Intelligent control method for ginger and lemon mixed juice production process

Abstract

The present application relates to the technical field of fruit juice production control, and particularly relates to an intelligent control method for ginger and lemon mixed fruit juice production process. The method comprises the following steps: constructing a structure by using pressure, time and temperature during juicing; obtaining the number of initial parent structures; obtaining the first weight of juice yield, energy consumption and vitamin C retention rate according to historical fruit juice data; obtaining the second weight of juice yield, energy consumption and vitamin C retention rate by using quantitative experiments of the current batch of ginger and lemon; then obtaining the final weight of juice yield, energy consumption and vitamin C retention rate; then constructing a fitness calculation formula; constructing a parent structure set and obtaining the fitness of the structure; obtaining the optimal structure based on the parent structure set by using genetic algorithm and combining the fitness of each structure in the iteration process; and juicing the current batch of ginger and lemon based on the optimal structure. The present application can effectively control the juicing process of ginger and lemon fruit juice.

Description

Technical Field

[0001] This invention relates to the field of juice production control technology, specifically to an intelligent control method for the production process of ginger and lemon mixed juice. Background Technology

[0002] In traditional ginger and lemon mixed juice production processes, many key steps rely on manual operation and experience. For example, in the raw material cleaning stage, manual cleaning makes it difficult to ensure the thorough removal of dirt, pesticide residues, and other impurities from the surface of ginger and lemons. Furthermore, the cleaning intensity and time are difficult to control precisely, easily leading to incomplete cleaning or over-cleaning of some raw materials, resulting in nutrient loss. During the juicing process, manually adjusting the pressure and speed of the juicing equipment cannot be precisely tailored to the texture and moisture content of different batches of raw materials, resulting in inconsistent juicing results, unstable juice yield, and impacting product quality and production efficiency.

[0003] Given the varying textures and moisture content of different batches of raw materials, the optimal settings for various parameters during the juicing process can be obtained using a genetic algorithm. However, genetic algorithms suffer from several drawbacks: parameter selection is sensitive, potentially leading to slow or premature convergence. Furthermore, the design of the fitness function is complex, with weight allocation exhibiting strong subjectivity (e.g., the weight ratio of juice yield to energy consumption), which may affect the reasonableness of the results. Improper constraint handling can result in an excessively low proportion of feasible solutions, reducing search efficiency. Summary of the Invention

[0004] To address the aforementioned technical problems, the present invention aims to provide an intelligent control method for the production process of ginger and lemon mixed juice, the specific technical solution of which is as follows: One embodiment of the present invention provides an intelligent control method for the production process of ginger and lemon mixed juice, the method comprising: The structure is constructed using the pressure, time, and temperature during juicing; the initial number of parent structures is obtained by using the range of various parameters in the structure and the maximum number of parent structures. The first weights for juice yield, energy consumption, and vitamin C retention rate were obtained by using the average juice yield, average energy consumption, and average vitamin C retention rate of each batch of juice from historical data. The second weights for juice yield, energy consumption, and vitamin C retention rate were obtained by conducting quantitative experiments on the current batch of ginger and lemon juice based on the pressure, time, and temperature during juicing. The final weights for juice yield, energy consumption, and vitamin C retention rate were obtained based on the first and second weights. A fitness calculation formula was constructed using the juice yield, energy consumption, vitamin C retention rate, and the corresponding final weights. A set of parent structures is constructed based on the initial number of parent structures; the fitness of the structures is obtained according to the fitness calculation formula and combined with experiments; the optimal structure is obtained by using a genetic algorithm to iterate based on the set of parent structures and combining the fitness of each structure during the iteration process; the pressure, time and temperature in the optimal structure are used as parameters for juicing ginger and lemon in the current batch.

[0005] Preferably, the initial number of parent structures is obtained by utilizing the range of various parameters in the structure and the maximum number of parent structures, including: Based on the pressure range and the pressure adjustment unit, obtain the number of pressure values; based on the time range and the time adjustment unit, obtain the number of time values; based on the temperature range and the temperature adjustment unit, obtain the number of temperature values; the product of the number of pressure values, the number of time values, and the number of temperature values ​​is the maximum population size. The proportional coefficients corresponding to pressure are obtained by comparing the number of pressure values ​​used when juicing each batch of raw materials with the number of pressure values. Similarly, the proportional coefficients corresponding to time and temperature are obtained. The proportional coefficients corresponding to pressure, time and temperature are multiplied together, and the result of multiplying the result by the maximum population size and the number of parameter types is rounded to obtain the initial number of parent structures.

[0006] Preferably, the first weights for juice yield, energy consumption, and vitamin C retention rate are obtained by using the average juice yield, average energy consumption, and average vitamin C retention rate of historical batches of juice, respectively, including: The average energy consumption ratio is obtained by comparing the average energy consumption of each batch of juice in history with the sum of the energy consumption of each batch of juice in history. The average energy consumption result is obtained by subtracting the average energy consumption ratio from the first preset value. The first weights of juice yield, vitamin C retention rate and energy consumption are obtained by dividing the average juice yield, the average vitamin C retention rate and the average energy consumption result of each batch of juice in history by the sum of the average juice yield, the average vitamin C retention rate and the average energy consumption result of each batch of juice in history.

