nutritional extraction, the overall nutritional value of macadamia oil may be unbalanced, which is not conducive to the application quality and effect of health and safety. Therefore, it is necessary to clarify the range of nutritional ratios that can be selected when each nutrient reaches the required nutritional needs. This method can accurately analyze, plan and include the nutritional ratio value range of each nutrient in macadamia oil when it is applied in the specified application scenario, so as to provide a reliable value basis for the precise allocation of various nutrients in subsequent macadamia oil, and ensure that the nutritional extraction and preparation of macadamia oil meets the nutritional application needs while improving the nutritional balance of macadamia oil. Among them, although macadamia oil is rich in a variety of high-value nutrients, there are interaction mechanisms between certain specific nutrients. For example, antioxidants and phytochemicals can protect the fatty acids in the oil from oxidative damage and extend its shelf life. The interaction effect between nutrients will affect the overall nutritional value of macadamia oil to a certain extent. Therefore, when limiting the ratio range, it is necessary to fully consider its interaction mechanism to ensure that the interaction effect between nutrients will not destroy or deviate from the nutritional value requirements of macadamia oil, and further improve the stability of macadamia oil nutritional formulation. Among them, the Belief network model, whose Chinese name is the belief network model, is used to represent a set of variables and their conditional dependencies. It can clearly and intuitively express the interaction relationship between nutrient variables, so as to accurately determine the interaction influence index between various nutrients, and provide an effective reference value for the analysis basis of the ratio range limitation.

[0077] It should be noted that if the interaction mechanism index exceeds the expected interaction mechanism index, it means that the interaction effect between the nutrients does not meet the expected requirements. Therefore, it is necessary to eliminate the deviation between the two to further ensure that the value planning of the nutrient ratio can tend to the expected interaction mechanism index. Therefore, the establishment of the discrete inclusion constraint matrix needs to rely on the deviation value of the interaction mechanism index. This method can accurately define and plan the reference range of nutrient ratio values ​​that meet the requirements for the accurate allocation of various nutrients in macadamia oil, greatly improve the balance of nutrient allocation of macadamia oil, and achieve the high nutritional value of macadamia oil with high reliability.

[0078] More specifically, the step S104 includes the following steps:

[0079] Constructing a sparse feature dictionary, performing sparse feature encoding on the nutritional ratio value range of each nutritional component through the sparse feature dictionary, and generating a nutritional ratio sparse feature matrix;

[0080] Obtain one or more target state characteristics of macadamia oil through nutritional preparation requirements, and set linear weights, global bias terms and latent vectors of the target state characteristics according to the nutritional ratio sparse feature matrix;

[0081] Based on the linear weight, global bias term and latent vector, a state characteristic mean square error regression equation is defined and established, and the first-order moment estimation parameters and the second-order moment estimation parameters of the state characteristic mean square error regression equation are initialized based on the nutrient ratio coefficient characteristic matrix;

[0082] An adaptive gradient algorithm is introduced to calculate an adaptive gradient value between a sparse characteristic matrix of a nutrient ratio and a mean square error of a state characteristic, and a first-order moment estimation parameter and a second-order moment estimation parameter are updated based on the adaptive gradient value to obtain an updated first-order moment estimation parameter and an updated second-order moment estimation parameter;

[0083] Correcting the deviation between the updated first-order moment estimation parameter and the first-order moment estimation parameter to obtain the first-order moment estimation deviation correction parameter, correcting the deviation between the updated second-order moment estimation parameter and the second-order moment estimation parameter to obtain the second-order moment estimation deviation correction parameter;

[0084] Based on the first-order moment estimation deviation correction parameter and the second-order moment estimation deviation correction parameter, the learning rate of each nutrient ratio sparse feature in the nutrient ratio sparse feature matrix for the state characteristic mean square error is solved to obtain an adaptive gradient learning rate for each nutrient ratio sparse feature;

[0085] Designing an interactive term update rule of the sparse features of the nutrient ratio through an adaptive gradient learning rate, updating and iterating the sparse feature matrix of the nutrient ratio based on the interactive term update rule, and generating a nutrient ratio interactive network that conforms to the state characteristics;

[0086] The nutrient ratio interaction network is deployed in the nutrient ratio value domain, and finally outputs an ideal mixing ratio for each nutrient component that meets the state characteristics.

