A method for dynamically adjusting tea fermentation environment using fuzzy logic control

Abstract

The present invention discloses a method for dynamically adjusting the environment of tea fermentation using fuzzy logic control, comprising the following steps: S1, real-time collection and acquisition of environmental parameter data of each layer; S2, fuzzy processing of temperature, humidity and oxygen concentration in each time period by a fuzzy logic controller, generating fuzzy control outputs of each time period; S3, combining a hybrid neural network architecture with meta-learning technology to construct an environmental parameter change model; S4, optimizing the fuzzy rule base using the environmental parameter change model; S5, converting the fuzzy control outputs of each time period into control instructions; S6, optimizing the fermentation environment through a cross-layer parameter balancing strategy, and obtaining the optimal state of environmental parameters of the fermentation process; S7, real-time display of environmental parameters of each spatial level and time stage, and supporting manual and automatic mode switching. The present invention combines fuzzy logic control with deep learning technology to achieve accurate dynamic adjustment of the tea fermentation environment, and has the advantages of strong adaptability.

Description

Technical Field

[0001] The present invention relates to the field of fuzzy logic control and deep learning technology, and in particular to a method for dynamically adjusting a tea fermentation environment using fuzzy logic control. Background Art

[0002] During the tea fermentation process, environmental parameters such as temperature, humidity and oxygen concentration have a crucial impact on the quality of the final tea. Traditional tea fermentation control methods usually rely on manual experience and timed manual adjustments, which makes it difficult to achieve real-time and accurate control of the fermentation environment, resulting in unstable tea quality and consistency differences between batches. In addition, although the existing automated control systems can replace manual operations to a certain extent, most systems still show problems of lag in regulation and insufficient control accuracy when facing complex changes in the fermentation environment. These problems are mainly reflected in the following aspects.

[0003] First, existing automated control systems are often unable to effectively handle the differences in spatial levels and time stages during tea fermentation. In the tea fermentation environment, there are usually significant differences in temperature, humidity, and oxygen concentration between different layers (such as the upper layer, middle layer, and lower layer), and this difference has an important impact on the quality of tea fermentation. However, existing control methods mostly adopt a global unified control strategy, lacking the ability to independently adjust environmental parameters at each level, resulting in the inability to accurately control the specific conditions of different levels. In addition, the tea fermentation process usually goes through three time stages: the initial, middle, and final stages. Each stage has different requirements for temperature, humidity, and oxygen concentration, and the existing system is difficult to achieve refined control of these time stages, which often leads to the inadaptability of environmental parameters during the fermentation process, thereby affecting the final quality of the tea.

[0004] Secondly, in terms of data processing and decision optimization, existing fermentation control systems usually rely on preset rules or simple feedback control, which makes it difficult to adapt to dynamic changes in the environment. Especially when dealing with complex historical data and real-time monitoring data, traditional rule bases and control algorithms often lack sufficient flexibility and learning capabilities, making it difficult to cope with the changing fermentation environment. Even if some systems have introduced advanced technologies such as fuzzy logic control, they still cannot maintain efficient control accuracy in a rapidly changing environment due to the fixed nature of their rule bases and the limitations of update frequency. In addition, most existing systems are unable to effectively use historical data and real-time data for self-learning and optimization, resulting in unsatisfactory control effects when faced with nonlinear changes in environmental parameters.

[0005] Thirdly, the existing fermentation environment control system has obvious deficiencies in cross-layer parameter adjustment. In the fermentation environment, environmental parameters at different levels often show complex interactions, and parameter adjustments at a single level often affect environmental conditions at other levels. Therefore, the lack of a coordinated regulation mechanism for cross-layer parameters will lead to limitations in the control effect. For example, when the temperature of the upper layer is higher and the temperature of the lower layer is lower, if the system cannot achieve balanced temperature regulation between the layers, some tea leaves may be under-fermented or over-fermented, thus affecting the overall fermentation quality. In existing systems, cross-layer parameter adjustments are usually passive and lagging, lacking real-time coordinated regulation capabilities.

[0006] Finally, the inadequacy of existing technologies in terms of intelligence and adaptability is also a major reason that affects the quality of fermentation. Traditional tea fermentation control systems mostly use preset control logic. Once the external environment or fermentation conditions change, the system cannot adjust the control strategy in time, which is prone to control errors. In addition, the existing system relies on manual intervention in the optimization of control strategies and lacks an intelligent automatic adjustment mechanism, which leads to poor adaptability and untimely adjustment of the system in complex and changing environments. Especially when dealing with emergencies during the fermentation process, the existing system is often difficult to respond quickly and take effective measures, which affects the consistency and quality stability of tea fermentation.

[0007] Therefore, how to provide a method for dynamically adjusting the tea fermentation environment using fuzzy logic control is an urgent problem that technicians in this field need to solve. Summary of the invention

[0008] One purpose of the present invention is to propose a method for dynamically adjusting the tea fermentation environment using fuzzy logic control. The present invention fully combines fuzzy logic control with deep learning technology, and optimizes the temperature, humidity and oxygen concentration regulation during tea fermentation through dynamic division of spatial levels and time stages. The system ensures the refined control and coordination of the fermentation environment through real-time data analysis and cross-layer parameter balancing strategy. Its intelligent adaptive mechanism can automatically adjust the control strategy and respond quickly to environmental changes, thereby significantly improving the consistency of fermentation quality and the overall quality of the product. It has the significant advantages of high intelligence, strong adaptability and stable fermentation quality.