[0007] Preferably, a second weighting of juice yield, energy consumption, and vitamin C retention rate is obtained by quantitatively analyzing the pressure, time, and temperature during juicing of the current batch of ginger and lemon, including: Pressure, time, and temperature are denoted as different types of process parameters, and juice yield, energy consumption, and vitamin C retention rate are denoted as different types of fitness parameters. One fitness parameter is randomly selected as the target fitness parameter. A sequence of target fitness parameters is obtained when a process parameter changes by the minimum unit of change in a quantitative experiment; this sequence is denoted as the target fitness parameter sequence corresponding to the minimum unit of change of that process parameter. The average value of the first-order difference sequence of the target fitness parameter sequence corresponding to that process parameter is calculated and denoted as the average change value of the target fitness parameter corresponding to the minimum unit of change of that process parameter. A sequence of different values ​​of that process parameter when it changes by the minimum unit of change is obtained and denoted as the process parameter sequence. The process parameter sequence is used as the horizontal axis, and the target fitness parameter sequence corresponding to the minimum unit of change of that process parameter is used as the vertical axis for linear fitting to obtain the fitted linear slope. The product of the absolute value of the fitted linear slope and the average change value of the target fitness parameter corresponding to the minimum unit of change of that process parameter is mapped using an exponential function with a base of the natural constant to obtain the first mapping value of that process parameter for the target fitness parameter. Calculate the difference between the maximum and minimum values ​​of a process parameter when it changes by one unit of variation, and record this as the maximum difference of the process parameter corresponding to that unit of variation. Obtain the ratio of this unit of variation to the maximum difference of the process parameter corresponding to that unit of variation, and record this as the change ratio of the process parameter when it changes by that unit of variation. Divide the average change value of the target fitness parameter corresponding to the change of the process parameter by the difference between the maximum and minimum values ​​in the sequence of target fitness parameters corresponding to the change of the process parameter by that unit of variation to obtain the average change ratio of the target fitness parameter when it changes by that unit of variation. Calculate the difference between the average change ratio of the target fitness parameter when it changes by that unit of variation and the change ratio of the process parameter when it changes by that unit of variation, and record this as the change difference corresponding to the change of the process parameter when it changes by that unit of variation. Calculate the mean of the change differences corresponding to the change of the process parameter with respect to various units of variation, and record this as the average change difference of the process parameter with respect to the target fitness parameter. The influence characteristic value of the process parameter on the target fitness parameter is obtained by multiplying the first mapping value of the process parameter on the target fitness parameter with the average change difference of the process parameter on the target fitness parameter. The second weight of the target fitness parameter is obtained by obtaining the difference of the influence characteristic value of each process parameter on the target fitness parameter. Similarly, the second weights of the other two fitness parameters are obtained.

[0008] Preferably, the final weights of juice yield, energy consumption, and vitamin C retention rate are obtained based on a first weight and a second weight, including: The final weight of juice yield is obtained by multiplying the first and second weights of juice yield; the final weight of energy consumption is obtained by multiplying the first and second weights of energy consumption; and the final weight of vitamin C retention rate is obtained by multiplying the first and second weights of vitamin C retention rate.

[0009] Preferably, the fitness calculation formula is as follows: , in, Indicates fitness; , , The final weights for juice yield, energy consumption, and vitamin C retention rate are respectively represented. , , These represent juice yield, energy consumption, and vitamin C retention rate, respectively.

[0010] Preferably, the optimal structure is obtained by iteratively applying a genetic algorithm based on the parent structure set and combining the fitness of each structure during the iteration process, including: Based on the set of parent structures, a genetic algorithm is used for iteration. When the iteration stops, the iteration stops, and the structure with the highest fitness among the offspring is taken as the optimal structure.

[0011] Preferably, the iteration stopping condition is as follows: When using a genetic algorithm to iterate using the set of parent structures, the standard deviation of the fitness of each structure in a generation is obtained as the fitness standard deviation of that generation. If the fitness standard deviation of N consecutive generations is less than or equal to a preset threshold, then population reduction is performed, and the parents of the latest generation and the top CN1 structures with the highest fitness in the latest generation are retained as new parents. If the fitness standard deviation of the new parents is less than or equal to the preset threshold, then the iteration stops.

[0012] The embodiments of this invention have at least the following beneficial effects: This application encodes and obtains structures based on the pressure, time, and temperature during juicing, and then uses the range of various parameters in the structure and the maximum number of parent structures to obtain the initial number of parent structures; then, it uses the average juice yield, average energy consumption, and average vitamin C retention rate of each batch of juice in history to obtain the first weights of juice yield, energy consumption, and vitamin C retention rate respectively; it takes the current batch of ginger and lemon and conducts quantitative experiments based on the pressure, time, and temperature during juicing to obtain the second weights of juice yield, energy consumption, and vitamin C retention rate, and then combines the first and second weights of juice yield, energy consumption, and vitamin C retention rate to obtain the final weights of juice yield, energy consumption, and vitamin C retention rate, and then constructs a fitness calculation formula; then, it constructs a set of parent structures based on the initial number of parent structures; it obtains the fitness of the structures according to the fitness calculation formula and combined with experiments, and finally, it uses a genetic algorithm to iterate based on the set of parent structures and combines the fitness of each structure during the iteration process to obtain the optimal structure; it uses the pressure, time, and temperature in the optimal structure as parameters for juicing the current batch of ginger and lemon. This allows for control over the process parameters during the juicing of each batch of raw materials, thereby improving the production quality of ginger and lemon mixed juice. Attached Figure Description

[0013] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0014] Figure 1 A flowchart of a method for intelligent control of a ginger and lemon mixed juice production process provided in an embodiment of the present invention. Detailed Implementation

[0015] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of an intelligent control method for the production process of ginger and lemon mixed juice according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0016] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0017] The following description, in conjunction with the accompanying drawings, details the specific scheme of the intelligent control method for the production process of ginger and lemon mixed juice provided by the present invention.

