Ziyang oolong tea processing system and method based on group algorithm

Abstract

The invention discloses a group algorithm-based Ziyang oolong tea processing system and method, and the method comprises the following steps: S1, collecting fresh tea leaves according to a picking standard, and carrying out the pretreatment; s2, performing fine manipulation, fixation, rolling and drying operations in stages, and establishing a data set; s3, constructing a tea sensory quality database, binding a processing object and a picking standard, and establishing a mapping relationship between processing parameters and quality; s4, processing parameters are optimized by adopting a TPHCHA algorithm, and the main group and the auxiliary group cooperate with each other; s5, the processing technology is optimized in combination with an immune genetic algorithm, and fine tuning is carried out; s6, realizing relaxation and optimization of constraint conditions in different processing stages, and adjusting a fine adjustment scheme of tea flavor; s7, updating the tea sensory quality database in real time, and optimizing the processing technology; and S8, completing production of the Ziyang oolong tea, and performing quality detection. The quality stability and the production efficiency of the Ziyang oolong tea are improved, and the cost is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent control and production automation, and in particular to a Ziyang Oolong tea processing system and method based on a swarm algorithm. Background Art

[0002] As one of China's traditional famous teas, Ziyang Oolong tea enjoys widespread market demand for its unique taste, aroma, and meticulous craftsmanship. The traditional Ziyang Oolong tea processing process includes multiple steps, including picking, withering, greening, killing the green leaves, rolling, and drying. Each step directly impacts the quality of the final product. Due to the high technical requirements of Ziyang Oolong tea production, improving processing efficiency and minimizing the impact of human intervention while ensuring tea quality have long been key research areas in the industry.

[0003] Traditional Ziyang Oolong tea processing relies on the experience of tea growers and tea technicians, manually adjusting operational parameters at every stage. During each stage, such as the tea leaves being processed, withering, rolling, and drying, processing parameters such as temperature, humidity, time, and turning frequency must be fine-tuned according to the specific characteristics of the tea leaves. However, because tea quality is affected by numerous factors, including the environment, raw materials, equipment, and operator skills, traditional empirical processing methods often suffer from significant uncertainty, leading to inconsistent product quality. Therefore, achieving precise control over the processing of Ziyang Oolong tea has become a key technological advancement.

[0004] Existing Ziyang Oolong tea processing methods are often adjusted based on empirical data and experimental operations. However, this approach is limited by its heavy reliance on manual labor and accumulated experience, making large-scale, standardized production difficult. Although some research has attempted to introduce automated and intelligent equipment for tea processing, the application of these technologies is still limited to controlling a single processing step and lacks comprehensive optimization of the entire process. In particular, when dealing with complex process parameters across multiple processing stages, precisely controlling each stage to ensure high-quality final tea remains a challenge.

[0005] Meanwhile, while existing tea processing technology can collect some data through sensors, such as temperature, humidity, and stirring frequency, this data is often limited to a single processing step and lacks a comprehensive understanding of the interplay between these steps. Furthermore, existing technology often relies on manual fine-tuning of processing parameters, making it difficult to achieve consistency and standardization in large-scale production. Furthermore, due to the limitations of manual operation, product quality often fluctuates, making it impossible to guarantee the stability and consistency of each batch of tea.

[0006] Furthermore, current tea processing technology lacks the flexibility and precision to adapt to different batches of tea raw materials, particularly the unique characteristics of Ziyang Oolong tea. Because raw tea quality fluctuates with factors such as season and climate, automatically adjusting processing parameters to accommodate these varying raw material characteristics remains a challenge in the tea processing industry. Traditional methods typically rely on fixed processing procedures and manual experience to address these variations, making it difficult to achieve truly automated and intelligent adjustments.

[0007] Against this backdrop, swarm algorithms, as advanced intelligent optimization methods, have found widespread application in various fields, demonstrating significant potential in solving complex problems. Swarm algorithms, such as particle swarm optimization (PSO), ant colony optimization (ACO), and genetic algorithms (GA), can optimize complex multi-objective problems by simulating the behavior of swarms in nature. Swarm algorithms are capable of performing global optimization in complex, high-dimensional, multivariable problems, avoiding local optimal solutions. Therefore, they hold significant potential for application in the tea processing industry.

[0008] However, there are few existing applications of swarm optimization algorithms to comprehensively optimize Ziyang Oolong tea processing. While some studies have explored the use of machine learning and optimization algorithms to adjust certain parameters in tea processing, these approaches often focus on optimizing a single step and lack a systematic approach to optimizing the entire processing process. Furthermore, most existing technologies rely on traditional processing data and manual input, lacking dynamic data updates and feedback mechanisms based on sensory quality, and thus cannot adjust processing parameters in real time.

[0009] Therefore, how to provide a Ziyang Oolong tea processing system and method based on swarm algorithm is a problem that those skilled in the art urgently need to solve. Summary of the Invention

[0010] One objective of this invention is to develop a Ziyang Oolong tea processing system and method based on a swarm algorithm. This invention leverages swarm algorithms and intelligent optimization techniques to describe in detail a swarm algorithm-based method for optimizing Ziyang Oolong tea processing parameters. This method includes the collection of picking criteria, optimization and adjustment of processing stages, the construction of a sensory quality database, and the synergistic effects of the TPHCEA algorithm and the immune genetic algorithm, achieving intelligent control and optimization of the Ziyang Oolong tea processing process.

[0011] The Ziyang Oolong tea processing method based on a swarm algorithm according to an embodiment of the present invention comprises the following steps:

[0012] S1. Collect fresh tea leaves according to the picking standards and complete pretreatment within four hours after picking;

[0013] S2. Based on the pre-treated fresh tea leaves, perform the greening, fixing, rolling, and drying operations in stages, set the processing objects and processing parameters, and establish a processing data set;

[0014] S3. Based on the processing data set, a tea sensory quality database is constructed, and processing objects are labeled and bound to picking standards. At the same time, a mapping relationship between processing parameters and tea sensory quality is established;

[0015] S4, the TPHCEA algorithm is used to optimize and adjust the processing parameters. The main group and the two auxiliary groups cooperate and dynamically choose between strong cooperation and weak cooperation;

[0016] S5. Combine the immune genetic algorithm to genetically optimize the processing process and simulate the selection and memory mechanism of the immune system to fine-tune the processing technology of different batches of tea;

[0017] S6. Through the synergistic effect of the main and auxiliary populations in the TPHCEA algorithm, the constraints at different processing stages are relaxed and optimized, and the fine-tuning scheme of tea flavor is adjusted;

[0018] S7. During the tea processing process, the tea sensory quality database is updated in real time, and the processing technology is optimized by combining the TPHCEA algorithm and the immune genetic algorithm;

[0019] S8. Based on the optimized processing technology, complete the production of Ziyang Oolong tea and conduct quality inspection to ensure that the sensory quality and physical and chemical indicators meet the preset standards.