[0009] A method for dynamically adjusting a tea fermentation environment using fuzzy logic control according to an embodiment of the present invention comprises the following steps:

[0010] S1. Divide the tea fermentation environment into the first layer, the second layer and the third layer according to the spatial dimension, and arrange temperature, humidity and oxygen concentration sensors on each layer to collect and obtain the environmental parameter data of each layer in real time;

[0011] S2, dividing the fermentation process into a first time period, a second time period and a third time period according to the time stage, and based on the environmental parameter data of each layer and combined with the fermentation time, fuzzy processing is performed on the temperature, humidity and oxygen concentration in each time period through a fuzzy logic controller to generate a fuzzy control output for each time period;

[0012] S3, using the fuzzy control output of each time period and historical fermentation data, combined with the hybrid neural network architecture and meta-learning technology to analyze the changes in environmental parameters at the spatial level and time stage, and build an environmental parameter change model;

[0013] S4, using the environmental parameter change model to optimize the fuzzy rule base;

[0014] S5. Based on the optimized fuzzy rule base, the fuzzy control output of each time period is converted into control instructions to dynamically adjust the temperature, humidity and oxygen concentration in the fermentation environment at each spatial level and each time stage;

[0015] S6, real-time detection of environmental parameter differences between different levels, optimization of the fermentation environment through cross-layer parameter balancing strategy, and obtaining the optimal state of environmental parameters during the fermentation process;

[0016] S7, real-time display of environmental parameters such as temperature, humidity and oxygen concentration at each spatial level and time stage, and support switching between manual and automatic modes.

[0017] Optionally, the S2 specifically includes:

[0018] S21, dividing the fermentation process into a first time period, a second time period and a third time period according to the time stage, and performing fuzzy processing on the environmental parameter data of each layer obtained in step S1 in each time period;

[0019] S22, in the first time period, the real-time temperature, humidity and oxygen concentration of each layer are input into the fuzzy logic controller, and the real-time temperature, humidity and oxygen concentration are fuzzified through the fuzzy membership function to generate a fuzzy set ;

[0020] S23, in the second time period, the real-time temperature, humidity and oxygen concentration of each layer are input into the fuzzy logic controller, and the real-time temperature, humidity and oxygen concentration are fuzzified through the fuzzy membership function to generate a fuzzy set ;

[0021] S24, in the third time period, the real-time temperature, humidity and oxygen concentration of each layer are input into the fuzzy logic controller, and the real-time temperature, humidity and oxygen concentration are fuzzified by the fuzzy membership function to generate a fuzzy set ;

[0022] S25. Through fuzzy logic reasoning, use the inference rules:

[0023] :if and and ,but ;

[0024] in, Indicates Fuzzy rules, Indicates The serial number of the fuzzy rules, , and represents the input fuzzy set, represents the output fuzzy set, represents the temperature fuzzy set in the input fuzzy set, represents the humidity fuzzy set in the input fuzzy set, represents the oxygen concentration fuzzy set in the input fuzzy set, represents the result set of fuzzy logic reasoning, Indicates temperature, Indicates humidity, Indicates oxygen concentration, Represents the final fuzzy control output:

[0025] ;

[0026] in, represents the membership function value of the input temperature, represents the membership function value of the input humidity, represents the membership function value of the input oxygen concentration, Represents the total number of fuzzy rules.

[0027] Optionally, the S3 specifically includes:

[0028] S31. Collect fuzzy control output data of each time period during the fermentation process , , ,in represents the fuzzy control output of temperature, represents the fuzzy control output of humidity, The fuzzy control output representing the oxygen concentration, Indicates time period;

[0029] S32, integrating the fuzzy control output data of each time period with the historical fermentation data, the historical fermentation data including the historical temperature data , Historical humidity data and historical oxygen concentration data , and the actual quality results of tea fermentation ;

[0030] S33, using a hybrid neural network architecture to process the integrated data, the hybrid neural network architecture includes a convolutional neural network, a recurrent neural network and an adaptive weight adjustment mechanism, and the extraction of spatial features is done through a nonlinear function conduct:

[0031] ;

[0032] in, Represents spatial features, represents a convolutional neural network;

[0033] S34, extracting spatial features Input into the recurrent neural network and extract time series features through weighted recurrent calculation:

[0034] ;

[0035] in, Represents the time series characteristics, represents the weight of adaptive learning, represents the number of layers of the recurrent neural network, represents a recurrent neural network;

[0036] S35. Adaptive learning weights in hybrid neural network architectures via meta-learning algorithms and nonlinear functions The parameters are optimized, and the optimization goal is to minimize the nonlinear loss function :

[0037] ;

[0038] in, represents the adaptive adjustment coefficient, represents the predicted tea quality result, Represents historical quality results, represents the number of samples, A serial number indicating the number of samples;

[0039] S36. Generate an environmental parameter change model through the processed and optimized model. The environmental parameter change model includes prediction curves of temperature, humidity and oxygen concentration in each time period. , and , and control recommendations.