[0018] Example: The main application scenario of this invention is: This application uses a genetic algorithm to optimize the process parameters in the production of ginger and lemon juice, thereby achieving intelligent control of the production of ginger and lemon juice.

[0019] Please see Figure 1 The diagram illustrates a method flowchart for intelligent control of a ginger-lemon mixed juice production process according to an embodiment of the present invention. The method includes the following steps: Step S1: Construct a structure using the pressure, time, and temperature during juicing; obtain the initial number of parent structures using the range of various parameters in the structure and the maximum number of parent structures.

[0020] This application primarily involves adjusting various parameters of the juicer during juicing to control the juicing process. This requires obtaining historical data on the pressure, time, and temperature used by the juicer to juic each batch of raw materials. It also requires obtaining the range of pressure, time, and temperature values ​​during juicing, as well as their respective adjustment units. For example, if the juicing time range is 30 minutes and the adjustment unit is 1 minute, then adjustments are made every one minute.

[0021] In order to provide feedback on the impact of different combinations of pressure, time and temperature on juicing, it is also necessary to obtain the juice yield, energy consumption and vitamin C retention rate (vitamin C retention rate) after the juicer has juiced each batch of raw materials in history. Juice yield = (weight of extracted juice ÷ weight of raw materials) × 100%. Energy consumption is obtained from the energy management system (EMS). The vitamin C retention rate is obtained by separating and quantitatively analyzing vitamin C and its derivatives using high performance liquid chromatography (HPLC). The vitamin C content in lemons is 50mg / 100g, and the juice produced is 40mg / 100g, so the vitamin C retention rate is 80%.

[0022] This application primarily involves adjusting various parameters of the juicer during juicing to control the juicing process. This requires obtaining historical data on the pressure, time, and temperature used by the juicer to juic each batch of raw materials. It also requires obtaining the range of pressure, time, and temperature values ​​during juicing, as well as their respective adjustment units. For example, if the juicing time range is 30 minutes and the adjustment unit is 1 minute, then adjustments are made every one minute.

[0023] In order to provide feedback on the impact of different combinations of pressure, time and temperature on juicing, it is also necessary to obtain the juice yield, energy consumption and vitamin C retention rate (vitamin C retention rate) after the juicer has juiced each batch of raw materials in history. Juice yield = (weight of extracted juice ÷ weight of raw materials) × 100%. Energy consumption is obtained from the energy management system (EMS). The vitamin C retention rate is obtained by separating and quantitatively analyzing vitamin C and its derivatives using high performance liquid chromatography (HPLC). The vitamin C content in lemons is 50mg / 100g, and the juice produced is 40mg / 100g, so the vitamin C retention rate is 80%.

[0024] Furthermore, the genetic algorithm needs to encode the parameters that affect the results and construct chromosomes, which are referred to as structures in this application. A structure is constructed by the pressure, time and temperature during juicing, that is, the structure of the structure is [pressure, time, temperature], using real number parameters, that is, directly using real numbers to represent the parameters (e.g., pressure = 32.5 MPa).

[0025] Since the population size parameter affects the algorithm's search capability, a population that is too small can lead to insufficient solution space coverage and a tendency to get stuck in local optima (e.g., only finding solutions with "high juice yield but poor flavor"), while a population that is too large can lead to a surge in computation and a decrease in convergence speed, but a more comprehensive search. It also affects convergence speed; for example, small populations converge quickly initially (due to small differences between individuals), but are prone to stagnation later (due to a lack of new information), while large populations converge slowly initially (requiring more iterations to select high-quality individuals), but converge more stably later (due to high diversity). Furthermore, it affects solution quality; small populations result in large fluctuations in solution quality (potentially getting stuck in suboptimal solutions due to randomness), while large populations offer more stable solution quality (due to multiple independent searches reducing the impact of randomness). Therefore, the population size setting needs to be determined based on the specific data conditions.

[0026] Therefore, the maximum population size can be obtained. Since the structure is [pressure, time, temperature], the maximum population size is the permutation and combination of these three data types. Specifically, it needs to be calculated based on the range of pressure, time, and temperature values. Specifically, based on the pressure range and the unit of adjustment, the number of pressure values ​​is obtained; based on the time range and the unit of adjustment, the number of time values ​​is obtained; based on the temperature range and the unit of adjustment, the number of temperature values ​​is obtained. The product of the number of pressure values, the number of time values, and the number of temperature values ​​is the maximum population size.