[0087] It should be noted that after determining the nutritional ratio range for each nutrient in macadamia oil, a reasonable blending ratio can be further selected based on its range, but nutrients with different blending ratios will affect the different state characteristics of macadamia oil. For example, oil with a higher oleic acid content may be smoother and lighter, while oil with a higher palmitoleic acid content may be denser, which affects the sensory state characteristics; or a higher proportion of oleic acid has the effect of lowering bad cholesterol levels, which is helpful for cardiovascular health, and affects the health state characteristics. Therefore, in order to ensure that macadamia oil meets the specified state characteristic requirements, the specific blending ratio of each nutrient must be screened according to different state characteristics. Therefore, this method constructs a state characteristic mean square error regression equation through the target state characteristic. This state characteristic mean square error regression equation can capture the interactive behavior between each nutrient ratio feature based on the target state characteristic regression, so that the final state formed when each nutrient is mixed and formulated is within the target state characteristic requirements, thereby improving the stability of the state characteristics of the macadamia oil nutrient blend. The adaptive gradient algorithm can dynamically adjust the learning rate for each nutrient ratio in the value range, which means that different step sizes can be selected for different parameters according to the changes in the target state characteristic gradient, which can further improve the accuracy of learning interactive adaptation of nutrient ratios based on target state characteristics and ensure the reliability of determining the ideal adjustment ratio results.

[0088] It should be noted that this method estimates the first-order moment and second-order moment of the adaptive gradient. The first-order moment is the momentum, and the second-order moment is the sum of the squares of the gradient. On the one hand, it takes into account the direction of the gradient, and on the other hand, it takes into account the fluctuation of the gradient, which helps to update the parameters more stably and reduce the oscillation in the updating process. At the same time, the estimation of the first-order and second-order moments greatly improves the computational efficiency compared with the traditional methods, thereby improving the update screening accuracy of each nutrient ratio in the nutrient ratio value range that is within the premise of the target state characteristic requirements, ensuring that the final ideal blending ratio can meet the target state characteristic requirements, avoiding the instability of the state characteristics of macadamia oil, and ensuring the nutritional value and nutritional balance, thereby improving the health factor of macadamia oil in actual application scenarios.

[0089] More specifically, the step S106 is as follows: Figure 2 As shown, the specific steps include:

[0090] S202: blending and extracting macadamia nut oil in a macadamia nut oil nutrient extraction and preparation process according to the ideal blending ratio of each nutrient component that meets the state characteristics, and obtaining process state variables and process control input parameters;

[0091] S204: obtaining the actual blending ratio of each nutrient component in the extracted macadamia nut oil, presetting an allowable blending ratio drift threshold based on nutrient preparation requirements and an optimal nutrient supply base, calculating a blending ratio error between the actual blending ratio and the corresponding ideal blending ratio, and obtaining a blending ratio deviation value;

[0092] S206: If the blending ratio deviation value is greater than the allowable blending ratio drift threshold, a genetic algorithm is introduced, and based on the fitness of the offspring individuals of the process state variables of the blending ratio deviation value in the genetic algorithm, the parent generation is selected through the offspring individual fitness and a crossover mutation process is performed to obtain a new offspring individual;

[0093] S208: Repeat the above steps to continuously update and iterate the new offspring individuals to replace all the old offspring individuals, forming a new process state variable population, and extracting the process state variable with the maximum offspring individual fitness from the new process state variable population, which is defined as the expected process state variable;

[0094] S210: constructing a state control model of the macadamia oil nutrient extraction and preparation process based on the process state variables and process control parameters, calculating a hash state value between the expected process state variables and the process state variables, and determining a cost function of the control input for the process state difference according to the hash state value;

[0095] S212: constructing an expected state control algebraic differential equation of the state control model according to the expected process state variable analysis, solving the expected state control algebraic differential equation through a cost function to obtain a feedback gain matrix, and globally calibrating the state control model based on the feedback gain matrix to obtain a calibrated state control model;

[0096] S214: Applying the calibrated state control model to the macadamia oil nutrient extraction and preparation process to obtain accurate formulation process control parameters of the macadamia oil nutrient extraction and preparation process.

[0097] It should be noted that, since the ratio of each nutrient in macadamia oil is in natural form, that is, the ratio of each component is relatively fixed, it can only be changed by certain specific means, and the most direct and feasible way is to adjust the nutritional extraction and preparation process of macadamia oil. However, the traditional process adjustment method usually requires manual adjustment of each process control parameter one by one with the help of specific tools, which often produces a large manual adjustment error, so that the actual blending ratio of various nutrients in macadamia oil does not meet the ideal blending ratio, or the finished product finally obtained when extracting and preparing macadamia oil based on the ideal blending ratio does not meet the nutritional preparation requirements, resulting in nutritional imbalance of macadamia oil, greatly destroying the overall quality and nutritional value of macadamia oil, and reducing the nutritional extraction and preparation efficiency of macadamia oil. Therefore, this method constructs a state control model through the process state variables and process control input parameters in the process of extracting macadamia oil by the preparation process, and this state control model is used to describe the actual process state of the expression process to prepare macadamia oil. Then, the cost function of the control input of the state control model for the process state difference is analyzed to feedback the degree of gap between the current process and the expected process state when each nutrient reaches the ideal blending ratio. According to this degree of gap, the corresponding feedback gain matrix is ​​designed to compensate for the state gap in which the process state output by the current preparation process cannot extract the ideal blending ratio effect, thereby optimizing the extraction and preparation performance of the macadamia oil nutrient extraction preparation process for the ideal blending ratio, thereby ensuring that the blending effect extracted and prepared by each nutrient can meet the ideal blending effect, effectively improving the accuracy of the preparation process for the nutrient blending, and effectively ensuring the overall nutritional balance and value of macadamia oil.