[0040] Optionally, the S4 specifically includes:

[0041] S41, based on the environmental parameter change model generated in S36, the temperature, humidity and oxygen concentration prediction curves for each time period are calculated. , and As input, a preliminary analysis of the existing fuzzy rule base is conducted, including the shape and parameters of the membership functions, as well as the rationality of the rule weight distribution;

[0042] S42, optimizing the membership function, adjusting the parameters of the membership function using the environmental parameter change model, and obtaining an optimized temperature membership function , humidity membership function and oxygen concentration membership function ;

[0043] S43, based on optimizing the membership function, using the error between historical data and predicted data Adjust the parameters of the membership functions:

[0044] ;

[0045] in, represents the predicted temperature, represents the predicted humidity, represents the predicted oxygen concentration, , and represents the initial membership function value, represents the number of samples, A serial number indicating the number of samples;

[0046] S44, using the time series features extracted in step S34 Dynamic optimization of fuzzy rules, including readjusting the weight of each fuzzy rule according to the characteristics of the time series;

[0047] S45, verifying the optimized fuzzy rule base, and testing the optimized rule base using real environment parameters and historical data;

[0048] S46. Apply the optimized and verified fuzzy rule base to the system to replace the original fuzzy rule base.

[0049] Optionally, the S5 specifically includes:

[0050] S51, based on the optimized and verified fuzzy rule base, the fuzzy control output generated in each time period is defuzzified to generate control instructions , and ,in Used to regulate the temperature in the fermentation environment. Used to regulate the humidity in the fermentation environment. Used to adjust the oxygen concentration in the fermentation environment. Indicates time period;

[0051] S52, the generated control instruction , and Input to the control system, the regulation process includes controlling the output power or flow of heaters, humidifiers and ventilation systems;

[0052] S53, real-time monitoring of the actual temperature in the fermentation environment , Real-time humidity and actual oxygen concentration , the actual temperature , Real-time humidity and actual oxygen concentration and temperature target parameters , Humidity target parameters and oxygen concentration target parameters Compare and introduce adaptive error adjustment coefficient Calculate dynamic error :

[0053] ;

[0054] in, Represents the adaptive error adjustment coefficient, which is dynamically adjusted based on the fermentation stage;

[0055] S54, based on the calculated dynamic error , the increment of the control command is calculated by the nonlinear adjustment model , and :

[0056] ;

[0057] ;

[0058] ;

[0059] in, represents the nonlinear adjustment coefficient of temperature, represents the nonlinear adjustment coefficient of humidity, Indicates the nonlinear adjustment coefficient of oxygen concentration;

[0060] S55, updating the control instruction after the incremental adjustment to , and And re-input it into the control system to optimize the environmental parameter adjustment process;

[0061] S56. Continuously monitor and record the temperature, humidity and oxygen concentration at each level during the fermentation process, conduct real-time data analysis and feedback, and dynamically adjust the control strategy.

[0062] Optionally, the S6 specifically includes:

[0063] S61, real-time detection of temperature, humidity and oxygen concentration parameters of the first level, the second level and the third level in the fermentation environment, and inputting them into the central control system for analysis;

[0064] S62. Based on the real-time collected data, calculate the parameter differences between the layers and define the cross-layer temperature difference , Humidity difference and oxygen concentration difference ;

[0065] S63. Based on the cross-layer parameter differences, a cross-layer parameter balancing strategy is applied to dynamically adjust the environmental parameters:

[0066] ;

[0067] in, Indicates the intensity of temperature regulation. represents the initial temperature adjustment coefficient, Indicates the temperature regulation threshold, represents the temperature nonlinear amplification coefficient, represents the temperature intensity adjustment coefficient, represents the logarithmic temperature adjustment coefficient, represents the critical point of the piecewise function of temperature regulation intensity;

[0068] ;

[0069] in, Indicates the intensity of humidity regulation. represents the initial humidity adjustment coefficient, Indicates the humidity adjustment threshold, represents the humidity nonlinear amplification factor, represents the humidity intensity adjustment coefficient, represents the logarithmic adjustment coefficient of humidity, represents the critical point of the piecewise function of humidity regulation intensity;

[0070] ;

[0071] in, Indicates the intensity of regulation of oxygen concentration. Indicates the initial adjustment coefficient of oxygen concentration, Indicates the oxygen concentration adjustment threshold, represents the nonlinear amplification factor of oxygen concentration, Indicates the oxygen concentration intensity adjustment coefficient, represents the logarithmic adjustment coefficient of oxygen concentration, Represents the critical point of the piecewise function of oxygen concentration regulation intensity;

[0072] S64, the calculated adjustment intensity , and Input to the control system to adjust the output power of the heater, humidifier and ventilation system at each level;

[0073] S65, provide real-time feedback on the changes in parameters at all levels, update the model parameters in the central control system, and obtain the optimal state of the environmental parameters of the fermentation process.

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

[0075] First, the present invention realizes independent and refined control of each level and stage by dividing the fermentation environment into different levels according to the spatial dimension and combining it with the dynamic division of the time stage. This not only effectively copes with the differences in environmental parameters between levels during the fermentation process, but also flexibly adjusts the temperature, humidity and oxygen concentration according to the specific stage of fermentation, ensuring that each level of tea is fermented under suitable conditions, thereby improving the consistency and overall quality of the fermented product.