[0027] For example, assuming the pressure range is 0.3MPa-3.0MPa and the adjustment unit is 0.1, there are actually 28 possible data points. The time range is 1-10 minutes and the adjustment unit is 30 seconds, so there are actually 20 possible data points. The temperature range is 5-35℃ and the adjustment unit is 1℃, so there are actually 31 possible data points. Therefore, the maximum population size is the permutation and combination of the above data, that is, 28*20*31=17360.

[0028] Using the maximum population size would result in an excessively large calculation, and this scale only applies to one raw material. For two raw materials, the amount of data required for experiments would be even greater, making this unnecessary. Therefore, a more reasonable value needs to be set for the population size, and the optimal structure can be obtained through a genetic algorithm.

[0029] To ensure comprehensive population coverage, historical parameter statistics can be analyzed. While different batches of ginger and lemons may vary, similar raw materials are typically selected during initial screening. Therefore, the usage of various parameters may primarily fall within a certain range or value. Population size can be reduced based on this range. The combined coefficients of all parameters used during juicing in historical data are calculated based on their respective statistical proportions. This coefficient, combined with the maximum population size, yields the minimum population size. Since the structure involves three parameters, to ensure the population size covers key regions of the solution space, the parameter dimensions are expanded based on the minimum population size to obtain the initial population size, which is also the initial number of parent structures.

[0030] Specifically, the proportional coefficient corresponding to pressure is obtained by comparing the number of pressure values ​​used when juicing each batch of raw materials with the number of pressure values; similarly, the proportional coefficient corresponding to time and the proportional coefficient corresponding to temperature are obtained; the proportional coefficient corresponding to pressure, the proportional coefficient corresponding to time and the proportional coefficient corresponding to temperature are multiplied together, and the result of multiplying the result by the maximum population size and the number of parameter types and then rounding down is obtained to obtain the initial number of parent structure.

[0031] The specific calculation model for the initial number of parent structure elements is as follows: , Where CN represents the initial number of parent structures, and round represents the rounding function. This represents the total number of optional data for the i-th type of parameter in the structure, such as the number of pressure values. This represents the number of historical statistical data types used for the i-th type of parameter in the structure, such as the number of pressure values ​​used when juicing each batch of raw materials. This represents the proportionality coefficient corresponding to the i-th type of parameter, such as the proportionality coefficient corresponding to pressure. This represents the product of the proportionality coefficients corresponding to pressure, time, and temperature; n represents the number of parameter types in the structure, which is 3. This indicates the maximum population size.

[0032] After obtaining the initial number of parent structures, the population size may still be too large during subsequent iterations using the genetic algorithm. Therefore, in subsequent iterations, iteration rules can be set according to the actual situation, and the population size can be reduced after a certain number of iterations.

[0033] Step S2: Obtain the first weights of juice yield, energy consumption, and vitamin C retention rate by using the average values ​​of juice yield, energy consumption, and vitamin C retention rate of each batch of juice in history; take the current batch of ginger and lemon and conduct quantitative experiments based on the pressure, time, and temperature during juicing to obtain the second weights of juice yield, energy consumption, and vitamin C retention rate; obtain the final weights of juice yield, energy consumption, and vitamin C retention rate based on the first and second weights; construct the fitness calculation formula using the juice yield, energy consumption, vitamin C retention rate, and the corresponding final weights.

[0034] When determining the formula for calculating the fitness of juicing production process parameters, it is necessary to select key indicators that significantly affect juice quality and production efficiency based on market taste survey results and process optimization goals. This scheme uses juice yield, energy consumption, and vitamin C retention rate as the core parameters for building fitness in juicing production.

[0035] The weighting of these parameters needs to consider two aspects: First, the priority ranking based on the intended use of the juice. Different application scenarios have different sensitivities to parameters, and weights need to be allocated according to the objective. For example, if the objective is economic benefit (such as industrial production), then the juice yield is directly related to the raw material utilization rate and production cost, and its importance is relatively high, so it can be given a higher weight. Energy consumption affects long-term operating costs and is also relatively important, so it can be given a higher weight. The importance of vitamin C retention rate can be appropriately reduced and a lower weight set if basic nutritional needs are met. If the objective is quality, then the importance of vitamin C retention rate increases, maintaining a similar importance to juice yield, while the importance of energy consumption decreases. If the objective is function (such as anti-oxidation, enhancing immunity), then vitamin C retention rate is the most important, and the other parameters are less important. Therefore, it is necessary to analyze the pressure, time, and temperature during the juicing of each batch of raw materials in history.

[0036] The second approach is sensitivity analysis based on the correlation between process parameters and results. This mainly focuses on the degree of impact of changes in process parameters on the result data. For example, when different process parameters change, even a small numerical change can lead to a significant change in a certain result parameter, and the affected parameters are also different. For instance, juice yield is mainly affected by pressure data, but the sensitivity of the impact varies at different stages. For example, increasing the pressure from 10MPa to 20MPa may increase the juice yield by 15%-20%, but at 20-30MPa, it only increases by 5%. Energy consumption is affected not only by pressure data but also by juicing time. For example, for every 10MPa increase in pressure, energy consumption increases by about 10%-15%, while time has a greater impact on pressure. Vitamin C retention rate is affected not only by pressure but also, due to the oxidation characteristics of vitamin C, by temperature and time. Time mainly affects its oxygen exposure time. In a prolonged oxygen environment, oxidation reactions are more likely to occur, leading to a decrease in vitamin C retention rate. At the same time, vitamin C is heat-sensitive; for every 10°C increase in temperature, the degradation rate increases by 2-4 times (e.g., 80% retention rate at 40°C for 10 minutes, but only 50% at 60°C). Results data with a high degree of impact should be given priority and assigned a larger weight, while those with a lower impact should be assigned a smaller weight.