[0098] It should be noted that if the blending ratio deviation value is greater than the allowable blending ratio drift threshold, it means that the actual blending ratio of macadamia oil prepared by the current process is significantly different from the ideal blending ratio, and the process state needs to be calibrated and optimized; and this method uses a genetic algorithm to globally iteratively optimize the process state of the preparation process based on the blending ratio deviation value, to ensure that the actual process state can ultimately reach the expected process state level for preparing and extracting the ideal blending ratio, thereby improving the nutritional extraction and preparation quality of macadamia oil; and it can replace the tedious steps of traditional manual adjustment and optimization, saving time and effort, avoiding process preparation errors caused by manual errors, and has high reliability.

[0099] More specifically, the step S108 is as follows: Figure 3 As shown, the specific steps include:

[0100] S302: Obtaining the variety information of the current macadamia nut, and controlling the macadamia nut oil nutrient extraction and preparation process to extract and prepare the current macadamia nut according to the accurate blending process control parameters to obtain a macadamia nut oil sample;

[0101] S304: Detecting the nutritional components of the macadamia oil sample by liquid chromatography to obtain the actual ratio of each nutritional component in the macadamia oil sample and the corresponding concentration peak area;

[0102] S306: constructing a bubble chart, presetting a bubble volume limit threshold based on the ideal mixing ratio, defining and planning the bubble volume of each nutrient component in the bubble chart until the corresponding bubble volume limit threshold is reached, and generating an ideal mixing bubble volume range corresponding to each nutrient component;

[0103] S308: Preset the initial bubble color cumulative hash value based on the ideal blended bubble volume range, bundle the actual ratio of each nutrient component with the corresponding concentration peak area into a bubble color code character, and introduce a hash algorithm to calculate the bubble color hash value of each bubble color code character;

[0104] S310: Generate a hash accumulation table for each nutrient component by accumulating and splicing each bubble color code character within the ideal blending bubble volume range of the corresponding nutrient component, and continuously updating the initial bubble color cumulative hash value based on the bubble color hash value of each bubble color code character during the accumulative splicing process;

[0105] S312: traverse the actual ratio and the corresponding concentration peak area in the metering hash accumulation table to output the final bubble color cumulative hash value of the ideal mixing bubble volume range, and determine the final concentration metering value of each nutrient component when it reaches the corresponding ideal mixing ratio according to the final bubble color hash cumulative value;

[0106] S314: Based on the rationality of the concentration effect of the final concentration measurement value when each nutrient component reaches the corresponding ideal blending ratio, the quality analysis and determination of the macadamia oil product is performed to obtain the measurement result.

[0107] It should be noted that the ideal blending ratio of various nutrients is extracted and prepared by accurately blending process control parameters of the macadamia nut oil nutrient extraction and preparation process, which can only ensure the nutritional value balance effect and state stability of various nutrients, and whether the nutritional health of a single nutrient is reasonable or not cannot be intuitively judged. For example, although the oleic acid ratio reaches a certain nutrient supply coefficient, the concentration of the final extraction and preparation after reaching the ideal blending ratio may be low. Low oleic acid will reduce the moisturizing effect and make it difficult to better lock the moisture in the skin. Based on the above situation, it is very important to measure the concentration content of each nutrient in macadamia nut oil. Therefore, this method first detects the actual ratio of each nutrient in the macadamia nut oil sample and the corresponding concentration peak area by liquid chromatography. The concentration peak area is an important parameter used in chromatographic analysis, which represents the integral value of the signal intensity and time detected by the detector during the chromatographic analysis, and can quantitatively analyze the content of each component in the sample from a side angle. Then, the ideal blending bubble volume range corresponding to each nutrient is planned by constructing a bubble chart. The range is used to visualize the proportion of nutrients in macadamia oil when the nutrient reaches the ideal blending ratio. Then, the bubble color coding characters formed by bundling the actual ratio with the corresponding concentration peak area are cumulatively spliced ​​within the ideal blending bubble volume range until the entire ideal blending bubble volume range is filled. The initial bubble color cumulative hash value updated by the bubble color hash value cumulative splicing will form a hash cumulative table, and the bubble color finally presented in the hash cumulative table within the ideal blending bubble volume range is the final metering concentration value of the nutrient, that is, the final bubble color hash cumulative value. This method can efficiently and accurately accumulate and measure the concentration of ideally blended nutrients, can improve the reliability of the metering results on the basis of minimizing the traditional manual metering interference steps, and ensure the rationality of the nutrient concentration of macadamia oil; at the same time, it can also analyze the quality of the extracted and prepared macadamia oil products according to the concentration metering results of the nutrient concentration, which is conducive to the optimization and improvement of the subsequent macadamia oil extraction and preparation quality, so as to reduce the defective rate of macadamia oil and improve the extraction and preparation quality and efficiency.