[0076] Secondly, thanks to the introduction of hybrid neural network architecture and meta-learning technology, the system of the present invention can use a large amount of historical fermentation data and real-time monitoring data to build an environmental parameter change model and dynamically optimize the fuzzy rule base. Compared with the traditional fixed rule base, the system of the present invention can continuously learn and adapt to environmental changes and automatically adjust the control strategy, which enables the system to maintain a high control accuracy when facing a complex and changeable fermentation environment. Through in-depth analysis of historical data and dynamic feedback of real-time data, the system can predict and optimize fermentation environmental parameters in advance, effectively reducing the negative impact of environmental fluctuations on the fermentation process and ensuring the stability of tea fermentation quality.

[0077] In addition, the cross-layer parameter balancing strategy in the present invention realizes the coordinated regulation of environmental parameters at different levels through nonlinear adjustment functions. Specifically, when there are significant differences in temperature, humidity or oxygen concentration between the layers, the system can automatically activate the cross-layer adjustment mechanism and dynamically adjust the control equipment of each layer, such as heaters, humidifiers and ventilation systems, to ensure that the environmental parameters of each layer tend to be balanced. This innovative adjustment mechanism avoids the limitations of single-level control commonly seen in traditional systems, and can better cope with the complex interactions in the fermentation environment, thereby improving the efficiency and balance of the entire fermentation process, and ultimately improving the quality of the fermented products.

[0078] Finally, the present invention greatly improves the intelligence and adaptive capabilities of the system. By combining fuzzy logic control with deep learning models, the system can not only operate efficiently under normal fermentation conditions, but also quickly respond and adjust the control strategy when abnormal or sudden environmental changes occur. This intelligent adaptive mechanism reduces the necessity of manual intervention, improves the automation of the system, and makes the tea fermentation process more stable and reliable. BRIEF DESCRIPTION OF THE DRAWINGS

[0079] 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:

[0080] Figure 1 This is a flow chart of a method for dynamically adjusting tea fermentation environment using fuzzy logic control proposed by the present invention;

[0081] Figure 2 This is a schematic diagram of spatial hierarchical division of a method for dynamically adjusting tea fermentation environment using fuzzy logic control proposed by the present invention;

[0082] Figure 3 A real-time dynamic adjustment schematic diagram of a method for dynamically adjusting a tea fermentation environment using fuzzy logic control proposed by the present invention. DETAILED DESCRIPTION

[0083] The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams, which only illustrate the basic structure of the present invention in a schematic manner, and therefore only show the components related to the present invention.

[0084] refer to Figure 1-3 A method for dynamically adjusting the tea fermentation environment using fuzzy logic control comprises the following steps:

[0085] S1. Divide the tea fermentation environment into the first layer, the second layer and the third layer according to the spatial dimension, and arrange temperature, humidity and oxygen concentration sensors on each layer to collect and obtain the environmental parameter data of each layer in real time;

[0086] S2, dividing the fermentation process into a first time period, a second time period and a third time period according to the time stage, and based on the environmental parameter data of each layer and combined with the fermentation time, fuzzy processing is performed on the temperature, humidity and oxygen concentration in each time period through a fuzzy logic controller to generate a fuzzy control output for each time period;

[0087] S3, using the fuzzy control output of each time period and historical fermentation data, combined with the hybrid neural network architecture and meta-learning technology to analyze the changes in environmental parameters at the spatial level and time stage, and build an environmental parameter change model;

[0088] S4, using the environmental parameter change model to optimize the fuzzy rule base;

[0089] S5. Based on the optimized fuzzy rule base, the fuzzy control output of each time period is converted into control instructions to dynamically adjust the temperature, humidity and oxygen concentration in the fermentation environment at each spatial level and each time stage;

[0090] S6, real-time detection of environmental parameter differences between different levels, optimization of the fermentation environment through cross-layer parameter balancing strategy, and obtaining the optimal state of environmental parameters during the fermentation process;

[0091] S7, real-time display of environmental parameters such as temperature, humidity and oxygen concentration at each spatial level and time stage, and support switching between manual and automatic modes.

[0092] In this implementation, S2 specifically includes:

[0093] S21, dividing the fermentation process into a first time period, a second time period and a third time period according to the time stage, and performing fuzzy processing on the environmental parameter data of each layer obtained in step S1 in each time period;

[0094] S22, in the first time period, the real-time temperature, humidity and oxygen concentration of each layer are input into the fuzzy logic controller, and the real-time temperature, humidity and oxygen concentration are fuzzified through the fuzzy membership function to generate a fuzzy set ;

[0095] S23, in the second time period, the real-time temperature, humidity and oxygen concentration of each layer are input into the fuzzy logic controller, and the real-time temperature, humidity and oxygen concentration are fuzzified through the fuzzy membership function to generate a fuzzy set ;

[0096] S24, in the third time period, the real-time temperature, humidity and oxygen concentration of each layer are input into the fuzzy logic controller, and the real-time temperature, humidity and oxygen concentration are fuzzified by the fuzzy membership function to generate a fuzzy set ;

[0097] S25. Through fuzzy logic reasoning, use the inference rules:

[0098] :if and and ,but ;

[0099] in, Indicates Fuzzy rules, Indicates The serial number of the fuzzy rules, , and represents the input fuzzy set, represents the output fuzzy set, represents the temperature fuzzy set in the input fuzzy set, represents the humidity fuzzy set in the input fuzzy set, represents the oxygen concentration fuzzy set in the input fuzzy set, represents the result set of fuzzy logic reasoning, Indicates temperature, Indicates humidity, Indicates oxygen concentration, Represents the final fuzzy control output:

[0100] ;

[0101] in, represents the membership function value of the input temperature, represents the membership function value of the input humidity, represents the membership function value of the input oxygen concentration, Represents the total number of fuzzy rules.

