Tea leaf fixation process adjusting method and system and medium

Abstract

The invention discloses a method and a system for adjusting a tea leaf fixation process and a medium, and aims to realize cooperative regulation and control of enzyme activity and aroma substances in a fixation process. The method comprises the following steps: collecting multi-modal sensing information in a fixation process, and carrying out feature extraction on the multi-modal sensing information to obtain multi-dimensional features; dynamically constructing an LSTM prediction model based on the multi-dimensional features, and outputting an enzyme activity prediction value sequence; the enzyme activity predicted value sequence is input into a fuzzy PID controller for reinforcement learning optimization, a temperature-rotating speed adjusting instruction is generated, and adjusting parameters of the controller are dynamically updated through a reward function of the chlorophyll retention rate and the energy consumption ratio; according to the temperature-rotating speed adjusting instruction, adjusting the hot air temperature of each temperature area of the enzyme deactivation chamber, the rotating speed of the enzyme deactivation cylinder and the axial segmentation wind speed; and collecting fixation end point judgment information, and terminating fixation when a preset condition is met. According to the invention, the problems of unstable quality and overhigh energy consumption caused by parameter solidification in the traditional enzyme deactivation process are solved, and dynamic optimization and accurate control of the enzyme deactivation process are realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of tea processing, and in particular to a tea fixation process adjustment method, system and medium. BACKGROUND

[0002] Tea fixation is a key process in tea processing, and its core is to quickly inactivate enzyme activity (such as polyphenol oxidase) through high temperature, while retaining quality ingredients such as chlorophyll, and promoting the formation of aroma substances. Traditional fixation processes mostly rely on fixed temperature curves or manual experience to adjust hot air temperature, drum speed and other parameters. However, only monitoring temperature through a thermocouple cannot capture the relationship between enzyme activity dynamic changes and aroma substance volatilization in real time, which can easily lead to a lag in fixation degree judgment, and traditional PID control cannot adapt to tea moisture content, variety differences and other variables, which can easily cause local over-fixation or fixation deficiency. In addition, only using time or apparent temperature as a standard, ignoring the synergistic indicators of chlorophyll retention rate and key aroma components, can lead to large fluctuations in fixation quality and high energy consumption, and long downtime is required to adjust parameters when switching tea varieties. SUMMARY

[0003] Therefore, the purpose of the present application is to provide a tea fixation process adjustment method, system and medium, to realize the synergistic and accurate regulation of enzyme activity and aroma substances in the fixation process, and to solve the problems of unstable quality and high energy consumption in traditional fixation processes.

[0004] In order to achieve the above technical purposes, in a first aspect, the present application provides a tea fixation process adjustment method, comprising: Collecting multi-modal sensing information in the fixation process, the multi-modal sensing information including infrared temperature time series information, chlorophyll fluorescence intensity and volatile organic matter concentration; Extracting features from the multi-modal sensing information to obtain multi-dimensional features, the multi-dimensional features including enzyme activity decay curve, aroma volatilization characteristic spectrum and temperature fluctuation characteristics; Dynamically constructing an LSTM prediction model according to the multi-dimensional features, and outputting an enzyme activity prediction value sequence of a prediction time segment; Inputting the enzyme activity prediction value sequence into a fuzzy PID controller optimized by reinforcement learning to generate temperature-rotation speed adjustment instructions, and the adjustment parameters of the fuzzy PID controller are configured to be dynamically updated by a reward function of chlorophyll retention rate and energy consumption ratio; Controlling edge execution nodes in each temperature zone in the fixation chamber according to the temperature-rotation speed adjustment instructions, and the edge execution nodes adjust hot air temperature, fixation drum speed and axial segmented wind speed; Collecting fixation end judgment information, the fixation end judgment information including chlorophyll fluorescence threshold, volatile organic matter ratio and leaf surface temperature, and terminating fixation when the fixation end judgment information meets a preset judgment condition.

[0005] In some embodiments, feature extraction is performed on the multi-modal sensing information to obtain multi-dimensional features, including: The infrared temperature time series information is segmented by a sliding window to obtain a plurality of temperature sequence windows; The temperature data in each temperature sequence window is subjected to frequency spectrum analysis to extract the principal frequency components and the energy distribution characteristics corresponding to each principal frequency component, thereby obtaining temperature fluctuation characteristics; The chlorophyll fluorescence intensity is subjected to digital filtering processing to remove high-frequency noise, thereby obtaining smooth fluorescence intensity information; The smooth fluorescence intensity information is subjected to differential calculation to obtain an instantaneous change rate of enzyme activity; The instantaneous change rate of enzyme activity is subjected to time fitting by a non-linear least squares method to obtain an enzyme activity decay curve of the fluorescence intensity with respect to time; The concentrations of the components in the volatile organic compounds are monitored in real time, and a dynamic concentration ratio of a preset aroma substance is calculated; Based on the dynamic concentration ratio, a conversion relationship model between the components of the volatile organic compounds is established by time series regression analysis; The dynamic change characteristics of the aroma substance are extracted according to the conversion relationship model, and an aroma volatilization characteristic spectrum is constructed; The temperature fluctuation characteristics, the enzyme activity decay curve, and the aroma volatilization characteristic spectrum are subjected to normalization processing to generate multi-dimensional features.

[0006] In some embodiments, an LSTM prediction model is dynamically constructed according to the multi-dimensional features, and an enzyme activity prediction value sequence of a prediction time section is output, including: The temperature fluctuation characteristics, the enzyme activity decay curve, and the aroma volatilization characteristic spectrum are subjected to time sequence alignment processing to construct a multi-dimensional feature input sequence; The multi-dimensional feature data collected in real time is input into the LSTM network model which has been trained; An original enzyme activity prediction value sequence of the prediction time section is output by forward propagation calculation of the LSTM network; The original enzyme activity prediction value sequence is subjected to smoothing processing to obtain an enzyme activity prediction value sequence; The LSTM prediction model is constructed by the following steps: The basic structure of the LSTM prediction model is initialized, and the number of neurons of the input layer, the hidden layer, and the output layer is configured; The sample training features of the same tea category are divided into a training sample set in a sliding time window manner; The LSTM prediction model is iteratively trained using the training sample set, and the network weight parameters are optimized until the LSTM prediction model is trained.

[0007] In some embodiments, the enzyme activity prediction value sequence is input into a fuzzy PID controller optimized by reinforcement learning to generate temperature-rotation speed adjustment instructions, including: The membership functions of the input variables and the membership functions of the output variables of the fuzzy PID controller are established, the input variables including the deviation and the rate of change of the deviation of the enzyme activity prediction value sequence, and the output variables including the temperature adjustment amount and the rotation speed adjustment amount; A reinforcement learning agent is constructed, and the state space, action space and reward function of the reinforcement learning agent are configured, the state space containing the current PID parameters and the control effect evaluation index, the action space being the adjustment amplitude of the PID parameters, and the reward function being calculated based on the chlorophyll retention rate and the energy consumption ratio; The interaction information generated by the interaction between the reinforcement learning agent and the fixation device is used to dynamically optimize the parameter correction strategy of the fuzzy PID controller; The enzyme activity prediction value sequence is input into the optimized fuzzy PID controller to calculate the fuzzy control output information of the temperature adjustment amount and the rotation speed adjustment amount; The fuzzy control output information is de-fuzzified to generate accurate initial temperature-rotation speed adjustment instructions; The initial temperature-rotation speed adjustment instructions are checked for feasibility according to the current working state in the fixation chamber; The initial temperature-rotation speed adjustment instructions that pass the check are recorded as temperature-rotation speed adjustment instructions and output.