[0108] More specifically, the quality analysis and determination of the macadamia oil product based on the rationality of the concentration effect of the final concentration measurement value when each nutrient component reaches the corresponding ideal blending ratio in step S108 to obtain the measurement result specifically includes the following steps:

[0109] According to the nutritional preparation requirements, the expected effect coefficient of macadamia nut oil in the application scenario is obtained, and based on the expected effect coefficient, the appropriate concentration range of each nutrient component is retrieved in the big data network;

[0110] Constructing a multidimensional binning domain, dividing the multidimensional binning domain into sub-binning domains representing each nutrient component based on multiple nutrients, and obtaining a plurality of gradient critical points of each of the suitable concentration ranges;

[0111] Taking the root node as the starting node, each gradient critical point is used to split the appropriate concentration value between two adjacent gradient critical points in the corresponding appropriate concentration range, and a number of branch nodes are generated. Repeat the above steps to continuously split the appropriate concentration value of each branch node, and finally form a binning decision tree;

[0112] Obtain all leaf node values ​​of the binning decision tree, determine the binning points of each suitable concentration range according to all leaf node values, divide each sub-binning area into N bins based on the binning points, and calculate the upper and lower suitable concentration values ​​at the binning boundaries in each bin;

[0113] If the final concentration measurement value is less than the upper suitable concentration value and greater than the lower suitable concentration value, the final concentration measurement value is embedded in the box compartment corresponding to the upper suitable concentration value and the lower suitable concentration value to obtain a fixed-point box compartment;

[0114] Obtain the fixed-point bins and cells whose final concentration measurement values ​​of all nutrients are in the corresponding sub-binning areas, strip off the multidimensional binning areas formed by all the fixed-point bins and cells, define them as the first multidimensional binning areas, and plan out the expected multidimensional binning areas where all nutrients reach the expected effect coefficients, define them as the second multidimensional binning areas;

[0115] The overlapping area between the first multidimensional binning area and the second multidimensional binning area is eliminated to obtain a non-overlapping multidimensional binning area. If the volume value of the non-overlapping multidimensional binning area is less than the preset volume value, the macadamia oil is calibrated as a qualified product. If the volume value of the non-overlapping multidimensional binning area is greater than the preset volume value, the macadamia oil is calibrated as an unqualified product to obtain a measurement result.

[0116] It should be noted that if the concentration of each nutrient in macadamia oil cannot be maintained at a more suitable level, it may cause macadamia oil to be difficult to achieve the expected effect coefficient when it is used in the required application scenario. Therefore, the quality of macadamia oil can be further determined by analyzing whether the concentration of each nutrient in the extracted and prepared macadamia oil is at an appropriate level. This method constructs a multidimensional binning field, which is used to visualize whether the final concentration measurement value when all nutrients reach the corresponding ideal blending ratio meets the expected concentration effect level. First, by dividing the multidimensional binning field into multiple sub-binning fields according to the number of nutrients, the concentration measurement level of each nutrient can be displayed globally without the need to construct them one by one, thereby reducing the number of calculation steps; then, binning is performed on each sub-binning field. This method uses a decision tree binning method to automatically determine the best split point without specifying the number or range of bins in advance, thereby improving the flexibility and accuracy of binning. Then, according to the upper suitable concentration value and the lower suitable concentration value at the sub-bin boundary in each box compartment, the final concentration measurement value of each nutrient component is embedded in the corresponding box compartment. If the final concentration measurement value is less than the upper suitable concentration value and greater than the lower suitable concentration value, it means that the final concentration measurement value is within the suitable concentration range of the box compartment. Finally, after all the final concentration measurement values ​​are positioned in the box compartment, a separate sub-bin field will be formed in the multi-dimensional sub-bin field, which reflects the degree of superiority of the final concentration measurement value compared to the concentration level of the expected effect. If the volume value of the non-overlapping multi-dimensional sub-bin field is less than the preset volume value, it means that the concentration level of the final concentration measurement value is not much different from the concentration level required for the expected effect, so the macadamia oil is qualified, otherwise the gap is large, that is, it is an unqualified product. This method can analyze the concentration level of the extracted and prepared macadamia oil according to the final concentration measurement value, so as to ensure that the concentration of the prepared nutrients can achieve the expected application effect quality, and improve the nutritional rationality and application health safety factor of macadamia oil.

[0117] It should be noted that the gradient critical point refers to the gradient dividing point that defines different concentration gradient ranges in the suitable concentration range. For example, the concentration of 80 micrograms to 100 micrograms in the suitable concentration range is one suitable concentration gradient range, and the concentration of 100 micrograms to 150 micrograms is the next suitable concentration gradient range. The gradient critical point can improve the segmentation accuracy of each box compartment.