[0102] In this implementation, S3 specifically includes:

[0103] S31. Collect fuzzy control output data of each time period during the fermentation process , , ,in represents the fuzzy control output of temperature, represents the fuzzy control output of humidity, The fuzzy control output representing the oxygen concentration, Indicates time period;

[0104] S32, integrating the fuzzy control output data of each time period with the historical fermentation data, the historical fermentation data including the historical temperature data , Historical humidity data and historical oxygen concentration data , and the actual quality results of tea fermentation ;

[0105] S33, using a hybrid neural network architecture to process the integrated data, the hybrid neural network architecture includes a convolutional neural network, a recurrent neural network and an adaptive weight adjustment mechanism, and the extraction of spatial features is done through a nonlinear function conduct:

[0106] ;

[0107] in, Represents spatial features, represents a convolutional neural network;

[0108] S34, extracting spatial features Input into the recurrent neural network and extract time series features through weighted recurrent calculation:

[0109] ;

[0110] in, Represents the time series characteristics, represents the weight of adaptive learning, represents the number of layers of the recurrent neural network, represents a recurrent neural network;

[0111] S35. Adaptive learning weights in hybrid neural network architectures via meta-learning algorithms and nonlinear functions The parameters are optimized, and the optimization goal is to minimize the nonlinear loss function :

[0112] ;

[0113] in, represents the adaptive adjustment coefficient, represents the predicted tea quality result, Represents historical quality results, represents the number of samples, A serial number indicating the number of samples;

[0114] S36. Generate an environmental parameter change model through the processed and optimized model. The environmental parameter change model includes prediction curves of temperature, humidity and oxygen concentration in each time period. , and , and control recommendations.

[0115] In this implementation manner, the S4 specifically includes:

[0116] S41, based on the environmental parameter change model generated in S36, the temperature, humidity and oxygen concentration prediction curves for each time period are calculated. , and As input, a preliminary analysis of the existing fuzzy rule base is conducted, including the shape and parameters of the membership functions, as well as the rationality of the rule weight distribution;

[0117] S42, optimizing the membership function, adjusting the parameters of the membership function using the environmental parameter change model, and obtaining an optimized temperature membership function , humidity membership function and oxygen concentration membership function ;

[0118] S43, based on optimizing the membership function, using the error between historical data and predicted data Adjust the parameters of the membership functions:

[0119] ;

[0120] in, represents the predicted temperature, represents the predicted humidity, represents the predicted oxygen concentration, , and represents the initial membership function value, represents the number of samples, A serial number indicating the number of samples;

[0121] S44, using the time series features extracted in step S34 Dynamic optimization of fuzzy rules, including readjusting the weight of each fuzzy rule according to the characteristics of the time series;

[0122] S45, verifying the optimized fuzzy rule base, and testing the optimized rule base using real environment parameters and historical data;

[0123] S46. Apply the optimized and verified fuzzy rule base to the system to replace the original fuzzy rule base.

[0124] In this implementation manner, S5 specifically includes:

[0125] S51, based on the optimized and verified fuzzy rule base, the fuzzy control output generated in each time period is defuzzified to generate control instructions , and ,in Used to regulate the temperature in the fermentation environment. Used to regulate the humidity in the fermentation environment. Used to adjust the oxygen concentration in the fermentation environment. Indicates time period;

[0126] S52, the generated control instruction , and Input to the control system, the regulation process includes controlling the output power or flow of heaters, humidifiers and ventilation systems;

[0127] S53, real-time monitoring of the actual temperature in the fermentation environment , Real-time humidity and actual oxygen concentration , the actual temperature , Real-time humidity and actual oxygen concentration and temperature target parameters , Humidity target parameters and oxygen concentration target parameters Compare and introduce adaptive error adjustment coefficient Calculate dynamic error :

[0128] ;

[0129] in, Represents the adaptive error adjustment coefficient, which is dynamically adjusted based on the fermentation stage;

[0130] S54, based on the calculated dynamic error , the increment of the control command is calculated by the nonlinear adjustment model , and :

[0131] ;

[0132] ;

[0133] ;

[0134] in, represents the nonlinear adjustment coefficient of temperature, represents the nonlinear adjustment coefficient of humidity, Indicates the nonlinear adjustment coefficient of oxygen concentration;

[0135] S55, updating the control instruction after the incremental adjustment to , and And re-input it into the control system to optimize the environmental parameter adjustment process;

[0136] S56. Continuously monitor and record the temperature, humidity and oxygen concentration at each level during the fermentation process, conduct real-time data analysis and feedback, and dynamically adjust the control strategy.