[0008] In some embodiments, a reinforcement learning agent is constructed, and the state space, action space and reward function of the reinforcement learning agent are configured, the state space containing the current PID parameters and the control effect evaluation index, the action space being the adjustment amplitude of the PID parameters, and the reward function being calculated based on the chlorophyll retention rate and the energy consumption ratio, including: The state space is represented by formula (1), and formula (1) is as follows: ; In formula (1), is the state space, is the parameter combination of the fuzzy PID controller, is the proportional coefficient of the fuzzy PID controller, is the integral coefficient of the fuzzy PID controller, is the differential coefficient of the fuzzy PID controller, is the control effect evaluation index, is the chlorophyll retention rate deviation, is the energy consumption ratio deviation; The action space is represented by formula (2), and formula (2) is as follows: ; In formula (2), is an action space, is an adjustment amount of a proportional coefficient, , is a limit adjustment amplitude, is an adjustment amount of an integral coefficient, , is a limit adjustment amplitude, is an adjustment amount of a differential coefficient, , is a limit adjustment amplitude; The reward function is represented by formula (3), and formula (3) is as follows: ; In formula (3), is a reward function, is a weight coefficient of a chlorophyll retention rate, is a weight coefficient of energy consumption efficiency, is an absolute value deviation of the chlorophyll retention rate, is an energy consumption ratio deviation truncated by an upper limit, ; The policy network and the value network of the reinforcement learning agent are initialized, and training hyperparameters are configured, including a learning rate and a discount factor; The reinforcement learning agent interacts with the kill green equipment to generate interaction information, and dynamically optimizes the parameter correction strategy of the fuzzy PID controller according to the interaction information, including: The interaction information generated in the real-time interaction process of the reinforcement learning agent and the kill green equipment is collected, and the interaction information includes control state, execution action, obtained reward and new control state; Based on the interaction information, an experience sample library is constructed, and training data is extracted from the experience sample library by random sampling; The training data is used to update the parameter weights of the policy network and the value network, and a new parameter adjustment strategy is generated; The new parameter adjustment strategy is applied to the parameter update of the fuzzy PID controller, and the parameter update includes proportional coefficient update, integral coefficient update and differential coefficient update; The updated fuzzy PID controller is evaluated in value, and the fuzzy PID controller passing the value evaluation is verified for stability, so that the control parameters are within a preset safe range; The new parameter adjustment strategy passing the stability verification is output and executed.

[0009] In some embodiments, the enzyme activity prediction value sequence is input into the optimized fuzzy PID controller, and the fuzzy control output information of the temperature adjustment amount and the speed adjustment amount is calculated, including: The enzyme activity prediction value sequence is subjected to bias calculation to obtain enzyme activity prediction value bias and bias change rate; The enzyme activity prediction value bias and bias change rate are input into a fuzzification interface to be converted into corresponding fuzzy language variables; The applicable temperature adjustment fuzzy rule and the speed adjustment fuzzy rule are matched through the fuzzy PID controller; The matched temperature adjustment fuzzy rule and speed adjustment fuzzy rule are subjected to reasoning operation to generate fuzzy control output information containing temperature adjustment amount and speed adjustment amount; The fuzzy control output information is subjected to defuzzification processing to generate accurate initial temperature-speed adjustment instructions, including: The temperature adjustment amount in the fuzzy control output information is calculated by using the gravity method to obtain the initial temperature adjustment amount; The speed adjustment amount in the fuzzy control output information is calculated by using the maximum membership degree method to obtain the initial speed adjustment amount; The initial temperature adjustment amount and the initial speed adjustment amount are combined in time sequence to generate the initial temperature-speed adjustment instructions.

[0010] In some embodiments, the edge execution nodes of each temperature zone in the fixation chamber are controlled according to the temperature-speed adjustment instructions, and the edge execution nodes adjust the hot air temperature, the fixation drum speed and the axial segmented air speed, including: The temperature-speed adjustment instructions are decomposed into independent control instructions of each temperature zone, and the independent control instructions of each temperature zone include preheating zone temperature instructions, main fixation zone temperature instructions and stable zone temperature instructions; The preheating zone temperature instructions are subjected to signal conversion to generate a gas valve opening degree control signal to adjust the gas combustion intensity of the preheating zone; The main fixation zone temperature instructions are subjected to power modulation conversion to obtain the output power of the heating element to adjust the hot air temperature of the main fixation zone; The stable zone temperature instructions are subjected to PID operation to obtain the fan speed adjustment amount to adjust the temperature interval of the stable zone; And the temperature-speed adjustment instructions are input into a variable frequency drive to adjust the operating frequency of the drive unit of the fixation drum.

[0011] In some embodiments, fixation endpoint determination information is collected, the fixation endpoint determination information includes chlorophyll fluorescence threshold value, volatile organic matter ratio and leaf surface temperature, and the fixation is terminated when the fixation endpoint determination information meets a preset determination condition, including: The chlorophyll fluorescence threshold value is subjected to real-time sampling to obtain a current fluorescence intensity value; The current fluorescence intensity value is compared with a preset fluorescence threshold value to calculate a fluorescence intensity decay rate; When the fluorescence intensity decay rate reaches a preset enzyme inactivation threshold value, a first termination signal is obtained; In addition, the concentration of each component in the volatile organic compound is monitored in real time, and a volatile organic compound ratio is calculated, the volatile organic compound ratio including a real-time concentration ratio of linalool and jasmone; When the volatile organic compound ratio reaches a preset aroma balance threshold value, a second termination signal is obtained; In addition, the leaf surface temperature is collected in real time, and an average value of the leaf surface temperature is calculated to obtain a leaf surface temperature average value; When the leaf surface temperature average value is lower than a preset surface temperature threshold value, a third termination signal is obtained; The method further comprises: When the first termination signal and the second termination signal are received simultaneously, a fixation termination instruction is triggered immediately; When only the first termination signal or the second termination signal is received, a delay determination procedure is started, and the third termination signal is detected within a preset time.

[0012] In a second aspect, the present application further provides a tea leaf fixation process adjustment system, which is suitable for the adjustment method in the first aspect, and comprises a fixation device, a driving unit, a hot air assembly, a temperature sensor, an olfactory sensor, an image acquisition unit and a control unit. The fixation device comprises a fixation cylinder, and the fixation cylinder has a fixation chamber in which the tea leaves to be fixed are contained. The driving unit is in transmission connection with the fixation cylinder. The hot air assembly comprises at least two fans, and the at least two fans are distributed in the fixation cylinder in a preset manner. The temperature sensor is arranged in the fixation cylinder. The olfactory sensor is arranged in the fixation cylinder. The image acquisition unit is arranged in the fixation cylinder. The control unit is electrically connected with the driving unit, the hot air assembly, the temperature sensor, the olfactory sensor and the image acquisition unit respectively, and is used for executing the method in the first aspect.

[0013] In a third aspect, the present application further provides a computer readable storage medium, which stores computer program instructions, and the computer program instructions are used to realize the method in the first aspect when executed by a processor.