[0037] Therefore, the first weights of juice yield, energy consumption, and vitamin C retention rate are obtained by using the average values ​​of juice yield, energy consumption, and vitamin C retention rate of each batch of juice in history.

[0038] Specifically, the average energy consumption ratio is obtained by comparing the average energy consumption of each batch of juice in history with the sum of the energy consumption of each batch of juice in history; the average energy consumption result is obtained by subtracting the average energy consumption ratio from the first preset value; the first weights of the juice yield, vitamin C retention rate and energy consumption are obtained by dividing the average juice yield, vitamin C retention rate and energy consumption of each batch of juice in history by the average juice yield, vitamin C retention rate and energy consumption of each batch of juice in history by the sum of the average juice yield, vitamin C retention rate and energy consumption of each batch of juice in history.

[0039] When analyzing energy consumption, it is necessary to calculate the ratio of the average energy consumption of each batch of juice in history to the total energy consumption of each batch of juice in history. This is the average energy consumption percentage. The average energy consumption percentage is obtained by subtracting the average energy consumption percentage from the first preset value 1. Since a lower average energy consumption percentage indicates that production pays more attention to energy consumption, its initial importance will be greater. Therefore, the average energy consumption percentage is subtracted from the first preset value 1 as the average energy consumption result to participate in the calculation of the first weight of energy consumption.

[0040] After obtaining the initial first weight, the initial weight needs to be adjusted based on the sensitivity analysis of the correlation between process parameters and results. Since the influence of each parameter on different processes is different, it is necessary to calculate the adjustment weight of the first weight based on the influence of juice yield, vitamin C retention rate and energy consumption under different process parameters (pressure / time / temperature) according to quantitative experiments. This adjustment weight is the second weight of juice yield, energy consumption and vitamin C retention rate.

[0041] A quantitative experiment was conducted on the current batch of ginger and lemons to obtain the second weights of juice yield, energy consumption, and vitamin C retention rate based on the pressure, time, and temperature during juicing.

[0042] When conducting quantitative experiments using the current batch of ginger and lemons, it is necessary to prepare experimental materials and equipment, specifically: 1. Material preparation: Use fresh ginger and lemons from the current batch (to ensure uniform ripeness and reduce the impact of individual differences on the results).

[0043] 2. Equipment Preparation: Juicer, adjustable pressure and time screw press or hydraulic juicer; temperature control device, constant temperature water bath or heating mantle, for controlling juice temperature; measuring tools, electronic balance (for weighing raw materials and juice), graduated cylinder (for measuring juice volume), saccharimeter (to help determine juice quality), vitamin C test kit or high performance liquid chromatograph (HPLC, for accurate determination of vitamin C content); energy consumption monitoring equipment: power meter or smart meter, to record energy consumption during the juicing process.

[0044] Further, it is necessary to design experimental parameters, specifically: The effects of pressure, time, and temperature on juice yield, energy consumption, and vitamin C retention were studied using a single-factor experimental method.

[0045] 1. Pressure: Set various units of pressure variation, for example, one of the units of variation is 0.5MPa, and the initial value of the pressure is 0.5MPa. Then obtain the pressure gradients of 0.5MPa, 1.0MPa, 1.5MPa, and 2.0MPa, fix the time and temperature (using the average time and average temperature of each batch of raw material during juicing in history), observe the effect of pressure changes on juice yield, energy consumption, and vitamin C retention rate, and at the same time obtain the smallest unit of pressure variation among the various units of pressure variation.

[0046] 2. Time: Set various units of time variation, for example, one unit of variation is 5 minutes, and the initial pressure value is 5 minutes. Then obtain time gradients of 5 minutes, 10 minutes, 15 minutes, and 20 minutes, fix the pressure and temperature (using the average pressure and average temperature of each batch of raw materials during juicing in history), observe the effect of time variation on juice yield, energy consumption, and vitamin C retention rate, and at the same time obtain the smallest unit of pressure variation among the various units of time variation.

[0047] 3. Temperature: Set various units of temperature variation, for example, one of the units of variation is 5℃, and the initial value of pressure is 5℃. Then obtain time gradients of 5℃, 10℃, 15℃, and 20℃, fix the time and pressure (using the average pressure and average time of each batch of raw material juiced in history), observe the effect of temperature changes on juice yield, energy consumption, and vitamin C retention rate, and at the same time obtain the smallest unit of temperature variation among the various units of temperature variation.

[0048] When conducting experiments, ensure that the raw materials used are consistent for each experiment, and keep other parameters constant, only changing the target parameter (e.g., fixing time and temperature when studying the effect of pressure). Repeat each set of process parameters 3-5 times and take the average value to reduce random errors. During the experiment, when a process parameter is varied in one unit, collect the corresponding sequences for each fitness parameter (juice yield, energy consumption, vitamin C retention rate) when that process parameter is varied in that unit. The sequence order should be the same as the order of changes when that process parameter is varied in that unit.