[0118] In addition, the method for preparing and dispensing macadamia nut oil by nutrient extraction also includes the following steps:

[0119] When there are substandard products in macadamia nut oil, the final concentration measurement value of the nutrients in the substandard products is obtained, and the reasonable concentration measurement value of each nutrient is retrieved from the big data network based on the nutritional preparation requirements;

[0120] Preset an allowable concentration error interval, calculate the concentration measurement deviation value between the final concentration measurement value and the reasonable concentration measurement value, and if the concentration measurement deviation value is not within the allowable concentration error interval, determine the concentration compensation amplitude for compensating the nutrients to reach a reasonable concentration according to the concentration measurement deviation value;

[0121] Obtaining macadamia nut planting cases based on a big data network, extracting the enrichment concentration values ​​of nutrients in macadamia nuts under different planting conditions through the macadamia nut planting cases, calculating the hash function between the enrichment concentration values ​​and the concentration compensation values ​​under different planting conditions, and obtaining multiple hash functions;

[0122] Only the planting conditions corresponding to the enrichment concentration values ​​of the hash function greater than the preset hash function are extracted, and marked as the target planting conditions of macadamia nuts, so as to obtain one or more significant phenotypic characteristics of macadamia nuts under the target planting conditions;

[0123] Constructing a significant phenotypic model of macadamia nuts based on the one or more significant phenotypic characteristics, uploading the significant phenotypic model to a terminal of a macadamia nut nutrition preparation device, and constructing an identification library of the significant phenotypic model;

[0124] Obtain the actual phenotypic model of the current macadamia nut, compare and identify each actual phenotypic model of the current macadamia nut through the recognition library of significant phenotypic models, and calculate the Euclidean distance between the significant phenotypic model and each actual phenotypic model one by one;

[0125] If the Euclidean distance is less than the preset Euclidean distance, the current macadamia nuts are marked as macadamia nuts that can be compensated for and transported to the macadamia nut nutrition preparation equipment. If the Euclidean distance is greater than the preset Euclidean distance, the current macadamia nuts are marked as macadamia nuts that cannot be compensated for and are screened out.

[0126] It should be noted that in the traditional extraction and preparation of macadamia oil, macadamia nuts produced under different planting conditions are usually mixed and put into the macadamia nut nutrition preparation equipment for unified extraction, and the appropriate fruit is not selected for extraction and preparation according to the concentration content of the nutrients rich in macadamia nuts. This phenomenon often leads to a deviation between the final concentration measurement value of the nutrients in the final extracted and prepared macadamia oil and the required reasonable concentration measurement value, making the extracted and prepared macadamia oil nutritionally unbalanced or insufficient, and may also cause the state characteristics of macadamia oil to change and fail to meet the application requirements, reducing the overall nutritional value and application safety of macadamia oil. Therefore, this method can perform concentration compensation analysis on the final concentration measurement value of the prepared macadamia oil, so as to accurately select macadamia fruits that can best compensate for the missing concentration for the high-quality extraction and preparation of macadamia oil, so as to ensure that the nutrient concentration of macadamia oil reaches a reasonable concentration, improve the nutritional balance and application safety factor of macadamia oil, and greatly guarantee product quality.

[0127] In addition, the method for preparing and dispensing macadamia nut oil by nutrient extraction also includes the following steps:

[0128] Obtaining original design drawings of the macadamia nut nutrition preparation equipment, and obtaining the mechanical structure design of each nutrition extraction and preparation mechanism in the macadamia nut nutrition preparation equipment through the original design drawings;

[0129] Based on the mechanical structure design, the standard extraction and preparation posture of each nutrient extraction and preparation institution for macadamia oil is retrieved in the big data network, and several mechanical factors that cause the extraction and preparation metering loss of macadamia nuts are obtained;

[0130] Obtaining the metered loss values ​​of various nutrients in the macadamia oil product, constructing a loss kurtosis graph based on the metered loss values ​​of the various nutrients, and determining the loss trend characteristics of various nutrients in the macadamia oil product by calculating the sharpness of the loss kurtosis in the loss kurtosis graph;

[0131] The FCM algorithm is introduced to initialize the membership matrix of the loss trend characteristics of each nutrient component, the cluster center of each mechanical factor is calculated through the membership matrix, and the membership matrix is ​​updated and iterated based on the cluster center until a preset iteration frequency is reached to obtain the mechanical factor cluster label of each loss trend characteristic;

[0132] The correlation between each mechanical factor and each standard extraction preparation posture is calculated using the Spearson correlation coefficient, and only the extraction preparation operation posture corresponding to the maximum correlation of each mechanical factor is extracted, which is defined as the inducement extraction preparation operation posture;

[0133] Taking mechanical factors as index headers, an inverted index table of inducement extraction preparation operation posture is constructed based on the index header, and the mechanical factor clustering label of each loss trend feature is imported into the inverted index table for inverted retrieval to obtain the inducement extraction preparation operation posture that causes each loss trend feature;

[0134] The generation time nodes of each loss trend feature are obtained, the extraction and preparation operating condition parameters that induce the extraction and preparation operation posture at the generation time nodes are extracted, the extraction and preparation operating condition parameters are adjusted based on the loss kurtosis, and the extraction and preparation optimization strategy of macadamia nut nutrition preparation equipment is obtained.