[0014] By using the above technical solution, the present application has the following beneficial effects compared with the prior art: The application provides a tea leaf fixation process adjusting method, system and medium, aiming to realize the synergistic and accurate regulation of enzyme activity and aroma substances in the fixation process. The method comprises the following steps: collecting multi-modal sensing information in the fixation process, including infrared temperature time series information, chlorophyll fluorescence intensity and volatile organic compound concentration; performing feature extraction on the multi-modal sensing information to obtain an enzyme activity decay curve, an aroma volatile feature spectrum and a temperature fluctuation feature; dynamically constructing an LSTM prediction model based on the multi-dimensional features to output an enzyme activity prediction value sequence; inputting the enzyme activity prediction value sequence into a fuzzy PID controller optimized by reinforcement learning to generate a temperature-rotation speed adjusting instruction, and dynamically updating the adjusting parameters of the PID controller through a reward function of chlorophyll retention rate and energy consumption ratio; adjusting the edge execution nodes of each temperature zone of the fixation chamber according to the temperature-rotation speed adjusting instruction to adjust the hot air temperature, the fixation cylinder rotation speed and the axial segmented air speed; and collecting fixation end point judgment information and terminating the fixation when the preset condition is met. The application solves the problems of unstable quality and excessive energy consumption caused by the parameter fixation of the traditional fixation process, and realizes the dynamic optimization and accurate control of the fixation process. BRIEF DESCRIPTION OF DRAWINGS

[0015] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or the prior art description. Obviously, the drawings in the following description are only some embodiments of the present application, and all other drawings obtained by those skilled in the art without creative labor are within the protection scope of the present application.

[0016] Figure 1 is a method step diagram of steps S101 to S106 of the adjusting method described in the specific embodiment; Figure 2 is a method step diagram of steps S201 to S208 of the adjusting method described in the specific embodiment. DETAILED DESCRIPTION

[0017] The present application will be further described in detail below in combination with the drawings and embodiments. It is particularly pointed out that the following embodiments are only used to illustrate the present application, but do not limit the scope of the present application. Similarly, the following embodiments are only some embodiments of the present application, not all embodiments, and all other embodiments obtained by those skilled in the art without creative labor are within the protection scope of the present application.

[0018] Please refer to Figure 1 In the first aspect, the present embodiment provides a tea leaf fixation process adjusting method, comprising: S101, collecting multi-modal sensing information in the fixation process, the multi-modal sensing information comprising infrared temperature time series information, chlorophyll fluorescence intensity and volatile organic compound concentration; S102, feature extraction is performed on the multi-modal sensing information to obtain multi-dimensional features, including enzyme activity decay curve, aroma volatile characteristic spectrum, and temperature fluctuation characteristic; S103, an LSTM prediction model is dynamically constructed according to the multi-dimensional features, and an enzyme activity prediction value sequence of a prediction time section is output; S104, the enzyme activity prediction value sequence is input into a fuzzy PID controller optimized by reinforcement learning to generate a temperature-rotation speed adjustment instruction, and the adjustment parameters of the fuzzy PID controller are configured to be dynamically updated by a reward function of chlorophyll retention rate and energy consumption ratio; S105, the edge execution nodes of each temperature zone in the fixation chamber are controlled according to the temperature-rotation speed adjustment instruction, and the edge execution nodes adjust the hot air temperature, the fixation cylinder rotation speed, and the axial segmented air speed; S106, fixation endpoint determination information is collected, including chlorophyll fluorescence threshold, volatile organic matter ratio, and leaf surface temperature, and the fixation is terminated when the fixation endpoint determination information meets the preset determination condition; further, the fixation process parameters, the adjustment instruction, and the determination result are stored in the block chain, and a traceable process chain including enzyme activity dynamic curve, aroma quality score, and energy consumption distribution is generated.

[0019] In step S101, the adaptive sampling strategy is adopted in the collection process of multi-modal sensing information, preferably, the spatial resolution of infrared temperature time series information is dynamically adjusted according to the fixation cylinder rotation speed, and when the rotation speed is increased, the transverse temperature measurement point density is increased to compensate for the monitoring blind area caused by leaf turning; the detection wavelength of chlorophyll fluorescence intensity is adaptively selected according to the spectral characteristics of tea varieties to ensure the sensitivity to different chlorophyll isomers; the monitoring of volatile organic matter concentration focuses on establishing a dynamic targeted detection list for characteristic aroma components of varieties, and the monitoring intensity of terpenes is automatically enhanced in the later stage of fixation. The intelligent adjustment of the perception strategy effectively solves the problem of insufficient adaptability of traditional fixed parameter sensing system to process changes.

[0020] In step S102, preferably, the attention mechanism is introduced to enhance the recognition of key features, and when extracting the enzyme activity decay curve, the time domain attention weight is used to highlight the feature points in the temperature sensitive stage; the frequency domain attention is used to focus on the resonance peak of the characteristic aroma component in the aroma volatile characteristic spectrum; and the space attention mechanism is used to strengthen the representation of the high variation area. This step significantly improves the ability of the subsequent model to capture weak quality signals, especially the early identification of the fixation turning point.

[0021] In step S103, the LSTM prediction model adopts a multi-task learning architecture, the main task predicts the enzyme activity, while the auxiliary task simultaneously predicts the chlorophyll degradation trend and the aroma generation potential, the hidden layers of the three tasks share feature representations but the output layers are independent, the gradient mask mechanism is used to realize the cooperative optimization of different prediction targets, so that the model can capture the deep correlation between enzyme activity change and quality index, and the prediction result is more in line with the coupling characteristics of the actual process.

[0022] In step S104, preferably, the parameter optimization of the fuzzy PID controller adopts a hierarchical reinforcement learning strategy, the high-level strategy quickly initializes the control parameters at variety switching through a meta-learning framework, and the low-level strategy adjusts the proportional coefficient, integral coefficient and differential coefficient in real time through online learning; the reward function introduces a dynamic weight mechanism, which focuses on chlorophyll retention rate at the beginning of fixation and gradually increases the weight of energy consumption in the later stage. This time-varying optimization target is more in line with the phased characteristics of tea fixation.

[0023] In step S105, the regulation of the edge execution node can introduce a feedforward-feedback composite control, the feedforward channel adjusts the hot air temperature in advance according to the tea moisture content prediction model, and the feedback channel compensates for the regulation deviation based on real-time sensing data. The axial segmented wind speed control adopts a computational fluid dynamics simulation assisted inverse model to calculate the optimal opening combination of the guide vane, so as to realize the precise spatial regulation of the heat field in the fixation cylinder.

[0024] In step S106, the fixation endpoint determination condition adopts a fuzzy logic dynamic adjustment strategy, the determination interval of the chlorophyll fluorescence threshold is automatically scaled according to the initial chlorophyll content; preferably, the volatile organic matter ratio introduces a moving time window statistical verification mechanism, which requires that the standard be met for three consecutive sampling periods; the gradient condition of the leaf surface temperature is gradually tightened as the fixation stage advances. Further, the blockchain storage can adopt a lightweight Merkle tree structure to ensure traceability while adapting to the data throughput requirements of the industrial site. The quality scoring module in the process chain integrates the transfer learning capability, which can quickly adapt to the evaluation standards of new varieties.

[0025] The embodiment realizes the cooperative regulation of enzyme activity and aroma generation through the LSTM prediction model and the fuzzy PID controller optimized by reinforcement learning, based on the real-time capture of temperature distribution, chlorophyll stability and aroma substance dynamic change in the fixation process through multi-modal sensing information. Through dynamic extraction and modeling of multi-dimensional features, the quality fluctuation problem caused by the fixation of traditional process parameters is solved; the distributed control architecture of the edge execution node is adopted to realize the precise cooperative adjustment of hot air temperature, fixation cylinder speed and axial segmented wind speed; combined with the blockchain traceable process chain, the complete correlation between process parameters and quality indicators is ensured, which improves the fixation uniformity while reducing energy consumption, providing an intelligent control paradigm that combines standardization and self-adaptation for tea processing.