[0049] A quantitative experiment was conducted on the current batch of ginger and lemons to obtain the second weights of juice yield, energy consumption, and vitamin C retention rate based on the pressure, time, and temperature during juicing.

[0050] Specifically, pressure, time, and temperature are denoted as different types of process parameters, and juice yield, energy consumption, and vitamin C retention rate are denoted as different types of fitness parameters. One fitness parameter is selected as the target fitness parameter. A sequence of target fitness parameters is obtained when a process parameter changes by the minimum unit of variation in a quantitative experiment; this sequence is denoted as the target fitness parameter sequence corresponding to the minimum unit of variation of that process parameter. The average value of the first-order difference sequence of the target fitness parameter sequence corresponding to that process parameter is calculated and denoted as the average change value of the target fitness parameter corresponding to the minimum unit of variation of that process parameter. A sequence of different values ​​of that process parameter when it changes by the minimum unit of variation is obtained and denoted as the process parameter sequence. The process parameter sequence is used as the horizontal axis, and the target fitness parameter sequence corresponding to the minimum unit of variation of that process parameter is used as the vertical axis for linear fitting to obtain the fitted linear slope. The product of the absolute value of the fitted linear slope and the average change value of the target fitness parameter corresponding to the minimum unit of variation of that process parameter is mapped using an exponential function with a base of the natural constant to obtain the first mapping value of that process parameter for the target fitness parameter. Calculate the difference between the maximum and minimum values ​​of a process parameter when it changes by one unit of variation, and record this as the maximum difference of the process parameter corresponding to that unit of variation. Obtain the ratio of this unit of variation to the maximum difference of the process parameter corresponding to that unit of variation, and record this as the change ratio of the process parameter when it changes by that unit of variation. Divide the average change value of the target fitness parameter corresponding to the change of the process parameter by the difference between the maximum and minimum values ​​in the sequence of target fitness parameters corresponding to the change of the process parameter by that unit of variation to obtain the average change ratio of the target fitness parameter when it changes by that unit of variation. Calculate the difference between the average change ratio of the target fitness parameter when it changes by that unit of variation and the change ratio of the process parameter when it changes by that unit of variation, and record this as the change difference corresponding to the change of the process parameter when it changes by that unit of variation. Calculate the mean of the change differences corresponding to the change of the process parameter with respect to various units of variation, and record this as the average change difference of the process parameter with respect to the target fitness parameter. The influence characteristic value of the process parameter on the target fitness parameter is obtained by multiplying the first mapping value of the process parameter on the target fitness parameter with the average change difference of the process parameter on the target fitness parameter. The second weight of the target fitness parameter is obtained by obtaining the difference of the influence characteristic value of each process parameter on the target fitness parameter. Similarly, the second weights of the other two fitness parameters can be obtained.

[0051] The specific calculation model for the second weight of the target fitness parameter is as follows: , in, The second weight of the j-th fitness parameter is the second weight of the target fitness parameter when the j-th fitness parameter is used as the target fitness parameter. It represents the average influence of various process parameters on the j-th fitness parameter (juice yield / energy consumption / vitamin C retention rate). The average change value of the fitness parameter corresponding to the i-th process parameter when the i-th process parameter changes by the smallest unit of change; This represents the slope of the fitted linear curve when the i-th process parameter changes by the smallest unit of change. If the average change value of the j-th fitness parameter when the i-th process parameter changes by the smallest unit of change and the slope of the fitted linear curve when the i-th process parameter changes by the smallest unit of change are both large, it indicates that the j-th fitness parameter is significantly affected by the i-th process parameter changing by the smallest unit of change. Let represent the first mapping value of the i-th process parameter to the j-th fitness parameter, and exp represent an exponential function with the natural constant as the base. However, since the influence relationship may not be linear, meaning different parameters may have different effects, therefore, we introduce... and , This represents the percentage change of the i-th process parameter when it changes by the t-th unit of variation. This represents the average change ratio of the fitness parameter j when the i-th process parameter changes by the t-th unit of variation. If the fitness parameter j is significantly affected by the i-th process parameter changing by the t-th unit of variation, then... The value tends to be a larger positive value, and vice versa; Normalization needs to be performed using the maximum and minimum values; norm represents the normalization operation; T represents the number of types of variations in the i-th process parameter. This represents the average variation difference of the i-th process parameter with respect to the j-th fitness parameter; This represents the characteristic value of the influence of the i-th process parameter on the j-th fitness parameter. From this, the second weight of each fitness parameter can be obtained, which is also the second weight of juice yield, energy consumption, and vitamin C retention rate. Next, the final weights of juice yield, energy consumption, and vitamin C retention rate are obtained based on the first and second weights of juice yield, energy consumption, and vitamin C retention rate. Specifically, the final weight of the juice yield is obtained by multiplying the first and second weights of the juice yield; the final weight of the energy consumption is obtained by multiplying the first and second weights of the energy consumption; and the final weight of the vitamin C retention rate is obtained by multiplying the first and second weights of the vitamin C retention rate.

[0052] Finally, a fitness calculation formula is constructed using juice yield, energy consumption, vitamin C retention rate, and their corresponding final weights. The fitness calculation formula is as follows: , in, Indicates fitness; , , The final weights for juice yield, energy consumption, and vitamin C retention rate are respectively represented. , , These represent juice yield, energy consumption, and vitamin C retention rate, respectively. The energy consumption used in this calculation is the normalized energy consumption, which is obtained by dividing it by the total energy consumption.