[0020] Optionally, the picking standard refers to mainly picking tea leaves with two to four leaves and opposite leaves, and is divided into three levels according to the quality of fresh tea leaves: special grade, first grade, and second grade. The special grade means that the buds with two or three leaves account for more than 95%, the first grade means that the buds with two or three leaves account for more than 80%, and the second grade means that the buds with two or three leaves account for more than 60%.

[0021] Optionally, the pretreatment includes removing non-tea inclusions, controlling the moisture content of fresh tea leaves to below 70%, setting a constant temperature and humidity environment for withering treatment, and recording the temperature, humidity and treatment time.

[0022] Optionally, the processing objects include strips, color, wholeness, clarity, soup color, aroma, taste, and leaf bottom.

[0023] Optionally, the processing parameters include temperature, duration, tumbling frequency, kneading intensity and drying method.

[0024] Optionally, the S3 specifically includes:

[0025] S31. Based on the processing data set, a label-based binding relationship between the processing object and the picking criteria is set. The label-based binding includes the following steps:

[0026] The picking standards are divided into special grade, first grade and second grade, and the picking standards of each batch of tea are matched with various indicators of sensory quality;

[0027] Set the sensory quality standards and corresponding numerical ranges for each batch of tea processing objects, the processing objects including strips, color, wholeness, clarity, soup color, aroma, taste, and leaf base, the strip standard is X1∈[a1,b1], the color standard is X2∈[a2,b2], the wholeness standard is X3∈[a3,b3], the clarity standard is X4∈[a4,b4], the soup color standard is X5∈[a5,b5], the aroma standard is X6∈[a6,b6], the taste standard is X7∈[a7,b7], and the leaf base standard is X8∈[a8,b8], among which a1, b1, a2, b2, a3, b3, a4, b4, a5, b5, a6, b6, a7, b7, a8, b8 are constants, and these numerical ranges are adjusted according to the type and grade of tea;

[0028] S32. Construct a tea sensory quality database by collecting the processing objects of each batch of tea at each processing stage, recording the processing parameters and corresponding sensory quality values ​​of each batch. The structure of the tea sensory quality database is as follows:

[0029] Each record includes picking criteria, processing parameters, sensory quality score, and actual processing time;

[0030] The scoring value of each sensory quality is X1, X2, X3, X4, X5, X6, X7, X8;

[0031] The values ​​of each processing parameter include temperature, duration, tumbling frequency, kneading intensity and drying method;

[0032] The picking standard labels are "special grade", "first grade", and "second grade", which are bound to the processing objects;

[0033] S33. Based on the tea sensory quality database, a mapping relationship between processing parameters and tea sensory quality is constructed, and an influence function of the processing operation parameters at each stage is set. The mapping relationship includes:

[0034] In the greening stage, the processing parameters are set as temperature T1, duration t1, and flipping frequency f1;

[0035] In the fixing stage, the processing parameters are set as temperature T2, duration t2, and turning frequency f2;

[0036] In the rolling stage, the processing parameters are set as rolling intensity I and duration t3;

[0037] In the drying stage, the processing parameters are set as drying mode M, temperature T3, and duration t4;

[0038] The influence function includes:

[0039] f1(T1,t1,f1)=c1·T1+c2·t1+c3·f1;

[0040] f2(T2,t2,f2)=c4·T2+c5·t2+c6·f2;

[0041] f3(I,t3)=c7·I+c8·t3;

[0042] f4(M,T3,t4)=c9·M+c 10 T3+c 11 t4;

[0043] Among them, f1(T1, t1, f1) is the sensory quality score of the greening stage, f2(T2, t2, f2) is the sensory quality score of the fixing stage, f3(I, t3) is the sensory quality score of the rolling stage, f4(M, T3, t4) is the sensory quality score of the drying stage, c1, c2, c3, c4, c5, c6, c7, c8, c9, c 10 ,c 11 is a constant;

[0044] S34, based on the mapping relationship between processing parameters and sensory quality of tea, the relationship between the processing parameters at each stage and the sensory quality of tea is quantitatively expressed, so that ∑(Q i -Q ideal ) 2 Minimum, ensuring the best organoleptic quality at every stage:

[0045] Q=f1(T1,t1,f1)·W1+f2(T2,t2,f2)·W2+f3(I,t3)·W3+f4(M,T3,t4)·W4;

[0046] Among them, Q represents the sensory quality score of all stages of the entire processing process, W1, W2, W3, and W4 are weighted coefficients, W1 represents the weighted coefficient of the greening stage, W2 represents the weighted coefficient of the fixing stage, W3 represents the weighted coefficient of the rolling stage, and W4 represents the weighted coefficient of the drying stage. Q i is the sensory quality score of the i-th processing stage, Q ideal It is the ideal sensory quality target value for the ideal processing stage.

[0047] Optionally, the S4 specifically includes:

[0048] S41, use TPHCEA algorithm to optimize and adjust the processing parameters, and initialize the individuals of the main group to P1, P2, ..., P n , where n is the number of individuals in the main group, P i is the i-th individual in the main group, each of which consists of a set of processing parameters, including temperature T1, T2, T3, duration t1, t2, t3, t4, tumbling frequency f1, f2, kneading intensity I, and drying method M;

[0049] S42. Initialize the individuals of two auxiliary groups as A1, A2, ..., A m with B1, B2, …, B k , where A i and B i are individuals in two auxiliary groups, containing the same set of processing parameters, and m and k are the number of individuals in the auxiliary groups respectively;

[0050] S43, the main group and the auxiliary group are collaboratively calculated, and the main group individuals are scored according to the current processing parameters and sensory quality scores Q1, Q2, Q3, ..., Q n Perform fitness evaluation and perform dynamic selection based on the standards of strong cooperation and weak cooperation to determine the strength of each group. The dynamic selection refers to the selection of each main group individual P i , using the fitness function in the TPHCEA algorithm to select strong cooperation or weak cooperation, the selection result is determined by the weighted coefficient W strong and W weak The ratio is determined by:

[0051] S i =W strong Fit(P i )+W weak Fit(A i )+W weak Fit(B i );

[0052] Among them, S i is the selection result of the i-th individual in the main group, Fit(·) is the fitness function, and Fit(P i ),Fit(A i ),Fit(B i ) are fitness evaluations corresponding to the main group and the two auxiliary groups, W strong and W weak is the weight coefficient of strong cooperation and weak cooperation;

[0053] S44. Through the synergistic effect of the main group and the auxiliary group in the TPHCEA algorithm, the weights of the processing parameters in each stage are adjusted to generate a new processing parameter configuration to ensure the optimization of temperature T1, T2, T3, duration t1, t2, t3, t4, tumbling frequency f1, f2, kneading intensity I and drying method M, so that the sensory quality score of each stage reaches the best.

[0054] Optionally, the S5 specifically includes:

[0055] S51, combine the immune genetic algorithm to perform genetic optimization on the processing process, simulate the selection and memory mechanism of the immune system, and initialize the population to G1, G2, ..., G p , where G i represents the i-th individual in the population, each of which contains a set of processing parameters, including temperature T1, T2, T3, duration t1, t2, t3, t4, tumbling frequency f1, f2, kneading intensity I, and drying method M;

[0056] S52. According to the selection and memory mechanism in the immune genetic algorithm, individuals with higher fitness are selected for genetic operation. The operation process is as follows:

[0057] Select individuals with high fitness as memory cells, and set the memory cell set to M = {G1, G2, ..., G p}, where p is the number of individuals selected;

[0058] Performing a crossover operation on each of the two to generate the next generation of individuals, wherein the crossover operation generates new individuals by exchanging part of the processing parameters;

[0059] Performing mutation operations on the next generation of individuals, wherein the mutation operation optimizes processing parameters through slight changes to ensure that the generated new individuals meet higher fitness;

[0060] S53. Through the selection, crossover, and mutation operations of the immune genetic algorithm, the processing technology of different batches of tea is fine-tuned, and multiple rounds of optimization iterations are carried out.

[0061] Optionally, the S6 specifically includes:

[0062] S61, through the synergy of the main group and the auxiliary group in the TPHCEA algorithm, the processing parameters of each stage are constrained and defined, and the processing parameter constraints of each processing stage are set;

[0063] S62. During the collaborative optimization process of the TPHCEA algorithm, the parameter constraints of the processing stage are gradually relaxed through the dynamic selection mechanism of the main group and the auxiliary group, allowing the processing parameters to be adjusted within a reasonable range, so that the processing parameters of each stage can meet the best sensory quality output as much as possible;

[0064] S63. By combining the genetic operation of the immune genetic algorithm and the dynamic selection mechanism of the TPHCEA algorithm, an optimized processing parameter combination is generated, multiple rounds of optimization iterations are performed, and the fine-tuning plan of the tea flavor is adjusted.

[0065] The Ziyang Oolong tea processing system based on a swarm algorithm according to an embodiment of the present invention includes:

[0066] The picking and processing module is used to collect fresh tea leaves according to picking standards and complete pre-processing within four hours after picking;

[0067] The operation setting module is used to perform the greening, fixing, rolling and drying operations in stages based on the pre-processed fresh tea leaves, and set the processing objects and processing parameters to establish the processing data set;

[0068] The data construction module is used to build a tea sensory quality database based on the processing data set, label and bind the processing objects with the picking standards, and establish a mapping relationship between processing parameters and tea sensory quality;

[0069] Collaborative optimization module, which uses the TPHCEA algorithm to optimize and adjust the processing parameters. The main group and two auxiliary groups work together to dynamically choose between strong and weak cooperation.

[0070] Genetic optimization module, which is used to genetically optimize the processing process by combining the immune genetic algorithm and simulating the selection and memory mechanism of the immune system to fine-tune the processing technology for different batches of tea;

[0071] The constraint relaxation module is used to relax and optimize the constraints at different processing stages through the synergy between the main and auxiliary populations in the TPHCEA algorithm, and to adjust the fine-tuning scheme of tea flavor;

[0072] The update optimization module is used to update the tea sensory quality database in real time during the tea processing process and optimize the processing technology by combining the TPHCEA algorithm and the immune genetic algorithm;

[0073] The quality inspection module is used to complete the production of Ziyang Oolong tea based on the optimized processing technology and conduct quality inspection to ensure that the sensory quality and physical and chemical indicators meet the preset standards.

[0074] The beneficial effects of the present invention are:

[0075] First, the present invention achieves intelligent control of the entire Ziyang Oolong tea processing process by introducing a swarm algorithm-based Ziyang Oolong tea processing method. By leveraging the synergistic effect of the TPHCEA algorithm and the immune genetic algorithm, parameters at each processing step can be dynamically adjusted based on real-time data and sensory quality feedback, effectively addressing the traditional processing techniques' reliance on manual experience and lack of precise control. This intelligent adjustment not only improves processing accuracy but also ensures flavor stability and consistency across each batch of tea.

[0076] Secondly, the optimization method of the present invention can automatically adjust parameters at each processing stage based on the varying raw materials and environmental variations of Ziyang Oolong tea, ensuring that each batch of tea achieves the desired sensory quality. This flexible adjustment method avoids the limitations of traditional processing techniques that are unable to adapt to raw material differences and external environmental changes, further improving the overall quality of the tea and processing efficiency.

[0077] Finally, this invention effectively improves the automation and standardization of Ziyang Oolong tea production. By applying a swarm algorithm, the need for human intervention is significantly reduced while ensuring high-quality tea production and improving production efficiency. The application of this technology will help promote the intelligent and automated development of Ziyang Oolong tea production, further optimizing the production process, reducing production costs, and enhancing market competitiveness. BRIEF DESCRIPTION OF THE DRAWINGS

[0078] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0079] Figure 1 This is a flow chart of the Ziyang Oolong tea processing method based on swarm algorithm proposed in the present invention;

[0080] Figure 2 This is a schematic diagram of the relationship between the construction of the tea sensory quality database and the mapping of processing parameters for the Ziyang Oolong tea processing method based on the swarm algorithm proposed in the present invention;

[0081] Figure 3 This is a module structure diagram of the Ziyang Oolong tea processing system based on swarm algorithm proposed in the present invention. DETAILED DESCRIPTION

[0082] The present invention will now be described in further detail with reference to the accompanying drawings, which are simplified schematic diagrams that illustrate the basic structure of the present invention in a schematic manner.