[0026] Referring to Figure 2 In some embodiments, feature extraction is performed on the multi-modal sensing information to obtain multi-dimensional features, including: S201, performing sliding window segmentation on the infrared temperature time series information to obtain a plurality of temperature sequence windows; S202, performing spectral analysis on the temperature data in each temperature sequence window to extract the main frequency components and the energy distribution features corresponding to each main frequency component, and obtaining temperature fluctuation features; And, performing digital filtering processing on the chlorophyll fluorescence intensity to remove high-frequency noise and obtain smooth fluorescence intensity information; S203, performing difference calculation on the smooth fluorescence intensity information to obtain the instantaneous change rate of enzyme activity; S204, performing time fitting on the instantaneous change rate of enzyme activity by a nonlinear least squares method to obtain an enzyme activity decay curve of the fluorescence intensity over time; S205, monitoring the concentration of each component in the volatile organic compound in real time, and calculating the dynamic concentration ratio of the preset aroma substance; S206, based on the dynamic concentration ratio, a conversion relationship model between the components of the volatile organic compound is established by time series regression analysis; S207, extracting the dynamic change characteristics of the aroma substance based on the conversion relationship model, and constructing an aroma volatilization characteristic spectrum; S208, normalizing the temperature fluctuation features, the enzyme activity decay curve, and the aroma volatilization characteristic spectrum to generate multi-dimensional features.

[0027] In step S201, the sliding window segmentation adopts an adaptive window length strategy. Preferably, the window length is dynamically adjusted according to the fixation cylinder rotation speed to ensure that each window contains a complete temperature fluctuation period. The division of the temperature sequence window is realized by time stamp alignment to ensure the time sequence consistency with the chlorophyll fluorescence intensity data. The spectral analysis adopts the fast Fourier transform algorithm, the extraction of the main frequency component is completed by peak detection of the power spectral density, and the energy distribution features are obtained by calculating the energy proportion of each frequency band. These features together constitute the temperature fluctuation features reflecting the heat conduction characteristics in the fixation chamber.

[0028] In step S202, a Butterworth low-pass filter is used for digital filtering processing of the chlorophyll fluorescence intensity. The cutoff frequency is set according to the typical fluctuation range of the chlorophyll fluorescence signal, which can effectively suppress high-frequency noise while retaining the true fluorescence intensity change trend. Further, the acquisition of the smooth fluorescence intensity information is realized by recursive filtering, which minimizes the phase distortion while ensuring the smoothness of the data.

[0029] In step S203, the differential calculation adopts the central difference method, and the calculation step of the transient change rate of enzyme activity is synchronized with the temperature sequence window, so as to ensure the time alignment of the multi-modal characteristics.

[0030] In step S204, the exponential decay model is selected for the nonlinear least squares fitting, and the initial parameters are initialized by the prior knowledge of the change of enzyme activity. Preferably, the fitting process is optimized by using the Levenberg-Marquardt algorithm, and the obtained enzyme activity decay curve contains characteristic parameters representing the enzyme inactivation kinetics, such as half-life and decay rate constant. The fitting degree of the enzyme activity decay curve is verified by residual analysis to ensure that the change rule of the activity of the key enzyme in the fixation process can be accurately reflected.

[0031] In step S205, the preset aroma substance refers to the characteristic aromatic components generated in the fixation process of tea leaves, such as terpenes, alcohols and other key volatile organic compounds that determine the aroma type of tea leaves. Preferably, the dynamic concentration ratio of the preset aroma substance is calculated by using the relative peak area method, and the target component is determined according to the characteristic aroma fingerprint spectrum of tea varieties. The ratio calculation introduces moving average smoothing processing to eliminate the influence of instantaneous detection fluctuations. The update frequency of the dynamic concentration ratio is matched with the adjustment period of the fixation process, so as to provide a stable input sequence for subsequent conversion relationship modeling.

[0032] In steps S206 and S207, the time series regression analysis adopts the vector autoregressive model, and the lag order is automatically determined by the information criterion. The establishment of the conversion relationship model includes the calculation and significance test of the conversion rate between components, and only the statistically significant conversion path is retained. The construction of the aroma volatilization characteristic spectrum is completed by extracting the key conversion coefficients in the model, and a characteristic matrix reflecting the dynamic balance between the generation and degradation of aroma substances is formed.

[0033] In step S208, the normalization processing preferably adopts the dynamic extreme value method of each characteristic parameter, and the scaling coefficient is adaptively adjusted according to the process characteristics of the current batch of tea. The generation of multi-dimensional characteristics is realized by splicing the characteristic vectors, and the original dimensional relationship of the temperature fluctuation characteristics, enzyme activity decay curve and aroma volatilization characteristic spectrum is maintained. The normalized characteristic vectors are verified for the independence between characteristics by covariance matrix analysis, so as to ensure that the characteristic space input into the prediction model has good separability.

[0034] The embodiment extracts temperature fluctuation characteristics containing main frequency components and energy distribution characteristics by sliding window segmentation and spectral analysis of infrared temperature time series information; meanwhile, the enzyme activity instantaneous change rate is obtained by digital filtering and difference calculation of chlorophyll fluorescence intensity, and the enzyme activity decay curve is fitted; combined with the dynamic concentration ratio analysis of volatile organic compounds, the aroma substance conversion relationship model is established and the aroma volatilization characteristic spectrum is constructed. The combination of sliding window segmentation and spectral analysis completely retains the periodic characteristics of temperature fluctuations; through digital filtering processing and nonlinear least squares fitting, the dynamic process of enzyme activity change is accurately characterized; based on the time series regression analysis, the conversion relationship model is established, which effectively reflects the dynamic conversion law of aroma substances; finally, through normalization processing, the multi-dimensional characteristics are realized, which realizes the organic integration of temperature fluctuation characteristics, enzyme activity decay curve and aroma volatilization characteristic spectrum, and provides comprehensive and reliable characteristic basis for process optimization.

[0035] In some embodiments, an LSTM prediction model is dynamically constructed according to the multi-dimensional characteristics, and an enzyme activity prediction value sequence of a prediction time period is output, including: The temperature fluctuation characteristics, enzyme activity decay curve and aroma volatilization characteristic spectrum are processed in time sequence to construct a multi-dimensional characteristic input sequence; The real-time collected multi-dimensional characteristic data is input into the trained LSTM network model; The original enzyme activity prediction value sequence of the prediction time period is output through the forward propagation calculation of the LSTM network; The original enzyme activity prediction value sequence is smoothed to obtain an enzyme activity prediction value sequence; The LSTM prediction model is constructed by the following steps: The basic structure of the LSTM prediction model is initialized, and the number of neurons of the input layer, hidden layer and output layer is configured; The sample training characteristics of the same tea category are divided into a training sample set in a sliding time window manner; The LSTM prediction model is iteratively trained using the training sample set, and the network weight parameters are optimized until the LSTM prediction model is trained.

[0036] In the embodiment, the temperature fluctuation characteristics refer to the periodic characteristics extracted from the infrared temperature time series data by sliding window segmentation and spectral analysis, containing main frequency components and energy distribution information, for characterizing the heat conduction characteristics in the fixation chamber. The aroma volatilization characteristic spectrum is a conversion relationship model established based on the dynamic concentration ratio analysis of volatile organic compounds, reflecting the generation and degradation dynamic balance characteristics of aroma substances.

[0037] The time alignment processing refers to calibrating the temperature fluctuation characteristics, enzyme activity decay curve and aroma volatilization characteristic spectrum according to a unified time reference, so as to ensure that the multi-dimensional characteristic input sequence has consistency in the time dimension. In the basic structure configuration of the LSTM network model, the number of input layer neurons matches the dimension of the multi-dimensional characteristics, the number of hidden layer neurons is determined according to the characteristics of tea categories, and the output layer corresponds to the enzyme activity prediction value sequence of the prediction time section. The training characteristics of the same tea category sample are divided into a training sample set by using a sliding time window method, that is, the training samples are divided according to a fixed time interval, and the window length is adaptively adjusted according to the process cycle characteristics, so as to ensure that each window contains a complete process change cycle.