[0135] It should be noted that due to the long-term operation of the macadamia nut nutrition preparation equipment or the failure to maintain and service it regularly, some nutrient extraction and preparation mechanisms in the macadamia nut nutrition preparation equipment have experienced operation deviations, failures and damage, etc., resulting in the phenomenon of metering loss in the process of extracting and preparing macadamia nut oil, thereby reducing the overall product quality of macadamia nut oil and the rationality of the nutritional balance level, which is not conducive to the high-quality retention of nutrients. Therefore, this method can cluster and analyze the mechanical factors that cause the loss trend characteristics of various nutrients in macadamia nut oil products, and further locate the extraction and preparation operation posture of the inducement that produces the loss trend characteristics according to the clustering labels of the mechanical factors. By adjusting the extraction and preparation working parameters of these inducement extraction and preparation operation postures, the metering loss phenomenon in the process of extracting and preparing macadamia nut oil by the macadamia nut nutrition preparation equipment can be repaired or eliminated, the overall nutritional value of macadamia nut oil is retained to the greatest extent, and the cost increase caused by the loss of macadamia nut oil extraction and preparation is reduced, with high economic benefits. Among them, the FCM algorithm, whose Chinese name is fuzzy clustering algorithm, can process data sets that are not clearly divided by gradually optimizing the membership matrix and cluster centers, allowing data points to have different degrees of membership in multiple clusters. Compared with traditional clustering algorithms, it is more flexible and can better handle fuzzy or fuzzy-boundary data. At the same time, it can reduce the impact of noise or abnormal points on the clustering results of mechanical factors by allocating part of the membership to points that are not obviously classified into a certain category.

[0136] The second aspect of the present invention provides a mixing and metering system for preparing macadamia nut oil nutrition extraction, such as Figure 4 As shown, the mixing and metering system for preparing nutritional extraction of macadamia oil includes a memory 41 and a processor 42. The memory 41 stores a mixing and metering method program for preparing nutritional extraction of macadamia oil. When the mixing and metering method program for preparing nutritional extraction of macadamia oil is executed by the processor 42, any one of the steps of the mixing and metering method for preparing nutritional extraction of macadamia oil is implemented.

[0137] The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A method for preparing and dispensing macadamia nut oil by nutrient extraction, characterized in that: The following steps are involved: S102: Obtaining the application scenario and nutritional preparation requirements of macadamia nut oil, analyzing and calculating several groups of historical preparation interaction indicator variables of various nutritional components through the application scenario and nutritional preparation requirements, so as to determine the interaction mechanism index between the various nutritional components, constraining the nutritional ratio value according to the interaction mechanism index, and obtaining the nutritional ratio value range of various nutritional components; S104: obtaining one or more target state characteristics of macadamia nut oil, adaptively learning and constructing a nutrient ratio interaction network in which each nutrient ratio value in the nutrient ratio value range meets the state characteristics through a mean square error regression equation, and outputting an ideal blending ratio of each nutrient component meeting the state characteristics through the deployment of the nutrient ratio interaction network; S106: Obtaining the process state variables of the macadamia nut oil nutrient extraction preparation process, globally optimizing the process state variables through each of the ideal blending ratios to obtain the expected process state variables, analyzing the feedback gain matrix of the current process state control based on the expected process state variables, calibrating the current process using the feedback gain matrix, and obtaining accurate blending process control parameters; S108: Based on the accurate blending process control parameters, the actual ratio of each nutrient component in the macadamia oil sample and the corresponding concentration peak area are obtained, and the actual ratio and the corresponding concentration peak area are hashed and cumulatively measured and analyzed with the ideal blending ratio as the bubble volume benchmark to obtain the measurement result.