[0083] refer to Figure 1-2 The Ziyang Oolong tea processing method based on swarm algorithm comprises the following steps:

[0084] S1. Collect fresh tea leaves according to the picking standards and complete pretreatment within four hours after picking;

[0085] S2. Based on the pre-treated fresh tea leaves, perform the greening, fixing, rolling, and drying operations in stages, set the processing objects and processing parameters, and establish a processing data set;

[0086] S3. Based on the processing data set, a tea sensory quality database is constructed, and processing objects are labeled and bound to picking standards. At the same time, a mapping relationship between processing parameters and tea sensory quality is established;

[0087] S4, the TPHCEA algorithm is used to optimize and adjust the processing parameters. The main group and the two auxiliary groups cooperate and dynamically choose between strong cooperation and weak cooperation;

[0088] S5. Combine the immune genetic algorithm to genetically optimize the processing process and simulate the selection and memory mechanism of the immune system to fine-tune the processing technology of different batches of tea;

[0089] S6. Through the synergistic effect of the main and auxiliary populations in the TPHCEA algorithm, the constraints at different processing stages are relaxed and optimized, and the fine-tuning scheme of tea flavor is adjusted;

[0090] S7. During the tea processing process, the tea sensory quality database is updated in real time, and the processing technology is optimized by combining the TPHCEA algorithm and the immune genetic algorithm;

[0091] S8. Based on the optimized processing technology, complete the production of Ziyang Oolong tea and conduct quality inspection to ensure that the sensory quality and physical and chemical indicators meet the preset standards.

[0092] This invention introduces a swarm-based algorithm-based Ziyang oolong tea processing method, addressing the traditional reliance on experience and manual labor. Through the collaborative optimization of the TPHCEA algorithm and the immune genetic algorithm, precise control of processing parameters is achieved, ensuring that parameters at each processing stage are intelligently adjusted based on real-time feedback from the tea leaves. This significantly improves the processing precision, production efficiency, and consistency of product quality of Ziyang oolong tea.

[0093] In this embodiment, the picking standard refers to the buds with two to four leaves and the opposite leaves, and is divided into three levels according to the quality of the fresh tea leaves: special grade, first grade, and second grade. The special grade refers to the buds with two or three leaves accounting for more than 95%, the first grade refers to the buds with two or three leaves accounting for more than 80%, and the second grade refers to the buds with two or three leaves accounting for more than 60%.

[0094] By subdividing the picking standards for Ziyang Oolong tea into three grades: Special Grade, Grade One, and Grade Two, we effectively ensure the uniformity and quality of each batch of tea. This standardized picking requirement ensures more precise raw material quality for subsequent processing, further enhancing the sensory quality of the tea.

[0095] In this embodiment, the pretreatment includes removing non-tea inclusions, controlling the moisture content of fresh tea leaves to below 70%, setting a constant temperature and humidity environment for withering treatment, and recording the temperature, humidity and treatment time.

[0096] During the pretreatment of fresh tea leaves, this method removes non-tea inclusions and strictly controls the moisture content of the fresh leaves. This ensures the stability and consistency of the tea leaves during processing, avoids quality fluctuations caused by factors such as uneven moisture, and effectively improves the controllability of subsequent processing steps and the sensory quality of the tea leaves.

[0097] In this embodiment, the processing objects include strips, color, wholeness, clarity, soup color, aroma, taste, and leaf bottom.

[0098] By clearly defining the tea processing objects, including strips, color, wholeness, purity, soup color, aroma, taste and leaf bottom, and combining specific processing parameters for precise control, it is possible to comprehensively monitor and optimize the various sensory indicators of tea in each processing link, thereby improving the overall quality of tea.

[0099] In this embodiment, the processing parameters include temperature, duration, tumbling frequency, kneading intensity and drying method.

[0100] This method, by detailing processing parameters such as temperature, duration, rolling frequency, rolling intensity, and drying method, combined with swarm optimization, ensures that processing parameters at each stage can be dynamically adjusted based on specific circumstances. This method effectively reduces the imprecise parameter control issues associated with traditional processing methods, thereby ensuring high-quality tea output.

[0101] In this embodiment, S3 specifically includes:

[0102] S31. Based on the processing data set, a label-based binding relationship between the processing object and the picking criteria is set. The label-based binding includes the following steps:

[0103] The picking standards are divided into special grade, first grade and second grade, and the picking standards of each batch of tea are matched with various indicators of sensory quality;

[0104] Set the sensory quality standards and corresponding numerical ranges for each batch of tea processing objects, the processing objects including strips, color, wholeness, clarity, soup color, aroma, taste, and leaf base, the strip standard is X1∈[a1,b1], the color standard is X2∈[a2,b2], the wholeness standard is X3∈[a3,b3], the clarity standard is X4∈[a4,b4], the soup color standard is X5∈[a5,b5], the aroma standard is X6∈[a6,b6], the taste standard is X7∈[a7,b7], and the leaf base standard is X8∈[a8,b8], among which a1, b1, a2, b2, a3, b3, a4, b4, a5, b5, a6, b6, a7, b7, a8, b8 are constants, and these numerical ranges are adjusted according to the type and grade of tea;

[0105] S32. Construct a tea sensory quality database by collecting the processing objects of each batch of tea at each processing stage, recording the processing parameters and corresponding sensory quality values ​​of each batch. The structure of the tea sensory quality database is as follows:

[0106] Each record includes picking criteria, processing parameters, sensory quality score, and actual processing time;

[0107] The scoring value of each sensory quality is X1, X2, X3, X4, X5, X6, X7, X8;

[0108] The values ​​of each processing parameter include temperature, duration, tumbling frequency, kneading intensity and drying method;

[0109] The picking standard labels are "special grade", "first grade", and "second grade", which are bound to the processing objects;

[0110] S33. Based on the tea sensory quality database, a mapping relationship between processing parameters and tea sensory quality is constructed, and an influence function of the processing operation parameters at each stage is set. The mapping relationship includes:

[0111] In the greening stage, the processing parameters are set as temperature T1, duration t1, and flipping frequency f1;

[0112] In the fixing stage, the processing parameters are set as temperature T2, duration t2, and turning frequency f2;

[0113] In the rolling stage, the processing parameters are set as rolling intensity I and duration t3;

[0114] In the drying stage, the processing parameters are set as drying mode M, temperature T3, and duration t4;

[0115] The influence function includes:

[0116] f1(T1,t1,f1)=c1·T1+c2·t1+c3·f1;

[0117] f2(T2,t2,f2)=c4·T2+c5·t2+c6·f2;

[0118] f3(I,t3)=c7·I+c8·t3;

[0119] f4(M,T3,t4)=c9·M+c 10 T3+c 11 t4;

[0120] Among them, f1(T1, t1, f1) is the sensory quality score of the greening stage, f2(T2, t2, f2) is the sensory quality score of the fixing stage, f3(I, t3) is the sensory quality score of the rolling stage, f4(M, T3, t4) is the sensory quality score of the drying stage, c1, c2, c3, c4, c5, c6, c7, c8, c9, c 10 ,c 11 is a constant;

[0121] S34, based on the mapping relationship between processing parameters and sensory quality of tea, the relationship between the processing parameters at each stage and the sensory quality of tea is quantitatively expressed, so that ∑(Q i -Q ideal ) 2 Minimum, ensuring the best organoleptic quality at every stage:

[0122] Q=f1(T1,t1,f1)·W1+f2(T2,t2,f2)·W2+f3(I,t3)·W3+f4(M,T3,t4)·W4;

[0123] Among them, Q represents the sensory quality score of all stages of the entire processing process, W1, W2, W3, and W4 are weighted coefficients, W1 represents the weighted coefficient of the greening stage, W2 represents the weighted coefficient of the fixing stage, W3 represents the weighted coefficient of the rolling stage, and W4 represents the weighted coefficient of the drying stage. Q i is the sensory quality score of the i-th processing stage, Q ideal It is the ideal sensory quality target value for the ideal processing stage.

[0124] This invention builds a tea sensory quality database, combining picking standards with labeling and processing data to achieve comprehensive traceability and quantitative management of tea quality. This database not only provides data support for optimizing the processing process but also allows for real-time updates based on each batch's processing parameters, ensuring continuous improvement and quality stability.

[0125] In this embodiment, the S4 specifically includes:

[0126] S41, use TPHCEA algorithm to optimize and adjust the processing parameters, and initialize the individuals of the main group to P1, P2, ..., P n , where n is the number of individuals in the main group, P i is the i-th individual in the main group, each of which consists of a set of processing parameters, including temperature T1, T2, T3, duration t1, t2, t3, t4, tumbling frequency f1, f2, kneading intensity I, and drying method M;

[0127] S42. Initialize the individuals of two auxiliary groups as A1, A2, ..., A m with B1, B2, …, B k , where A i and B i are individuals in two auxiliary groups, containing the same set of processing parameters, and m and k are the number of individuals in the auxiliary groups respectively;

[0128] S43, the main group and the auxiliary group are collaboratively calculated, and the main group individuals are scored according to the current processing parameters and sensory quality scores Q1, Q2, Q3, ..., Q n Perform fitness evaluation and perform dynamic selection based on the standards of strong cooperation and weak cooperation to determine the strength of each group. The dynamic selection refers to the selection of each main group individual P i , using the fitness function in the TPHCEA algorithm to select strong cooperation or weak cooperation, the selection result is determined by the weighted coefficient W strong and W weak The ratio is determined by:

[0129] S i =W strong Fit(P i )+W weak Fit(A i )+W weak Fit(B i );

[0130] Among them, S i is the selection result of the i-th individual in the main group, Fit(·) is the fitness function, and Fit(P i ),Fit(A i ),Fit(B i ) are fitness evaluations corresponding to the main group and the two auxiliary groups, W strong and W weak is the weight coefficient of strong cooperation and weak cooperation;

[0131] S44. Through the synergistic effect of the main group and the auxiliary group in the TPHCEA algorithm, the weights of the processing parameters in each stage are adjusted to generate a new processing parameter configuration to ensure the optimization of temperature T1, T2, T3, duration t1, t2, t3, t4, tumbling frequency f1, f2, kneading intensity I and drying method M, so that the sensory quality score of each stage reaches the best.

[0132] The TPHCEA algorithm was used to optimize and adjust processing parameters. Through the collaborative efforts of the primary and auxiliary groups, the parameters at each stage were effectively optimized. This algorithm achieves a dynamic balance between strong and weak collaboration, avoiding the uncertainty inherent in traditional manual adjustments and further enhancing the intelligent level and efficiency of tea processing.

[0133] In this embodiment, the S5 specifically includes:

[0134] S51, combine the immune genetic algorithm to perform genetic optimization on the processing process, simulate the selection and memory mechanism of the immune system, and initialize the population to G1, G2, ..., G p , where G i represents the i-th individual in the population, each of which contains a set of processing parameters, including temperature T1, T2, T3, duration t1, t2, t3, t4, tumbling frequency f1, f2, kneading intensity I, and drying method M;

[0135] S52. According to the selection and memory mechanism in the immune genetic algorithm, individuals with higher fitness are selected for genetic operation. The operation process is as follows:

[0136] Select individuals with high fitness as memory cells, and set the memory cell set to M = {G1, G2, ..., G p}, where p is the number of individuals selected;

[0137] Performing a crossover operation on each of the two to generate the next generation of individuals, wherein the crossover operation generates new individuals by exchanging part of the processing parameters;

[0138] Performing mutation operations on the next generation of individuals, wherein the mutation operation optimizes processing parameters through slight changes to ensure that the generated new individuals meet higher fitness;

[0139] S53. Through the selection, crossover, and mutation operations of the immune genetic algorithm, the processing technology of different batches of tea is fine-tuned, and multiple rounds of optimization iterations are carried out.

[0140] By combining genetic optimization with an immune genetic algorithm, this invention mimics the immune system's selection and memory mechanisms. This allows for intelligent adaptation to the varying characteristics of raw materials and optimization of the tea processing process when fine-tuning different batches. This optimization approach enhances the process's adaptability and improves the sensory quality and stability of each batch of tea.