[0038] Further, the division of the training sample set needs to consider the process consistency of the same tea category, and the sample characteristics should include the temperature fluctuation, enzyme activity change and aroma substance conversion data under typical process conditions.

[0039] Preferably, the optimization of network weight parameters is realized by a back propagation algorithm, and an adaptive learning rate strategy is used to balance the training speed and convergence stability; the smoothing processing of the original enzyme activity prediction value sequence uses a moving average algorithm, and the window size is synchronized with the process adjustment period, so as to eliminate the instantaneous fluctuation interference.

[0040] The embodiment models the long-range dependence relationship of the time sequence characteristics by using the LSTM network, encodes the synergistic change law of the temperature fluctuation characteristics, enzyme activity decay curve and aroma volatilization characteristic spectrum into the network weight, and realizes accurate prediction of the enzyme activity change in the future time section. Among them, the temperature fluctuation characteristics reflect the external thermal action condition, the enzyme activity decay curve represents the internal biochemical reaction process, and the aroma volatilization characteristic spectrum indicates the quality formation trend, which together constitute a multi-dimensional characteristic space of the prediction model. The prediction model learns the nonlinear mapping relationship between the process parameters and the quality indicators through iterative training, and finally outputs the smoothed enzyme activity prediction value sequence, which provides a decision basis for real-time process regulation.

[0041] The embodiment realizes accurate prediction of the enzyme activity change by performing time alignment processing on the temperature fluctuation characteristics, enzyme activity decay curve and aroma volatilization characteristic spectrum and inputting them into the LSTM prediction model. The synergistic analysis of the multi-dimensional characteristics completely reflects the interaction of the heat conduction characteristics, enzyme inactivation kinetics and aroma conversion law in the fixation process; the long-range modeling capability of the LSTM network for time sequence characteristics ensures the reliability of the prediction result; the smoothed enzyme activity prediction value sequence provides a stable basis for process regulation, and effectively improves the accuracy and timeliness of tea quality control.

[0042] In some embodiments, the enzyme activity prediction value sequence is input into a fuzzy PID controller optimized by reinforcement learning to generate temperature-rotational speed adjustment instructions, including: The membership function of the input variable of the fuzzy PID controller and the membership function of the output variable are established, the input variable includes the deviation and the change rate of the enzyme activity prediction value sequence, and the output variable includes the temperature adjustment amount and the rotation speed adjustment amount; The reinforcement learning agent is constructed, the state space, the action space and the reward function of the reinforcement learning agent are configured, the state space includes the current PID parameter and the control effect evaluation index, the action space is the adjustment range of the PID parameter, and the reward function is calculated based on the chlorophyll retention rate and the energy consumption ratio; The interaction information generated by the interaction between the reinforcement learning agent and the fixation device is obtained, and the parameter correction strategy of the fuzzy PID controller is dynamically optimized according to the interaction information; The enzyme activity prediction value sequence is input into the optimized fuzzy PID controller, and the fuzzy control output information of the temperature adjustment amount and the rotation speed adjustment amount is calculated; The fuzzy control output information is de-fuzzified to generate accurate initial temperature-rotation speed adjustment instructions; The initial temperature-rotation speed adjustment instructions are checked for feasibility according to the current working state of the fixation chamber; The initial temperature-rotation speed adjustment instructions that pass the check are recorded as temperature-rotation speed adjustment instructions and are output.

[0043] In this embodiment, the input variable membership function of the fuzzy PID controller is used to define the fuzzy conversion rule of the deviation and the change rate of the enzyme activity prediction value sequence, and the output variable membership function corresponds to the fuzzy control output range of the temperature adjustment amount and the rotation speed adjustment amount.

[0044] The state space of the reinforcement learning agent includes the current PID parameter combination and the corresponding control effect evaluation index, the action space is defined as the adjustable range of the PID parameter, and the reward function is obtained by weighted calculation of the chlorophyll retention rate and the energy consumption ratio, which is used to guide the optimization direction of the agent.

[0045] The interaction information refers to the state transition data and control effect feedback generated by the reinforcement learning agent and the fixation device during the control process, which is used to dynamically optimize the parameter correction strategy of the fuzzy PID controller. Preferably, the de-fuzzification processing converts the fuzzy control output information into accurate initial temperature-rotation speed adjustment instructions by using the barycentric method. The feasibility check is based on the current working state parameters of the fixation chamber, including the real-time temperature distribution and the mechanical load capacity, to ensure the executability of the adjustment instructions.

[0046] The embodiment realizes adaptive control based on enzyme activity prediction value through the cooperative optimization of reinforcement learning agent and fuzzy PID controller. The fuzzy PID controller processes the deviation relationship between the prediction value and the target value, and the reinforcement learning agent optimizes the PID parameter adjustment strategy according to the long-term control effect. The combination of the two retains the rapid response characteristics of fuzzy control and the continuous optimization ability of reinforcement learning. The finally generated temperature-rotation speed adjustment instruction guarantees the process precision while considering the equipment operation stability, forming a closed-loop control system.

[0047] The embodiment realizes accurate regulation of enzyme activity prediction value through the fuzzy PID controller optimized by reinforcement learning. The membership function of the fuzzy PID controller ensures reasonable fuzzy processing of enzyme activity deviation. The reinforcement learning agent optimizes the PID parameter adjustment strategy based on chlorophyll retention rate and energy consumption ratio, improving the adaptive ability of the control system. The defuzzification processing and feasibility check guarantee the accuracy and executability of the temperature-rotation speed adjustment instruction, and finally realize intelligent optimization control of fixation process parameters.

[0048] In some embodiments, a reinforcement learning agent is constructed, and the state space, action space and reward function of the reinforcement learning agent are configured. The state space includes the current PID parameters and control effect evaluation indicators, the action space is the adjustment amplitude of the PID parameters, and the reward function is calculated based on the chlorophyll retention rate and the energy consumption ratio, including: The state space is represented by formula (1), and formula (1) is as follows: ; In formula (1), is the state space, is the parameter combination of the fuzzy PID controller, is the proportional coefficient of the fuzzy PID controller, which controls the reaction intensity to the current error, is the integral coefficient of the fuzzy PID controller, which is used to eliminate steady-state error, is the differential coefficient of the fuzzy PID controller, which suppresses system oscillation, is the control effect evaluation indicator, is the chlorophyll retention rate deviation, is the energy consumption ratio deviation; The action space is represented by formula (2), and formula (2) is as follows: ; In formula (2), is the action space, is the adjustment amount of the proportional coefficient, , is the limit adjustment amplitude, is the adjustment amount of the integral coefficient, , is the limit adjustment amplitude, is the adjustment amount of the differential coefficient, , is the limit adjustment amplitude; The reward function is represented by formula (3), and formula (3) is as follows: ; In formula (3), is the reward function, is the weight coefficient of the chlorophyll retention rate, wherein , is the weight coefficient of the energy consumption efficiency, wherein , is the absolute value deviation of the chlorophyll retention rate, is the energy consumption ratio deviation truncated by the upper limit, ; The policy network and the value network of the reinforcement learning agent are initialized, and the training hyperparameters are configured, including the learning rate and the discount factor, wherein the learning rate controls the parameter update step, and the discount factor balances the immediate reward and the long-term return; The interaction information generated by the interaction between the reinforcement learning agent and the kill device is obtained, and the parameter correction strategy of the fuzzy PID controller is dynamically optimized according to the interaction information, including: The interaction information generated by the real-time interaction process between the reinforcement learning agent and the kill device is collected, and the interaction information includes the control state, the executed action, the obtained reward and the new control state, and the interaction information is expressed by formula (4) as , wherein is the state observation value at time , which includes , is the executed parameter adjustment action , is the immediate reward value, calculated according to the reward function , is the new state after executing the adjustment action; Based on the interaction information, an experience sample library is constructed, training data is extracted from the experience sample library by random sampling, and the time series difference error is calculated, which is expressed by formula (5) as: ; , wherein is the time series difference error, is the discount factor, used to balance the immediate reward and the long-term return, is the value network, used to evaluate the value of the current state, is the parameter of the value network; The parameters of the strategy network and the value network are updated by using the training data to generate a new parameter adjustment strategy, and the updated parameters of the value network are expressed by the formula , and the advantage function is further calculated and expressed by the following formula: ; Wherein, is the learning rate, is the current sampling batch, is the batch sampling index, is the time difference error of the sample in the batch, is the probability of the strategy network selecting action under state ; The parameters of the updated strategy network are expressed by the formula , wherein, is the strategy network, is the parameter of the strategy network; The new parameter adjustment strategy is applied to the parameter update of the fuzzy PID controller, and the parameter update includes proportional coefficient update, integral coefficient update and derivative coefficient update, and is expressed by the following formula (6): ; Wherein, is the proportional coefficient of the updated fuzzy PID controller, is the integral coefficient of the updated fuzzy PID controller, is the derivative coefficient of the updated fuzzy PID controller; The updated fuzzy PID controller is evaluated, and the fuzzy PID controller that passes the value evaluation is verified for stability to make the control parameters within a preset safe range; The new parameter adjustment strategy that passes the stability verification is output and executed.