[0053] Step S3: Construct a set of parent structures based on the initial number of parent structures; obtain the fitness of the structures according to the fitness calculation formula and in combination with experiments; use a genetic algorithm to iterate based on the set of parent structures and combine the fitness of each structure during the iteration process to obtain the optimal structure; use the pressure, time and temperature in the optimal structure as parameters for juicing ginger and lemon in the current batch.

[0054] The above steps complete the preparations for the genetic algorithm iteration. Next, a set of parent structures needs to be constructed based on the initial number of parent structures. Specifically, a set of parent structures is formed by randomly generating the initial number of parent structures. During the genetic algorithm iteration, the fitness of each structure needs to be calculated. This calculation can be done experimentally by obtaining the juice yield, energy consumption, and vitamin C retention rate when juicing using the pressure, time, and temperature specified in the structure. The experiments should use the current batch of raw materials, and all experimental conditions should remain consistent for each structure except for the parameters within that structure. After obtaining the juice yield, energy consumption, and vitamin C retention rate for each structure through experiments, these values ​​are then substituted into the fitness calculation formula to obtain the fitness of each structure.

[0055] When using a genetic algorithm to iterate based on the set of parent structures, the various process parameters in this application are range parameters. Therefore, the crossover method is single-point crossover, the mutation method is random replacement, and the mutation probability is set to 0.1. Random crossover and mutation occur, and offspring are generated. Each time the genetic algorithm is used, the fitness of each structure is calculated, and the top CN structures with the highest fitness in the parent and offspring generations are retained as new parents. Then, new offspring are generated, and this process is repeated.

[0056] Simultaneously, iterative stopping conditions need to be set. Specifically, when using the genetic algorithm to iterate using the parent structure set, the standard deviation of the fitness of each structure in a generation is obtained as the fitness standard deviation of that generation. If the fitness standard deviation of N consecutive generations is less than or equal to a preset threshold, then population reduction is performed, retaining the parent of the latest generation's offspring and the top CN1 structures with the highest fitness in the latest generation's offspring as new parents. If the fitness standard deviation of the new parents is less than or equal to the preset threshold, then iteration stops. If the fitness standard deviation of the new parents is greater than the preset threshold, then iteration continues with the new parents until the fitness standard deviation of N consecutive generations is less than or equal to the preset threshold, and the fitness standard deviation of the new parents generated after population reduction is less than or equal to the preset threshold, then iteration stops.

[0057] Preferably, the reference value of N in this application is 20, that is, if the fitness standard deviation of 20 consecutive generations is less than or equal to a preset threshold, then population reduction is performed. The preset threshold is 1%, and the value of CN1 is less than CN, which is 90% of CN. It should be noted that the value of N can be adjusted according to the actual situation. When performing population reduction, it is reduced according to a fixed ratio. The first population reduction is CN1=90%CN, the second population reduction is CN2=90%CN1, and so on, until the iteration stops. When the number of new parent generations is reduced according to a fixed ratio after population reduction, it is necessary to round to obtain the number of new parent generation structures.

[0058] Finally, the pressure, time, and temperature in the optimal structure are used as parameters for juicing the current batch of ginger and lemon juice, and then juicing is carried out.

[0059] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0060] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

[0061] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for intelligent control of the production process of ginger and lemon mixed juice, characterized in that, The method includes: The structure is constructed using the pressure, time, and temperature during juicing; the initial number of parent structures is obtained by using the range of various parameters in the structure and the maximum number of parent structures. The first weights for juice yield, energy consumption, and vitamin C retention rate were obtained by using the average juice yield, average energy consumption, and average vitamin C retention rate of historical batches of juice, respectively. The second weights for juice yield, energy consumption, and vitamin C retention rate were obtained by conducting quantitative experiments on the current batch of ginger and lemon juice based on the pressure, time, and temperature during juicing, including: Pressure, time, and temperature are denoted as different types of process parameters, and juice yield, energy consumption, and vitamin C retention rate are denoted as different types of fitness parameters. One fitness parameter is randomly selected as the target fitness parameter. A sequence of target fitness parameters is obtained when a process parameter changes by the minimum unit of change in a quantitative experiment; this sequence is denoted as the target fitness parameter sequence corresponding to the minimum unit of change of that process parameter. The average value of the first-order difference sequence of the target fitness parameter sequence corresponding to that process parameter is calculated and denoted as the average change value of the target fitness parameter corresponding to the minimum unit of change of that process parameter. A sequence of different values ​​of that process parameter when it changes by the minimum unit of change is obtained and denoted as the process parameter sequence. The process parameter sequence is used as the horizontal axis, and the target fitness parameter sequence corresponding to the minimum unit of change of that process parameter is used as the vertical axis for linear fitting to obtain the fitted linear slope. The product of the absolute value of the fitted linear slope and the average change value of the target fitness parameter corresponding to the minimum unit of change of that process parameter is mapped using an exponential function with a base of the natural constant to obtain the first mapping value of that process parameter for the target fitness parameter. Calculate the difference between the maximum and minimum values ​​of a process parameter when it changes by one unit of variation, and record this as the maximum difference of the process parameter corresponding to that unit of variation. Obtain the ratio of this unit of variation to the maximum difference of the process parameter corresponding to that unit of variation, and record this as the change ratio of the process parameter when it changes by that unit of variation. Divide the average change value of the target fitness parameter corresponding to the change of the process parameter by the difference between the maximum and minimum values ​​in the sequence of target fitness parameters corresponding to the change of the process parameter by that unit of variation to obtain the average change ratio of the target fitness parameter when it changes by that unit of variation. Calculate the difference between the average change ratio of the target fitness parameter when it changes by that unit of variation and the change ratio of the process parameter when it changes by that unit of variation, and record this as the change difference corresponding to the change of the process parameter when it changes by that unit of variation. Calculate the mean of the change differences corresponding to the change of the process parameter with respect to various units of variation, and record this as the average change difference of the process parameter with respect to the target fitness parameter. The influence characteristic value of the process parameter on the target fitness parameter is obtained by multiplying the first mapping value of the process parameter on the target fitness parameter with the average change difference of the process parameter on the target fitness parameter. The mean value of the influence characteristic value of each process parameter on the target fitness parameter is obtained to obtain the second weight of the target fitness parameter. Similarly, the second weights of the other two fitness parameters are obtained. The final weights of juice yield, energy consumption, and vitamin C retention rate are obtained based on the first and second weights. The fitness calculation formula is constructed using juice yield, energy consumption, vitamin C retention rate, and the corresponding final weights. A set of parent structures is constructed based on the initial number of parent structures; the fitness of the structures is obtained according to the fitness calculation formula and combined with experiments; the optimal structure is obtained by using a genetic algorithm to iterate based on the set of parent structures and combining the fitness of each structure during the iteration process; the pressure, time and temperature in the optimal structure are used as parameters for juicing ginger and lemon in the current batch.