[0141] In this embodiment, S6 specifically includes:

[0142] S61, through the synergy of the main group and the auxiliary group in the TPHCEA algorithm, the processing parameters of each stage are constrained and defined, and the processing parameter constraints of each processing stage are set;

[0143] S62. During the collaborative optimization process of the TPHCEA algorithm, the parameter constraints of the processing stage are gradually relaxed through the dynamic selection mechanism of the main group and the auxiliary group, allowing the processing parameters to be adjusted within a reasonable range, so that the processing parameters of each stage can meet the best sensory quality output as much as possible;

[0144] S63. By combining the genetic operation of the immune genetic algorithm and the dynamic selection mechanism of the TPHCEA algorithm, an optimized processing parameter combination is generated, multiple rounds of optimization iterations are performed, and the fine-tuning plan of the tea flavor is adjusted.

[0145] By combining the TPHCEA algorithm with an immune genetic algorithm, this method relaxes and optimizes processing constraints, enabling flexible adjustment of processing parameters based on the actual needs of each stage. This method improves the adaptability and flexibility of each processing stage, ensuring fine-tuning of tea flavor and consistent final quality.

[0146] refer to Figure 3 , Ziyang Oolong tea processing system based on swarm algorithm, including:

[0147] The picking and processing module is used to collect fresh tea leaves according to picking standards and complete pre-processing within four hours after picking;

[0148] The operation setting module is used to perform the greening, fixing, rolling and drying operations in stages based on the pre-processed fresh tea leaves, and set the processing objects and processing parameters to establish the processing data set;

[0149] The data construction module is used to build a tea sensory quality database based on the processing data set, label and bind the processing objects with the picking standards, and establish a mapping relationship between processing parameters and tea sensory quality;

[0150] Collaborative optimization module, which uses the TPHCEA algorithm to optimize and adjust the processing parameters. The main group and two auxiliary groups work together to dynamically choose between strong and weak cooperation.

[0151] Genetic optimization module, which is used to genetically optimize the processing process by combining the immune genetic algorithm and simulating the selection and memory mechanism of the immune system to fine-tune the processing technology for different batches of tea;

[0152] The constraint relaxation module is used to relax and optimize the constraints at different processing stages through the synergy between the main and auxiliary populations in the TPHCEA algorithm, and to adjust the fine-tuning scheme of tea flavor;

[0153] The update optimization module is used to update the tea sensory quality database in real time during the tea processing process and optimize the processing technology by combining the TPHCEA algorithm and the immune genetic algorithm;

[0154] The quality inspection module is used to complete the production of Ziyang Oolong tea based on the optimized processing technology and conduct quality inspection to ensure that the sensory quality and physical and chemical indicators meet the preset standards.

[0155] The swarm algorithm-based Ziyang Oolong tea processing system provided by this invention integrates multiple processes, including picking, processing, optimization, updating, and quality inspection, through a modular design, forming a comprehensive intelligent control system. This system optimizes the processing process in real time, ensuring efficient and stable production of Ziyang Oolong tea. It also implements automation and standardization throughout the production process, reducing manual intervention and improving production efficiency and tea quality consistency.

[0156] Example 1:

[0157] To verify the feasibility of the present invention, the invention was applied to a Ziyang Oolong tea production line in a certain region. This region boasts unique natural conditions and abundant tea resources. However, traditional Ziyang Oolong tea processing relies on manual fine-tuning based on experience, resulting in difficulty maintaining consistent product quality, low processing efficiency, and frequent fluctuations in tea flavor and unstable quality.

[0158] In the actual application of this production line, we implemented a swarm algorithm-based Ziyang Oolong tea processing method. First, the picking standard is strictly defined as primarily leaves with two to four leaves in the bud, as well as leaves in the middle. Fresh tea leaves are categorized into three grades based on their quality: special grade, first grade, and second grade. Under this standard, picking is performed by experienced tea farmers to ensure the consistency and high quality of the raw materials. After picking, the tea leaves undergo pre-processing within four hours. All tea leaves undergo a rigorous removal of non-tea impurities. The stability of each batch is guaranteed by withering, which maintains a moisture content below 70% and maintains a constant temperature and humidity environment.

[0159] Subsequently, according to the method described in this invention, the pretreated fresh tea leaves were sequentially processed through various stages, including greening, withering, rolling, and drying. Each stage was defined with specific processing targets and parameters. Using a sensory data acquisition system, we recorded parameters such as temperature, humidity, rolling frequency, rolling intensity, and drying method in real time during each processing stage. This data was used to establish a processing dataset, providing the foundation for a subsequent tea sensory quality database.

[0160] To further improve tea quality, we used the TPHCEA algorithm to optimize processing parameters. During the "greening" stage, temperature and duration were adjusted through collaboration between the main and two auxiliary groups, ultimately achieving the optimal combination of processing parameters. The TPHCEA algorithm also played a role during the "fixing" stage. Through a dynamic selection mechanism of strong and weak cooperation, each batch of tea was processed under the optimal conditions, maximizing the tea's aroma and flavor.

[0161] Next, during the rolling and drying stages, the process was further optimized using an immune genetic algorithm. Using the genetic optimization algorithm's selection and memory mechanism, the algorithm selects individuals with high fitness for genetic manipulation, generating new individuals. Optimization was achieved through crossover and mutation operations. Ultimately, the optimized process underwent multiple rounds of iterative fine-tuning, ensuring maximum flavor consistency across each batch of tea and improving product stability.

[0162] Throughout the entire production process, the tea sensory quality database is updated in real time, and the processing parameters and sensory quality scores of each batch of tea are recorded. By comparing with traditional processing methods, we have achieved significant improvements in time efficiency and quality stability. During the experiment, the production time of Ziyang Oolong tea was shortened by approximately 20% compared to traditional methods, and the production cycle from picking to finished product was shortened from 48 hours to 38 hours. In addition, the fluctuation range of sensory quality scores was significantly reduced, and indicators such as tea aroma, flavor, and tea soup color all showed high consistency. The standard deviation of sensory quality scores was reduced from 1.5 with traditional methods to 0.5.

[0163] The following table shows the quality comparison data of Ziyang Oolong tea in different batches produced under the processing method of the present invention and the traditional method.