[0137] In this implementation manner, S6 specifically includes:

[0138] S61, real-time detection of temperature, humidity and oxygen concentration parameters of the first level, the second level and the third level in the fermentation environment, and inputting them into the central control system for analysis;

[0139] S62. Based on the real-time collected data, calculate the parameter differences between the layers and define the cross-layer temperature difference , Humidity difference and oxygen concentration difference ;

[0140] S63. Based on the cross-layer parameter differences, a cross-layer parameter balancing strategy is applied to dynamically adjust the environmental parameters:

[0141] ;

[0142] in, Indicates the intensity of temperature regulation. represents the initial temperature adjustment coefficient, Indicates the temperature regulation threshold, represents the temperature nonlinear amplification coefficient, represents the temperature intensity adjustment coefficient, represents the logarithmic temperature adjustment coefficient, represents the critical point of the piecewise function of temperature regulation intensity;

[0143] ;

[0144] in, Indicates the intensity of humidity regulation. represents the initial humidity adjustment coefficient, Indicates the humidity adjustment threshold, represents the humidity nonlinear amplification factor, represents the humidity intensity adjustment coefficient, represents the logarithmic adjustment coefficient of humidity, represents the critical point of the piecewise function of humidity regulation intensity;

[0145] ;

[0146] in, Indicates the intensity of regulation of oxygen concentration. Indicates the initial adjustment coefficient of oxygen concentration, Indicates the oxygen concentration adjustment threshold, represents the nonlinear amplification factor of oxygen concentration, Indicates the oxygen concentration intensity adjustment coefficient, represents the logarithmic adjustment coefficient of oxygen concentration, Represents the critical point of the piecewise function of oxygen concentration regulation intensity;

[0147] S64, the calculated adjustment intensity , and Input to the control system to adjust the output power of the heater, humidifier and ventilation system at each level;

[0148] S65, provide real-time feedback on the changes in parameters at all levels, update the model parameters in the central control system, and obtain the optimal state of the environmental parameters of the fermentation process.

[0149] Embodiment 1:

[0150] In order to verify the feasibility of the present invention in implementation, the present invention is applied to the fermentation workshop of a large tea production enterprise. Since tea fermentation has extremely stringent requirements on ambient temperature, humidity and oxygen concentration, any slight parameter deviation may lead to a decline in the quality of the final product. Especially in large-scale production, due to the significant environmental differences between the upper and lower levels in the fermentation room and the frequent changes in fermentation requirements in different time periods, traditional manual and simple automated control methods cannot meet the needs of high-quality fermentation, resulting in challenges to the consistency and market competitiveness of the company's products.

[0151] In order to solve these problems, the company decided to introduce the dynamic adjustment method of tea fermentation environment using fuzzy logic control of the present invention. This method achieves refined management of the fermentation process by real-time monitoring and precise control of the environment in the fermentation workshop. The company selected two fermentation rooms for comparative experiments, one of which used a traditional temperature and humidity controller for conventional fermentation, and the other room applied the dynamic adjustment system of the present invention.

[0152] Before the experiment began, the two rooms were equipped with the same amount and type of tea leaves and the same initial environmental parameters. The traditional fermentation room used a conventional temperature and humidity controller, while the new system fermentation room was equipped with temperature, humidity and oxygen concentration sensors at different levels. The sensors monitored the environmental parameters at each level in real time and transmitted the data to the central control system. Based on the collected data, the central control system used a fuzzy logic controller to fuzzify the environmental parameters at each level, and combined the historical fermentation data with the real-time monitoring data to generate an environmental parameter change model through a hybrid neural network and meta-learning model. This model is used to dynamically optimize the fuzzy rule base and ultimately generate precise control instructions to adjust the temperature, humidity and oxygen concentration at each level.

[0153] During the fermentation process, the temperature and humidity controller of the traditional room is adjusted using preset rules, but due to the complex environmental changes, the control effect is lagging. Especially in the middle and late stages of fermentation, the frequency of changes in environmental parameters is high, and the response speed of the traditional system is insufficient, resulting in a gradual increase in the temperature and humidity differences between the upper and lower levels. In the room where the present invention is applied, the system dynamically adjusts the output power of the heater, humidifier and ventilation system by real-time monitoring of the parameter differences of each level, applying a cross-layer parameter balancing strategy, so that the environmental parameters of each level are always kept within the optimal range. In addition, the system can also automatically adjust the control strategy according to the real-time changes in the environment and the analysis results of historical data, avoiding errors caused by manual intervention.

[0154] Table 1 Comparison data of tea fermentation environmental control systems

[0155]

[0156] Table 1 above clearly shows that there are significant differences in the performance of traditional fermentation rooms and the rooms of the system of the present invention in multiple key parameters. First of all, from the perspective of temperature control, the temperature in traditional fermentation rooms fluctuates greatly, with the highest temperature difference reaching 3°C. Such fluctuations may cause uneven active enzyme reactions of tea leaves during the fermentation process, thereby affecting the quality of the final tea leaves. In the system of the present invention, through precise temperature regulation, the temperature is always maintained within the range of 30±0.5°C. This stable temperature control ensures the uniform fermentation of tea leaves during the fermentation process and improves the quality consistency of the finished product.

[0157] In terms of humidity control, the humidity fluctuations in traditional fermentation rooms are also quite significant, with a humidity difference of 7%. Humidity instability may cause some tea leaves to ferment excessively or insufficiently, thus affecting the aroma and taste of the tea leaves. In contrast, the humidity control performance of the system of the present invention is excellent, and the humidity is always stable within the range of 85±2%, ensuring the suitability of the fermentation environment and avoiding the negative impact caused by humidity changes.