[0049] In this embodiment, the state space is composed of the parameter combination of the fuzzy PID controller and the control effect evaluation index, and the control effect evaluation index characterizes the deviation degree of chlorophyll retention rate from the target value, reflects the deviation of actual energy consumption ratio from the expected value. The action space is defined as the adjustment range of the PID parameters, and the limit adjustment range of each parameter 、 、 is determined according to process sensitivity analysis.

[0050] The reward function is constructed by the weighted combination of chlorophyll retention rate deviation and energy consumption ratio deviation, and the weight coefficients and Satisfy the normalization constraint for balancing the dual goals of quality preservation and energy consumption optimization. Empirical sample library stores interaction information tuples , timing difference error The difference between the current reward and the state value evaluated by the value network, discount factor Adjust the proportion of long-term reward consideration.

[0051] Policy network Output probability distribution of parameter adjustment action, value network Evaluate state value. Parameter update process uses advantage function Perform policy gradient calculation through the current sampling batch Reduce training variance. The updated PID parameters are weighted and adjusted through the action selection probability of the policy network to ensure smooth transition of parameter changes. Stability verification is performed through pre-set safety range constraints to prevent system instability caused by parameter updates. Preferably, the pre-set safety range is determined by the allowable fluctuation range of process parameters, combined with equipment operating limits and product quality requirements, and is verified by experimental tests to ensure system stable operation.

[0052] The embodiment realizes adaptive adjustment of control parameters through iterative optimization of the reinforcement learning agent and the fuzzy PID controller. The state space completely characterizes the system control characteristics, the action space provides feasible parameter adjustment directions, and the reward function guides the optimization target. The empirical sample library and the time difference learning realize efficient use of experience, and the collaborative training of the policy network and the value network ensures the reliability of the parameter adjustment strategy. The finally formed closed-loop control system can dynamically optimize the PID parameters according to the process state, while ensuring the retention rate of chlorophyll and improving the energy utilization efficiency.

[0053] The embodiment realizes intelligent optimization of the fuzzy PID controller by constructing a state space containing PID parameters and control effect evaluation indicators, defining an action space for parameter adjustment amplitude, and defining a reward function based on chlorophyll retention rate and energy consumption ratio. The collaborative training mechanism of the policy network and the value network is adopted, combined with the empirical sample library and the time difference learning, to ensure the reliability and stability of the parameter adjustment strategy. Finally, an adaptive closed-loop control system is formed, which effectively optimizes the energy consumption efficiency while maintaining the chlorophyll retention rate.

[0054] In some embodiments, the enzyme activity prediction value sequence is input into the optimized fuzzy PID controller to calculate the fuzzy control output information of the temperature adjustment amount and the speed adjustment amount, including: Calculate the deviation of the enzyme activity prediction value sequence to obtain the enzyme activity prediction value deviation and the deviation change rate; Input the enzyme activity prediction value deviation and the deviation change rate into the fuzzification interface to convert them into corresponding fuzzy language variables; The fuzzy PID controller matches the applicable temperature adjustment fuzzy rules and the speed adjustment fuzzy rules; The matched temperature adjustment fuzzy rules and the speed adjustment fuzzy rules are subjected to inference operation to generate fuzzy control output information containing temperature adjustment amount and speed adjustment amount; The fuzzy control output information is subjected to defuzzification processing to generate accurate initial temperature-speed adjustment instructions, including: The temperature adjustment amount in the fuzzy control output information is calculated by the gravity method to obtain the initial temperature adjustment amount; The speed adjustment amount in the fuzzy control output information is calculated by the maximum membership degree method to obtain the initial speed adjustment amount; The initial temperature adjustment amount and the initial speed adjustment amount are combined according to the time sequence to generate the initial temperature-speed adjustment instructions.

[0055] In the embodiment, the enzyme activity prediction value sequence is time series data reflecting the change of enzyme activity in the fixation process, which is collected by a sensor in real time, the deviation is obtained by the difference between the current prediction value and the target value, and the deviation change rate represents the change trend of the deviation at adjacent time points. The fuzzification interface is used to convert continuous numerical variables into fuzzy language variables, and the conversion process is realized based on the preset membership function, which ensures the reasonable mapping of accurate quantities to fuzzy concepts.

[0056] The temperature adjustment fuzzy rules and the speed adjustment fuzzy rules are pre-set control knowledge bases, and each rule describes the fuzzy relationship between the input variable and the output variable in the form of "if-then". Preferably, the inference operation adopts the Mamdani fuzzy inference method, and the results of multiple applicable rules are integrated through fuzzy logical operation. The fuzzy control output information contains the fuzzy set of temperature adjustment amount and speed adjustment amount, and the output range matches the physical limit of the device actuator.

[0057] The defuzzification processing converts the fuzzy output into accurate control instructions, in which the temperature adjustment amount is calculated by the gravity method, and the gravity position of the area surrounded by the membership function curve and the abscissa is obtained; the speed adjustment amount is calculated by the maximum membership degree method, and the accurate value corresponding to the maximum membership point is directly selected. The initial temperature-speed adjustment instructions are combined by time sequence to ensure the time sequence coordination of the control action, and the time interval is determined according to the process response characteristics.

[0058] The implementation principle of the embodiment can be understood as follows: the enzyme activity prediction deviation and its change rate are converted into a continuously adjustable control instruction by a fuzzy PID controller. The fuzzy processing realizes the conversion of accurate quantity into fuzzy concept, the fuzzy reasoning simulates the experience of expert control, and the defuzzification ensures the executability of the control instruction. The coordinated control of temperature and rotating speed ensures the enzyme activity inhibition effect and maintains the stability of the fixation process. The whole control process forms a closed-loop regulation from prediction to execution, and realizes the intelligent optimization of the fixation process parameters.

[0059] The embodiment realizes the intelligent perception of the fixation process parameters through real-time deviation analysis of the enzyme activity prediction value sequence and accurate conversion of the fuzzy interface; ensures that the control decision meets the process requirements through the coordinated reasoning of the temperature regulation fuzzy rule and the rotating speed regulation fuzzy rule; generates accurate temperature-rotating speed regulation instructions through defuzzification processing, forms a closed-loop control system, effectively inhibits the enzyme activity while maintaining the process stability, and realizes the dynamic optimization control of the fixation process.