2. The intelligent control method for the production process of ginger and lemon mixed juice according to claim 1, characterized in that, The method of obtaining the initial number of parent structures by utilizing the range of various parameters in the structure and the maximum number of parent structures includes: Based on the pressure range and the pressure adjustment unit, obtain the number of pressure values; based on the time range and the time adjustment unit, obtain the number of time values; based on the temperature range and the temperature adjustment unit, obtain the number of temperature values; the product of the number of pressure values, the number of time values, and the number of temperature values ​​is the maximum population size. The proportional coefficients corresponding to pressure are obtained by comparing the number of pressure values ​​used when juicing each batch of raw materials with the number of pressure values. Similarly, the proportional coefficients corresponding to time and temperature are obtained. The proportional coefficients corresponding to pressure, time and temperature are multiplied together, and the result of multiplying the result by the maximum population size and the number of parameter types is rounded to obtain the initial number of parent structures.

3. The intelligent control method for the production process of ginger and lemon mixed juice according to claim 1, characterized in that, The method of obtaining the first weights for juice yield, energy consumption, and vitamin C retention rate by using the average juice yield, average energy consumption, and average vitamin C retention rate of historical batches of juice includes: The average energy consumption ratio is obtained by comparing the average energy consumption of each batch of juice in history with the sum of the energy consumption of each batch of juice in history. The average energy consumption result is obtained by subtracting the average energy consumption ratio from the first preset value. The first weights of juice yield, vitamin C retention rate and energy consumption are obtained by dividing the average juice yield, the average vitamin C retention rate and the average energy consumption result of each batch of juice in history by the sum of the average juice yield, the average vitamin C retention rate and the average energy consumption result of each batch of juice in history.

4. The intelligent control method for the production process of ginger and lemon mixed juice according to claim 1, characterized in that, The final weights for juice yield, energy consumption, and vitamin C retention rate are obtained based on a first weight and a second weight, including: The final weight of juice yield is obtained by multiplying the first and second weights of juice yield; the final weight of energy consumption is obtained by multiplying the first and second weights of energy consumption; and the final weight of vitamin C retention rate is obtained by multiplying the first and second weights of vitamin C retention rate.

5. The intelligent control method for the production process of ginger and lemon mixed juice according to claim 1, characterized in that, The fitness calculation formula is as follows: ， in, Indicates fitness; , , The final weights for juice yield, energy consumption, and vitamin C retention rate are respectively represented. , , These represent juice yield, energy consumption, and vitamin C retention rate, respectively.

6. The intelligent control method for the production process of ginger and lemon mixed juice according to claim 1, characterized in that, The optimal structure is obtained by iterating through the parent structure set using a genetic algorithm and combining the fitness of each structure during the iteration process, including: Based on the set of parent structures, a genetic algorithm is used for iteration. When the iteration stops, the iteration stops, and the structure with the highest fitness among the offspring is taken as the optimal structure.

7. The intelligent control method for the production process of ginger and lemon mixed juice according to claim 6, characterized in that, The specific iteration stopping condition is as follows: When using a genetic algorithm to iterate using the set of parent structures, the standard deviation of the fitness of each structure in a generation is obtained as the fitness standard deviation of that generation. If the fitness standard deviation of N consecutive generations is less than or equal to a preset threshold, then population reduction is performed, and the parents of the latest generation and the top CN1 structures with the highest fitness in the latest generation are retained as new parents. If the fitness standard deviation of the new parents is less than or equal to the preset threshold, then the iteration stops.

Images