[0158] Control of oxygen concentration is another key factor in the fermentation process. The oxygen concentration in the traditional fermentation room fluctuates greatly, with the maximum difference reaching 1.5%, while the oxygen concentration in the room of the system of the present invention is controlled within a stable range of 20±0.5%. Stable oxygen concentration helps maintain the activity of microorganisms during the fermentation of tea, thereby ensuring the uniformity of fermentation and the quality of tea.

[0159] In addition, the environmental parameters of the upper and lower layers of the traditional fermentation room are quite different, with the temperature difference, humidity difference and oxygen difference reaching 3°C, 7% and 1.5% respectively. This difference between the upper and lower layers can easily lead to uneven fermentation of tea leaves in the fermentation room, affecting the overall fermentation quality. The system of the present invention significantly reduces the difference between the upper and lower layers through a cross-layer parameter balancing strategy, and the temperature difference, humidity difference and oxygen difference are controlled within 1°C, 3% and 0.5% respectively, greatly improving the consistency of fermentation.

[0160] Finally, the quality of the finished tea leaves also directly reflects the difference between the two fermentation methods. The tea leaves in the traditional fermentation room are fermented unevenly, with a qualified rate of only 85% and a fermentation failure rate of up to 5%. In contrast, the tea leaves in the room of the system of the present invention are fermented evenly, with a qualified rate of up to 98% and a fermentation failure rate of less than 1%. These data fully demonstrate the significant advantages of the system of the present invention in improving the quality of tea fermentation, reducing production losses and improving product consistency.

[0161] In summary, through comparative analysis, it can be concluded that the system of the present invention shows higher stability and accuracy in controlling the tea fermentation environment. It not only solves the problem of unstable control of temperature, humidity and oxygen concentration in traditional fermentation methods, but also significantly improves the fermentation quality and consistency of the finished product.

[0162] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A method for dynamically adjusting the tea fermentation environment using fuzzy logic control, characterized in that: The steps include: S1. Divide the tea fermentation environment into the first layer, the second layer and the third layer according to the spatial dimension, and arrange temperature, humidity and oxygen concentration sensors on each layer to collect and obtain the environmental parameter data of each layer in real time; S2, dividing the fermentation process into a first time period, a second time period and a third time period according to the time stage, and based on the environmental parameter data of each layer and combined with the fermentation time, fuzzy processing is performed on the temperature, humidity and oxygen concentration in each time period through a fuzzy logic controller to generate a fuzzy control output for each time period; S3, using the fuzzy control output of each time period and historical fermentation data, combined with the hybrid neural network architecture and meta-learning technology to analyze the changes in environmental parameters at the spatial level and time stage, and build an environmental parameter change model; S4, using the environmental parameter change model to optimize the fuzzy rule base; S5. Based on the optimized fuzzy rule base, the fuzzy control output of each time period is converted into control instructions to dynamically adjust the temperature, humidity and oxygen concentration in the fermentation environment at each spatial level and each time stage; S6, real-time detection of environmental parameter differences between different levels, optimization of the fermentation environment through cross-layer parameter balancing strategy, and obtaining the optimal state of environmental parameters during the fermentation process; S7, real-time display of environmental parameters such as temperature, humidity and oxygen concentration at each spatial level and time stage, and support for manual and automatic mode switching; The S3 specifically includes: S31. Collect fuzzy control output data of each time period during the fermentation process , , ,in represents the fuzzy control output of temperature, represents the fuzzy control output of humidity, The fuzzy control output representing the oxygen concentration, Indicates time period; S32, integrating the fuzzy control output data of each time period with the historical fermentation data, the historical fermentation data including the historical temperature data , Historical humidity data and historical oxygen concentration data , and the actual quality results of tea fermentation ; S33, using a hybrid neural network architecture to process the integrated data, the hybrid neural network architecture includes a convolutional neural network, a recurrent neural network and an adaptive weight adjustment mechanism, and the extraction of spatial features is done through a nonlinear function conduct: ； in, Represents spatial features, represents a convolutional neural network; S34, extracting spatial features Input into the recurrent neural network and extract time series features through weighted recurrent calculation: ； in, Represents the time series characteristics, represents the weight of adaptive learning, represents the number of layers of the recurrent neural network, represents a recurrent neural network; S35. Adaptive learning weights in hybrid neural network architectures via meta-learning algorithms and nonlinear functions The parameters are optimized, and the optimization goal is to minimize the nonlinear loss function : ； in, represents the adaptive adjustment coefficient, represents the predicted tea quality result, Represents historical quality results, Indicates the number of samples; S36. Generate an environmental parameter change model through the processed and optimized model. The environmental parameter change model includes prediction curves of temperature, humidity and oxygen concentration in each time period. , and , and control recommendations; The S4 specifically includes: S41, based on the environmental parameter change model generated in S36, the temperature, humidity and oxygen concentration prediction curves for each time period are calculated. , and As input, a preliminary analysis of the existing fuzzy rule base is conducted, including the shape and parameters of the membership functions, as well as the rationality of the rule weight distribution; S42, optimizing the membership function, adjusting the parameters of the membership function using the environmental parameter change model, and obtaining an optimized temperature membership function , humidity membership function and oxygen concentration membership function ; S43, based on optimizing the membership function, using the error between historical data and predicted data Adjust the parameters of the membership functions: ； in, represents the predicted temperature, represents the predicted humidity, represents the predicted oxygen concentration, , and represents the initial membership function value, Indicates the number of samples; S44, using the time series features extracted in step S34 Dynamic optimization of fuzzy rules, including readjusting the weight of each fuzzy rule according to the characteristics of the time series; S45, verifying the optimized fuzzy rule base, and testing the optimized rule base using real environment parameters and historical data; S46. Apply the optimized and verified fuzzy rule base to the system to replace the original fuzzy rule base.