[0060] In some embodiments, the edge execution nodes of each temperature zone in the fixation chamber are controlled according to the temperature-rotating speed regulation instruction, and the edge execution nodes adjust the hot air temperature, the fixation cylinder rotating speed and the axial segmented wind speed, including: The temperature-rotating speed regulation instruction is decomposed into independent control instructions of each temperature zone, and the independent control instructions of each temperature zone include preheating zone temperature instructions, main fixation zone temperature instructions and stable zone temperature instructions; The preheating zone temperature instruction is subjected to signal conversion to generate a gas valve opening degree control signal to adjust the gas combustion intensity of the preheating zone; The main fixation zone temperature instruction is subjected to power modulation conversion to obtain the output power of the heating element to adjust the hot air temperature of the main fixation zone; The stable zone temperature instruction is subjected to PID operation to obtain the fan rotating speed regulation amount to adjust the temperature interval of the stable zone; And the temperature-rotating speed regulation instruction is input into a variable frequency drive to adjust the operating frequency of the drive unit of the fixation cylinder, including: According to the real-time feedback information of the load current of the drive unit, the output frequency of the variable frequency drive is fine-tuned; Actual temperature data of each temperature zone is collected to calculate the deviation value from the target temperature; The control instruction output value of each zone is dynamically corrected according to the temperature deviation value; The corrected control instruction is issued to the corresponding edge execution node for execution.

[0061] In this embodiment, the control strategy of the temperature zone edge execution node is further optimized by introducing a process parameter coupling analysis model to dynamically correlate the preheating zone gas valve opening degree with the main fixation zone heating power. When changes in the moisture content of the material are detected, the energy distribution ratio of the two zones is automatically coordinated. This ratio coefficient is dynamically calculated based on a regression model trained using historical process data. The temperature control of the main fixation zone uses an adaptive threshold mechanism. The allowable fluctuation range of the temperature is automatically adjusted based on the current material passing speed. When the material passes quickly, the threshold is appropriately relaxed to reduce energy waste. When the material passes slowly, the threshold is tightened to ensure fixation uniformity.

[0062] The self-tuning of the PID parameters in the stabilization zone is achieved through an online identification algorithm. The proportional coefficient, integral coefficient, and derivative coefficient are automatically updated based on the real-time collected temperature response curve characteristics. The adjustment amplitude of the proportional coefficient is positively correlated with the rate of change of the temperature deviation. The integral time constant is dynamically corrected based on the system lag characteristics. Furthermore, the frequency fine-tuning of the frequency converter driver introduces a load torque observer. The load torque variation of the fixation drum is indirectly estimated through harmonic analysis of the motor current. When a sudden increase in torque caused by material accumulation is detected, the frequency output is compensated in advance to prevent a drop in speed.

[0063] Preferably, the dynamic correction of the temperature deviation value uses a fuzzy prediction compensation strategy. Not only the current deviation value is considered, but also the deviation trend of the previous three control periods is predicted and compensated. The compensation weight is determined by the membership function. The response characteristic calibration of the edge execution node introduces digital twin technology. The execution delay characteristics under different working conditions are simulated through a virtual debugging platform to generate a device response time compensation parameter library. In actual control, the corresponding parameters are retrieved and used for advanced control according to the working conditions. Through the multi-level adaptive optimization mechanism, the fixation process control has stronger working condition adaptability and energy optimization potential. While ensuring the retention rate of chlorophyll, more precise energy consumption control is achieved.

[0064] In some embodiments, fixation endpoint determination information is collected, including chlorophyll fluorescence threshold, volatile organic matter ratio, and leaf surface temperature. When the fixation endpoint determination information meets the preset determination condition, the fixation is terminated, including: Real-time sampling of the chlorophyll fluorescence threshold is performed to obtain the current fluorescence intensity value. The current fluorescence intensity value is compared with the preset fluorescence threshold to calculate the fluorescence intensity decay rate. When the fluorescence intensity decay rate reaches the preset enzyme inactivation threshold, a first termination signal is obtained. In addition, the concentrations of various components in volatile organic matter are monitored in real time to calculate the volatile organic matter ratio, which includes the real-time concentration ratio of linalool and aldehyde. When the volatile organic matter ratio reaches the preset aroma balance threshold, a second termination signal is obtained. and, collecting the leaf surface temperature in real time, calculating the average value of the leaf surface temperature to obtain a leaf surface temperature average value; when the leaf surface temperature average value is lower than a preset surface temperature threshold value, obtaining a third termination signal; The method further comprises: when the first termination signal and the second termination signal are received simultaneously, triggering a kill-termination instruction immediately, and the execution of the kill-termination operation comprises: turning off the hot air supply system; maintaining the rotation of the drum for residual heat dissipation; turning on the cooling fan to accelerate the cooling; when only the first termination signal or the second termination signal is received, starting a delay determination program to detect the third termination signal within a preset time.

[0065] In the embodiment, the dynamic calibration of the chlorophyll fluorescence threshold introduces a machine learning optimization algorithm, and a fluorescence intensity-enzyme activity mapping model of different varieties of tea is established through historical process data training, so that the preset fluorescence threshold can be automatically fine-tuned according to the batch characteristics of the raw materials. The calculation of the fluorescence intensity decay rate uses an adaptive sliding window technology, and the window size is dynamically adjusted according to the current kill-green stage. A larger window is used to smooth the noise in the early kill-green stage, and a smaller window is used to improve the detection sensitivity in the critical stage.

[0066] The volatile organic matter ratio monitoring system integrates electronic nose technology, and distinguishes the characteristic peaks of linalool and chrysin through a pattern recognition algorithm. The preset aroma balance threshold is continuously optimized in combination with expert sensory evaluation data, forming a flavor control curve with self-learning ability. The average value calculation of the leaf surface temperature introduces a spatial weight coefficient, which gives different weights to the temperature measurement points at different positions of the drum, and focuses on monitoring the material accumulation area where temperature deviation is prone to occur.

[0067] Further, when only the first termination signal or the second termination signal is received, a delay determination program is started, and the preset time of the delay determination program uses fuzzy logic control, which comprehensively considers multiple factors such as the current kill-green progress, temperature change rate and material state, to realize intelligent adjustment of the determination time.

[0068] Preferably, the execution of the kill-termination operation introduces a gradient control strategy, the turning off of the hot air supply system is carried out in stages, the gas pressure is first reduced and then completely cut off, to avoid temperature sudden change affecting the quality; the rotation speed of the drum is intelligently adjusted according to the residual temperature distribution curve to ensure uniform residual heat dissipation; and the start and stop of the cooling fan is managed by a PID controller, and the air volume is dynamically adjusted according to the real-time cooling rate.

[0069] The embodiment introduces intelligent learning algorithm and adaptive control strategy, so that the end point determination system has process parameter self-optimization capability, improves adaptability to different raw materials and process conditions while ensuring determination accuracy, and realizes the best balance between quality control and energy consumption optimization.

[0070] In the second aspect, the embodiment also provides a tea leaf fixation process adjusting system suitable for the adjusting method in the first aspect. The adjusting system comprises a fixation device, a driving unit, a hot air assembly, a temperature sensor, an olfactory sensor, an image acquisition unit and a control unit. The fixation device comprises a fixation drum having a fixation chamber in which tea leaves to be fixed are contained. The driving unit is in transmission connection with the fixation drum. The hot air assembly comprises at least two fans which are distributed in the fixation drum in a preset mode. The temperature sensor is arranged in the fixation drum. The olfactory sensor is arranged in the fixation drum. The image acquisition unit is arranged in the fixation drum. The control unit is electrically connected with the driving unit, the hot air assembly, the temperature sensor, the olfactory sensor and the image acquisition unit respectively, and is used for executing the method in the first aspect.