2. The method for preparing and dispensing macadamia nut oil by nutrient extraction according to claim 1, characterized in that: The step S102 specifically includes the following steps: Obtain the application scenarios and nutritional preparation requirements of macadamia oil, and obtain the various nutrients necessary for macadamia oil in the application scenarios based on the big data network; Based on the nutritional preparation requirements, the optimal nutritional supply base corresponding to the application scenario of various nutritional components in macadamia nut oil and the logic of the interaction mechanism between each nutritional component are retrieved in the big data network; Obtain the nutritional extraction and preparation logs of macadamia nuts, and extract several groups of historical preparation interaction indicator variables on the optimal nutritional supply base between each nutritional component of macadamia nut oil through the nutritional extraction and preparation logs; Determine the dependency relationship between each nutrient component according to the interaction mechanism logic, construct a variable dependency directed graph based on the dependency relationship, import several groups of historically prepared interaction indicator variables into the variable dependency directed graph for fitting definition, and generate a Belief network model; The joint probability distribution between each nutrient component is calculated according to the Belief network model analysis, and the marginalized probability summation operation is performed on the joint probability distribution to obtain the marginal probability distribution. The interaction coefficient of each group of historical preparation interaction indicator variables is approximated based on the joint probability distribution and the marginal probability distribution, and the interaction mechanism index between various nutrients is determined according to the interaction coefficient; The expected interaction mechanism index is preset according to the nutritional preparation requirements. If the interaction mechanism index exceeds the expected interaction mechanism index, the index difference between the interaction mechanism index and the expected interaction mechanism index is calculated to obtain multiple interaction mechanism index deviation values, and a discrete inclusion constraint matrix is ​​established through the multiple interaction mechanism index deviation values. Through the big data network, various nutrients are obtained to achieve multiple nutritional ratios corresponding to the optimal nutritional supply base, and a backtracking algorithm is introduced. In the backtracking algorithm, each nutritional ratio of various nutrients is discretely combined until it approaches the expected interaction mechanism index, and multiple discrete combination sub-matrices are output; If the discrete combination submatrix exists in the discrete inclusion constraint matrix, the combined nutritional ratios of various nutrients in the discrete combination submatrix are retained, and finally the nutritional ratio value range of various nutrients is formed.

3. The method for preparing and dispensing macadamia nut oil by nutrient extraction according to claim 1, characterized in that: The step S104 specifically includes the following steps: Constructing a sparse feature dictionary, performing sparse feature encoding on the nutritional ratio value ranges of the various nutritional components through the sparse feature dictionary, and generating a nutritional ratio sparse feature matrix; Obtain one or more target state characteristics of macadamia oil through nutritional preparation requirements, and set linear weights, global bias terms and latent vectors of the target state characteristics according to the nutritional ratio sparse feature matrix; Based on the linear weight, global bias term and latent vector, a state characteristic mean square error regression equation is defined and established, and the first-order moment estimation parameters and the second-order moment estimation parameters of the state characteristic mean square error regression equation are initialized based on the nutrient ratio coefficient characteristic matrix; An adaptive gradient algorithm is introduced to calculate an adaptive gradient value between a sparse characteristic matrix of a nutrient ratio and a mean square error of a state characteristic, and a first-order moment estimation parameter and a second-order moment estimation parameter are updated based on the adaptive gradient value to obtain an updated first-order moment estimation parameter and an updated second-order moment estimation parameter; Correcting the deviation between the updated first-order moment estimation parameter and the first-order moment estimation parameter to obtain the first-order moment estimation deviation correction parameter, correcting the deviation between the updated second-order moment estimation parameter and the second-order moment estimation parameter to obtain the second-order moment estimation deviation correction parameter; Based on the first-order moment estimation deviation correction parameter and the second-order moment estimation deviation correction parameter, the learning rate of each nutrient ratio sparse feature in the nutrient ratio sparse feature matrix for the state characteristic mean square error is solved to obtain an adaptive gradient learning rate for each nutrient ratio sparse feature; Designing an interactive term update rule of the sparse features of the nutrient ratio through an adaptive gradient learning rate, updating and iterating the sparse feature matrix of the nutrient ratio based on the interactive term update rule, and generating a nutrient ratio interactive network that conforms to the state characteristics; The nutrient ratio interaction network is deployed in the nutrient ratio value domain, and finally outputs an ideal mixing ratio for each nutrient component that meets the state characteristics.

4. The method for preparing and dispensing macadamia nut oil by nutrient extraction according to claim 1, characterized in that: The step S106 specifically includes the following steps: According to the ideal blending ratio of each nutrient component in accordance with the state characteristics, the macadamia oil is blended and extracted in the macadamia oil nutrient extraction and preparation process, and the process state variables and process control input parameters are obtained at this time; Obtaining the actual blending ratio of each nutrient component in the extracted macadamia nut oil, presetting an allowable blending ratio drift threshold based on nutrient preparation requirements and an optimal nutrient supply base, calculating a blending ratio error between the actual blending ratio value and a corresponding ideal blending ratio, and obtaining a blending ratio deviation value; If the blending ratio deviation value is greater than the allowable blending ratio drift threshold, a genetic algorithm is introduced, and based on the fitness of the offspring individuals of the process state variables of the blending ratio deviation value in the genetic algorithm, the parent generation is selected through the offspring individual fitness and a crossover mutation process is performed to obtain a new offspring individual; Repeat the above steps to continuously update and iterate the new offspring individuals to replace all the old offspring individuals, forming a new population of process state variables, and extract the process state variable with the maximum offspring individual fitness from the new population of process state variables, which is defined as the expected process state variable; A state control model of the macadamia oil nutrient extraction and preparation process is constructed based on the process state variables and process control parameters, a hash state value between the expected process state variables and the process state variables is calculated, and a cost function of the control input for the process state difference is determined based on the hash state value; Constructing an expected state control algebraic differential equation of a state control model according to expected process state variables, solving the expected state control algebraic differential equation through a cost function to obtain a feedback gain matrix, and globally calibrating the state control model based on the feedback gain matrix to obtain a calibrated state control model; The calibrated state control model is applied to the macadamia oil nutrient extraction and preparation process to obtain accurate formulation process control parameters of the macadamia oil nutrient extraction and preparation process.