[0164] Table 1 Comparison of processing quality of Ziyang Oolong tea

[0165]

[0166] Table 1 shows that the production cycle of Ziyang Oolong tea using the method of the present invention was significantly shortened, and sensory quality scores were significantly improved compared to traditional methods. Furthermore, flavor consistency was significantly improved, demonstrating that optimizing the processing technology through swarm algorithms can effectively ensure consistent quality across batches of tea.

[0167] In general, the implementation of the present invention greatly improves the production process of Ziyang Oolong tea, which not only improves production efficiency and reduces manual intervention, but also ensures the sensory quality and flavor stability of the tea, achieving the expected technical effect.

[0168] The above description is only a preferred specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with the technical field, within the technical scope disclosed by the present invention, who makes equivalent replacements or changes based on the technical solution and inventive concept of the present invention, should be covered by the scope of protection of the present invention.

Claims

1. A Ziyang Oolong tea processing method based on a swarm algorithm, characterized in that: The steps include: S1. Collect fresh tea leaves according to the picking standards and complete pretreatment within four hours after picking; S2. Based on the pre-treated fresh tea leaves, perform the greening, fixing, rolling, and drying operations in stages, set the processing objects and processing parameters, and establish a processing data set; S3. Based on the processing data set, a tea sensory quality database is constructed, and processing objects are labeled and bound to picking standards. At the same time, a mapping relationship between processing parameters and tea sensory quality is established; S4, the TPHCEA algorithm is used to optimize and adjust the processing parameters. The main group and the two auxiliary groups cooperate and dynamically choose between strong cooperation and weak cooperation; S5. Combine the immune genetic algorithm to genetically optimize the processing process and simulate the selection and memory mechanism of the immune system to fine-tune the processing technology of different batches of tea; S6. Through the synergistic effect of the main and auxiliary populations in the TPHCEA algorithm, the constraints at different processing stages are relaxed and optimized, and the fine-tuning scheme of tea flavor is adjusted; S7. During the tea processing process, the tea sensory quality database is updated in real time, and the processing technology is optimized by combining the TPHCEA algorithm and the immune genetic algorithm; S8. Based on the optimized processing technology, complete the production of Ziyang Oolong tea and conduct quality inspection to ensure that the sensory quality and physical and chemical indicators meet the preset standards.

2. The Ziyang Oolong tea processing method based on swarm algorithm according to claim 1, characterized in that: The picking standard refers to the buds with two to four leaves and the opposite leaves. According to the quality of the fresh tea leaves, they are divided into three levels: special grade, first grade and second grade. The special grade means that the buds with two or three leaves account for more than 95%, the first grade means that the buds with two or three leaves account for more than 80%, and the second grade means that the buds with two or three leaves account for more than 60%.

3. The Ziyang Oolong tea processing method based on swarm algorithm according to claim 1, characterized in that: The pretreatment includes removing non-tea inclusions, controlling the moisture content of fresh tea leaves to below 70%, setting a constant temperature and humidity environment for withering treatment, and recording the temperature, humidity and treatment time.

4. The Ziyang Oolong tea processing method based on swarm algorithm according to claim 1, characterized in that: The processing objects include strips, color, wholeness, clarity, soup color, aroma, taste, and leaf bottom.

5. The Ziyang Oolong tea processing method based on swarm algorithm according to claim 1, characterized in that: The processing parameters include temperature, duration, tumbling frequency, kneading intensity and drying method.

6. The Ziyang Oolong tea processing method based on swarm algorithm according to claim 1, characterized in that: The S3 specifically includes: S31. Based on the processing data set, a label-based binding relationship between the processing object and the picking criteria is set. The label-based binding includes the following steps: The picking standards are divided into special grade, first grade and second grade, and the picking standards of each batch of tea are matched with various indicators of sensory quality; Set the sensory quality standards and corresponding numerical ranges for each batch of tea processing objects, the processing objects including strips, color, wholeness, clarity, soup color, aroma, taste, and leaf base, the strip standard is X1∈[a1,b1], the color standard is X2∈[a2,b2], the wholeness standard is X3∈[a3,b3], the clarity standard is X4∈[a4,b4], the soup color standard is X5∈[a5,b5], the aroma standard is X6∈[a6,b6], the taste standard is X7∈[a7,b7], and the leaf base standard is X8∈[a8,b8], among which a1, b1, a2, b2, a3, b3, a4, b4, a5, b5, a6, b6, a7, b7, a8, b8 are constants, and these numerical ranges are adjusted according to the type and grade of tea; S32. Construct a tea sensory quality database by collecting the processing objects of each batch of tea at each processing stage, recording the processing parameters and corresponding sensory quality values ​​of each batch. The structure of the tea sensory quality database is as follows: Each record includes picking criteria, processing parameters, sensory quality score, and actual processing time; The scoring value of each sensory quality is X1, X2, X3, X4, X5, X6, X7, X8; The values ​​of each processing parameter include temperature, duration, tumbling frequency, kneading intensity and drying method; The picking standard labels are "Special Grade", "First Grade", and "Second Grade", which are bound to the processing objects; S33. Based on the tea sensory quality database, a mapping relationship between processing parameters and tea sensory quality is constructed, and an influence function of the processing operation parameters at each stage is set. The mapping relationship includes: In the greening stage, the processing parameters are set as temperature T1, duration t1, and flipping frequency f1; In the fixing stage, the processing parameters are set as temperature T2, duration t2, and turning frequency f2; In the rolling stage, the processing parameters are set as rolling intensity I and duration t3; In the drying stage, the processing parameters are set as drying mode M, temperature T3, and duration t4; The influence function includes: f1(T1,t1,f1)=c1·T1+c2·t1+c3·f1; f2(T2,t2,f2)=c4·T2+c5·t2+c6·f2; f3(I,t3)=c7·I+c8·t3; f4(M,T3,t4)=c9·M+c 10 ·T3+c 11 ·t4; Among them, f1(T1, t1, f1) is the sensory quality score of the greening stage, f2(T2, t2, f2) is the sensory quality score of the fixing stage, f3(I, t3) is the sensory quality score of the rolling stage, f4(M, T3, t4) is the sensory quality score of the drying stage, c1, c2, c3, c4, c5, c6, c7, c8, c9, c 10 ,c 11 is a constant; S34, based on the mapping relationship between processing parameters and sensory quality of tea, the relationship between the processing parameters at each stage and the sensory quality of tea is quantitatively expressed, so that ∑(Q i -Q ideal ) 2 Minimum, ensuring the best organoleptic quality at every stage: Q=f1(T1,t1,f1)·W1+f2(T2,t2,f2)·W2+f3(I,t3)·W3+f4(M,T3,t4)·W4; Among them, Q represents the sensory quality score of all stages of the entire processing process, W1, W2, W3, and W4 are weighted coefficients, W1 represents the weighted coefficient of the greening stage, W2 represents the weighted coefficient of the fixing stage, W3 represents the weighted coefficient of the rolling stage, and W4 represents the weighted coefficient of the drying stage. Q i is the sensory quality score of the i-th processing stage, Q ideal It is the ideal sensory quality target value for the ideal processing stage.