2. A method for dynamically adjusting tea fermentation environment using fuzzy logic control according to claim 1, characterized in that: The S2 specifically includes: S21, dividing the fermentation process into a first time period, a second time period and a third time period according to the time stage, and performing fuzzy processing on the environmental parameter data of each layer obtained in step S1 in each time period; S22, in the first time period, the real-time temperature, humidity and oxygen concentration of each layer are input into the fuzzy logic controller, and the real-time temperature, humidity and oxygen concentration are fuzzified through the fuzzy membership function to generate a fuzzy set ; S23, in the second time period, the real-time temperature, humidity and oxygen concentration of each layer are input into the fuzzy logic controller, and the real-time temperature, humidity and oxygen concentration are fuzzified through the fuzzy membership function to generate a fuzzy set ; S24, in the third time period, the real-time temperature, humidity and oxygen concentration of each layer are input into the fuzzy logic controller, and the real-time temperature, humidity and oxygen concentration are fuzzified by the fuzzy membership function to generate a fuzzy set ; S25. Through fuzzy logic reasoning, use the inference rules: :if and and ,but ; in, Indicates Fuzzy rules, , and represents the input fuzzy set, represents the output fuzzy set, Indicates temperature, Indicates humidity, Indicates oxygen concentration, Represents the final fuzzy control output: ； in, represents the membership function value of the input temperature, represents the membership function value of the input humidity, represents the membership function value of the input oxygen concentration, Represents the total number of fuzzy rules.

3. The method for dynamically adjusting the tea fermentation environment using fuzzy logic control according to claim 1, characterized in that: The S5 specifically includes: S51, based on the optimized and verified fuzzy rule base, the fuzzy control output generated in each time period is defuzzified to generate control instructions , and ,in Used to regulate the temperature in the fermentation environment. Used to regulate the humidity in the fermentation environment. Used to regulate the oxygen concentration in the fermentation environment; S52, the generated control instruction , and Input to the control system, the regulation process includes controlling the output power or flow of heaters, humidifiers and ventilation systems; S53, real-time monitoring of the actual temperature in the fermentation environment , Real-time humidity and actual oxygen concentration , the actual temperature , Real-time humidity and actual oxygen concentration and temperature target parameters , Humidity target parameters and oxygen concentration target parameters Compare and introduce adaptive error adjustment coefficient Calculate dynamic error : ； in, Represents the adaptive error adjustment coefficient, which is dynamically adjusted based on the fermentation stage; S54, based on the calculated dynamic error , the increment of the control command is calculated by the nonlinear adjustment model , and : ； ； ； in, represents the nonlinear adjustment coefficient of temperature, represents the nonlinear adjustment coefficient of humidity, Indicates the nonlinear adjustment coefficient of oxygen concentration; S55, updating the control instruction after the incremental adjustment to , and And re-input it into the control system to optimize the environmental parameter adjustment process; S56. Continuously monitor and record the temperature, humidity and oxygen concentration at each level during the fermentation process, conduct real-time data analysis and feedback, and dynamically adjust the control strategy.

4. The method for dynamically adjusting the tea fermentation environment using fuzzy logic control according to claim 1, characterized in that: The S6 specifically includes: S61, real-time detection of temperature, humidity and oxygen concentration parameters of the first level, the second level and the third level in the fermentation environment, and inputting them into the central control system for analysis; S62. Based on the real-time collected data, calculate the parameter differences between the layers and define the cross-layer temperature difference , Humidity difference and oxygen concentration difference ; S63. Based on the cross-layer parameter differences, a cross-layer parameter balancing strategy is applied to dynamically adjust the environmental parameters: ； in, Indicates the intensity of temperature regulation. represents the initial temperature adjustment coefficient, Indicates the temperature regulation threshold, represents the temperature nonlinear amplification coefficient, represents the temperature intensity adjustment coefficient, represents the logarithmic temperature adjustment coefficient, represents the critical point of the piecewise function of temperature regulation intensity; ； in, Indicates the intensity of humidity regulation. represents the initial humidity adjustment coefficient, Indicates the humidity adjustment threshold, represents the humidity nonlinear amplification factor, represents the humidity intensity adjustment coefficient, represents the logarithmic humidity adjustment coefficient, represents the critical point of the piecewise function of humidity regulation intensity; ； in, Indicates the intensity of regulation of oxygen concentration. Indicates the initial adjustment coefficient of oxygen concentration, Indicates the oxygen concentration adjustment threshold, represents the nonlinear amplification factor of oxygen concentration, Indicates the oxygen concentration intensity adjustment coefficient, represents the logarithmic adjustment coefficient of oxygen concentration, Represents the critical point of the piecewise function of oxygen concentration regulation intensity; S64, the calculated adjustment strength , and Input to the control system to adjust the output power of the heater, humidifier and ventilation system at each level; S65, provide real-time feedback on the changes in parameters at all levels, update the model parameters in the central control system, and obtain the optimal state of the environmental parameters of the fermentation process.

Images