5. The method for preparing and dispensing macadamia nut oil by nutrient extraction according to claim 1, characterized in that: The step S108 specifically includes the following steps: Obtaining the variety information of the current macadamia nuts, and controlling the macadamia nut oil nutrient extraction and preparation process to extract and prepare the current macadamia nuts according to the accurate blending process control parameters to obtain a macadamia nut oil sample; Performing nutritional component testing on the macadamia nut oil sample by liquid chromatography to obtain the actual ratio of each nutritional component in the macadamia nut oil sample and the corresponding concentration peak area; Construct a bubble chart, preset a bubble volume limit threshold based on the ideal mixing ratio, define and plan the bubble volume of each nutrient in the bubble chart until the corresponding bubble volume limit threshold is reached, and generate an ideal mixing bubble volume range corresponding to each nutrient; Based on the preset initial bubble color cumulative hash value of the ideal blended bubble volume range, the actual ratio of each nutrient component and the corresponding concentration peak area are bundled and converted into a bubble color coding character, and a hash algorithm is introduced to calculate the bubble color hash value of each bubble color coding character; By accumulating and splicing each bubble color coding character within the ideal blending bubble volume range of the corresponding nutrient component, during the accumulative splicing process, the initial bubble color cumulative hash value is continuously updated based on the bubble color hash value of each bubble color coding character to generate a hash cumulative table for each nutrient component; Traverse the actual ratio and the corresponding concentration peak area in the metering hash accumulation table to output the final bubble color cumulative hash value of the ideal mixing bubble volume range, and determine the final concentration metering value of each nutrient when it reaches the corresponding ideal mixing ratio according to the final bubble color hash accumulation value; Based on the rationality of the concentration effect of the final concentration measurement value when each nutrient reaches the corresponding ideal blending ratio, the quality analysis and determination of the macadamia oil product are carried out to obtain the measurement results.

6. The method for preparing and dispensing macadamia nut oil by nutrient extraction according to claim 5, characterized in that: The S108 step is based on the rationality of the concentration effect of the final concentration measurement value when each nutrient component reaches the corresponding ideal blending ratio to perform a quality analysis and determination on the macadamia oil product to obtain the measurement result, which specifically includes the following steps: According to the nutritional preparation requirements, the expected effect coefficient of macadamia nut oil in the application scenario is obtained, and based on the expected effect coefficient, the appropriate concentration range of each nutrient component is retrieved in the big data network; Constructing a multidimensional binning domain, dividing the multidimensional binning domain into sub-binning domains representing each nutrient component based on multiple nutrients, and obtaining a plurality of gradient critical points of each of the suitable concentration ranges; Taking the root node as the starting node, each gradient critical point is used to split the appropriate concentration value between two adjacent gradient critical points in the corresponding appropriate concentration range, and a number of branch nodes are generated. Repeat the above steps to continuously split the appropriate concentration value of each branch node, and finally form a binning decision tree; Obtain all leaf node values ​​of the binning decision tree, determine the binning points of each suitable concentration range according to all leaf node values, divide each sub-binning area into N bins based on the binning points, and calculate the upper and lower suitable concentration values ​​at the binning boundaries in each bin; If the final concentration measurement value is less than the upper suitable concentration value and greater than the lower suitable concentration value, the final concentration measurement value is embedded in the box compartment corresponding to the upper suitable concentration value and the lower suitable concentration value to obtain a fixed-point box compartment; Obtain the fixed-point bins and cells whose final concentration measurement values ​​of all nutrients are in the corresponding sub-binning areas, strip off the multidimensional binning areas formed by all the fixed-point bins and cells, define them as the first multidimensional binning areas, and plan out the expected multidimensional binning areas where all nutrients reach the expected effect coefficients, define them as the second multidimensional binning areas; The overlapping area between the first multidimensional binning area and the second multidimensional binning area is eliminated to obtain a non-overlapping multidimensional binning area. If the volume value of the non-overlapping multidimensional binning area is less than the preset volume value, the macadamia oil is calibrated as a qualified product. If the volume value of the non-overlapping multidimensional binning area is greater than the preset volume value, the macadamia oil is calibrated as an unqualified product to obtain a measurement result.

7. A mixing and metering system for extracting and preparing macadamia nut oil, characterized in that: The preparation and metering system for the nutritional extraction and preparation of macadamia oil includes a memory and a processor, wherein the memory stores a program for the preparation and metering method for the nutritional extraction and preparation of macadamia oil. When the program for the preparation and metering method for the nutritional extraction and preparation of macadamia oil is executed by the processor, the steps of the preparation and metering method for the nutritional extraction and preparation of macadamia oil as described in any one of claims 1 to 6 are implemented.

Images