7. The Ziyang Oolong tea processing method based on swarm algorithm according to claim 1, characterized in that: The S4 specifically includes: S41, use TPHCEA algorithm to optimize and adjust the processing parameters, and initialize the individuals of the main group to P1, P2, ..., P n , where n is the number of individuals in the main group, P i is the i-th individual in the main group, each of which consists of a set of processing parameters, including temperature T1, T2, T3, duration t1, t2, t3, t4, tumbling frequency f1, f2, kneading intensity I, and drying method M; S42. Initialize the individuals of two auxiliary groups as A1, A2, ..., A m with B1, B2, …, B k , where A i and B i are individuals in two auxiliary groups, containing the same set of processing parameters, and m and k are the number of individuals in the auxiliary groups respectively; S43, the main group and the auxiliary group are collaboratively calculated, and the main group individuals are scored according to the current processing parameters and sensory quality scores Q1, Q2, Q3, ..., Q n Perform fitness evaluation and perform dynamic selection based on the standards of strong cooperation and weak cooperation to determine the strength of each group. The dynamic selection refers to the selection of each main group individual P i , using the fitness function in the TPHCEA algorithm to select strong cooperation or weak cooperation, the selection result is determined by the weighted coefficient W strong and W weak The ratio is determined by: S i =W strong ·Fit(P i )+W weak ·Fit(A i )+W weak ·Fit(B i ); Among them, S i is the selection result of the i-th individual in the main group, Fit(·) is the fitness function, and Fit(P i ),Fit(A i ),Fit(B i ) are fitness evaluations corresponding to the main group and the two auxiliary groups, W strong and W weak is the weight coefficient of strong cooperation and weak cooperation; S44. Through the synergistic effect of the main group and the auxiliary group in the TPHCEA algorithm, the weights of the processing parameters in each stage are adjusted to generate a new processing parameter configuration to ensure the optimization of temperature T1, T2, T3, duration t1, t2, t3, t4, tumbling frequency f1, f2, kneading intensity I and drying method M, so that the sensory quality score of each stage reaches the best.

8. The Ziyang Oolong tea processing method based on swarm algorithm according to claim 1, characterized in that: The S5 specifically includes: S51, combine the immune genetic algorithm to perform genetic optimization on the processing process, simulate the selection and memory mechanism of the immune system, and initialize the population to G1, G2, ..., G p , where G i represents the i-th individual in the population, each of which contains a set of processing parameters, including temperature T1, T2, T3, duration t1, t2, t3, t4, tumbling frequency f1, f2, kneading intensity I, and drying method M; S52. According to the selection and memory mechanism in the immune genetic algorithm, individuals with higher fitness are selected for genetic operation. The operation process is as follows: Select individuals with high fitness as memory cells, and set the memory cell set to M = {G1, G2, ..., G p }, where p is the number of individuals selected; Performing a crossover operation on each of the two to generate the next generation of individuals, wherein the crossover operation generates new individuals by exchanging part of the processing parameters; Performing mutation operations on the next generation of individuals, wherein the mutation operation optimizes processing parameters through slight changes to ensure that the generated new individuals meet higher fitness; S53. Through the selection, crossover, and mutation operations of the immune genetic algorithm, the processing technology of different batches of tea is fine-tuned, and multiple rounds of optimization iterations are carried out.

9. The Ziyang Oolong tea processing method based on swarm algorithm according to claim 1, characterized in that: The S6 specifically includes: S61, through the synergy of the main group and the auxiliary group in the TPHCEA algorithm, the processing parameters of each stage are constrained and defined, and the processing parameter constraints of each processing stage are set; S62. During the collaborative optimization process of the TPHCEA algorithm, the parameter constraints of the processing stage are gradually relaxed through the dynamic selection mechanism of the main group and the auxiliary group, allowing the processing parameters to be adjusted within a reasonable range, so that the processing parameters of each stage can meet the best sensory quality output as much as possible; S63. By combining the genetic operation of the immune genetic algorithm and the dynamic selection mechanism of the TPHCEA algorithm, an optimized processing parameter combination is generated, multiple rounds of optimization iterations are performed, and the fine-tuning plan of the tea flavor is adjusted.

10. A Ziyang Oolong tea processing system based on a swarm algorithm, which executes the Ziyang Oolong tea processing method based on a swarm algorithm according to any one of claims 1 to 9, characterized in that: include: The picking and processing module is used to collect fresh tea leaves according to picking standards and complete pre-processing within four hours after picking; The operation setting module is used to perform the greening, fixing, rolling and drying operations in stages based on the pre-processed fresh tea leaves, and set the processing objects and processing parameters to establish the processing data set; The data construction module is used to build a tea sensory quality database based on the processing data set, label and bind the processing objects with the picking standards, and establish a mapping relationship between processing parameters and tea sensory quality; Collaborative optimization module, which uses the TPHCEA algorithm to optimize and adjust the processing parameters. The main group and two auxiliary groups work together to dynamically choose between strong and weak cooperation. Genetic optimization module, which is used to genetically optimize the processing process by combining the immune genetic algorithm and simulating the selection and memory mechanism of the immune system to fine-tune the processing technology for different batches of tea; The constraint relaxation module is used to relax and optimize the constraints at different processing stages through the synergy between the main and auxiliary populations in the TPHCEA algorithm, and to adjust the fine-tuning scheme of tea flavor; The update optimization module is used to update the tea sensory quality database in real time during the tea processing process and optimize the processing technology by combining the TPHCEA algorithm and the immune genetic algorithm; The quality inspection module is used to complete the production of Ziyang Oolong tea based on the optimized processing technology and conduct quality inspection to ensure that the sensory quality and physical and chemical indicators meet the preset standards.

Images