[0071] In the embodiment, the fixation drum of the fixation device is preferably made of stainless steel and has a cylindrical structure, and a flow guide plate is arranged in the fixation chamber for optimizing the turning track of the tea leaves. The driving unit can be a variable frequency motor cooperating with a speed reduction mechanism for accurately controlling the rotation speed of the fixation drum, and the transmission connection is realized through a gear set or a belt pulley. The fans in the hot air assembly are high-temperature-resistant centrifugal fans which are distributed in the fixation drum in an axial symmetry mode to form a uniform hot air circulation field.

[0072] Preferably, the temperature sensor is an armored thermocouple array which is distributed along the axial direction and the circumferential direction of the fixation drum and is used for monitoring the temperature distribution of different regions in real time. The olfactory sensor is a gas detection module based on MEMS technology which can identify the component change of volatile organic compounds in the fixation process in real time. The image acquisition unit adopts a high-resolution industrial camera cooperating with a near-infrared light source and is used for capturing the surface morphology and color change of the tea leaves. The control unit is an embedded industrial control system which integrates a multi-channel signal acquisition card and a motion control card to realize the coordinated control of the execution units.

[0073] When the system is running, the fixation drum rotates at a constant speed under the driving of the driving unit, the hot air assembly generates an airflow with a controllable temperature, the temperature sensor, the olfactory sensor and the image acquisition unit collect process parameters in real time, and the control unit dynamically adjusts the hot air temperature and the rotation speed of the drum after comprehensively analyzing these parameters to realize accurate control of the fixation process. Through multi-sensor information fusion and closed-loop feedback mechanism, the system can automatically adapt to the fixation requirements of different varieties of tea leaves.

[0074] The embodiment integrates temperature, odor and visual multi-modal sensing technologies to build a comprehensive monitoring system for the fixation process, enabling the system to perceive changes in tea leaf status in all directions. The distributed layout of the hot air assembly combined with intelligent control strategies effectively solves the uneven temperature problem of traditional fixation equipment. The multi-parameter collaborative adjustment mechanism not only ensures the stability of fixation quality, but also improves energy utilization efficiency, providing an intelligent process solution for tea processing.

[0075] In a third aspect, the embodiment also provides a computer-readable storage medium having stored thereon computer program instructions, which when executed by a processor, implement the method of the first aspect.

[0076] The computer program involved in the embodiment can be stored in a computer device readable storage medium, including but not limited to magnetic disk, magnetic tape, magnetic card, floppy disk, flash memory, optical disc, optical card, read-only memory (ROM), random access memory (RAM), erasable programmable ROM (EPROM), and electrically erasable programmable ROM (EEPROM) and the like, and also includes other biological, physical or chemical structures that can realize similar or equivalent functions to the above-mentioned storage media, such as DNA, RNA, protein and other units with information storage ability, etc. In specific embodiments, the storage medium can be one of the above-mentioned medium types, or a combination of the above-mentioned medium types. In different embodiments, the computer program involved in the embodiment can be centrally stored in a single medium, or distributedly stored in multiple media. The memory containing the computer device readable storage medium can be a non-volatile memory or a random access memory. These computer device readable storage media can be built-in to the device, or connected to the device as an external device or part of an external device. In some embodiments, the memory with the computer device readable storage medium is deployed locally; in other embodiments, the memory can also be deployed remotely from the processor, such as network-attached storage accessed via RF circuitry or external port and communication network, where the communication network can be the Internet, one or more intranets, local area networks (LANs), wide area networks (WANs), storage area networks (SANs) and the like, or appropriate combinations thereof, as long as the computer device can access the memory. In addition, the computer program involved in the embodiment can be stored in plaintext / ciphertext form, or can be designed as training data, and integrated and reorganized by model training to be implicitly saved in the parameter state of a deep neural network or other machine learning model.

[0077] Compared with the prior art, the above technical scheme has the following beneficial effects: through real-time collection of multi-modal sensing information of infrared temperature time sequence information, chlorophyll fluorescence intensity and volatile organic matter concentration, a multi-dimensional feature space of enzyme activity attenuation curve, aroma volatile characteristic spectrum and temperature fluctuation characteristics is constructed, and real-time coupling analysis of enzyme activity dynamics and aroma substance generation in the fixation process is realized. Based on the multi-task learning architecture of the LSTM prediction model, the enzyme activity prediction value sequence, the chlorophyll degradation trend and the aroma generation potential are synchronously predicted, and the problem of lag in monitoring the dynamic change of enzyme activity in the traditional fixation process is solved. The fuzzy PID controller optimized by reinforcement learning dynamically updates the adjustment parameters through the reward function of chlorophyll retention rate and energy consumption ratio, and realizes intelligent generation of temperature-rotation speed regulation instructions. The edge execution node accurately adjusts the hot air temperature and the rotation speed of the fixation cylinder of each temperature zone according to the axial segmented wind speed control model, and effectively improves the uneven temperature distribution phenomenon of the traditional fixation. The end point determination system judges through the multi-condition fusion of chlorophyll fluorescence threshold, volatile organic matter ratio and leaf surface temperature, establishes the determination standard of enzyme activity inhibition and aroma quality optimization, and realizes the coordinated optimization of enzyme activity and aroma quality. The process chain stored by the block chain completes the record of the enzyme activity curve, the aroma quality score and the energy consumption distribution, and provides traceable data support for process optimization. The above technical scheme realizes the coordinated and accurate regulation and control of enzyme activity and aroma substances in the fixation process, improves the fixation quality stability, and significantly reduces the energy consumption.

[0078] The above only describes some embodiments of the present application, and does not limit the protection scope of the present application, and any equivalent device or equivalent process transformation using the content of the specification and drawings, or direct or indirect application in other related technical fields, are also included in the patent protection scope of the present application.

Claims

1. A method for adjusting the tea leaf fixing process, characterized in that: include: Collecting multimodal sensor information during the fixing process, wherein the multimodal sensor information includes infrared temperature time series information, chlorophyll fluorescence intensity, and volatile organic compound concentration; Extracting features from the multimodal sensing information to obtain multidimensional features, the multidimensional features including an enzyme activity decay curve, an aroma volatilization characteristic spectrum, and a temperature fluctuation characteristic; Dynamically construct an LSTM prediction model based on the multidimensional features and output a sequence of enzyme activity prediction values ​​for the prediction time segment; Inputting the enzyme activity prediction value sequence into a reinforcement learning optimized fuzzy PID controller to generate a temperature-speed adjustment instruction, wherein the adjustment parameters of the fuzzy PID controller are configured to be dynamically updated according to a reward function of chlorophyll retention rate and energy consumption ratio; According to the temperature-speed adjustment instruction, the edge execution nodes of each temperature zone in the fixing chamber are controlled, and the edge execution nodes adjust the hot air temperature, the fixing drum speed and the axial segmented wind speed; The fixing endpoint determination information is collected, and the fixing endpoint determination information includes a chlorophyll fluorescence threshold, a volatile organic compound ratio, and a leaf surface temperature. When the fixing endpoint determination information meets a preset determination condition, the fixing is terminated.

2. The tea leaf fixing process adjustment method according to claim 1, wherein: Feature extraction is performed on the multimodal sensing information to obtain multidimensional features, including: Perform sliding window segmentation on the infrared temperature time series information to obtain multiple temperature sequence windows; Perform spectrum analysis on the temperature data in each temperature sequence window, extract the main frequency component and the energy distribution characteristics corresponding to each main frequency component, and obtain the temperature fluctuation characteristics; Furthermore, digital filtering is performed on the chlorophyll fluorescence intensity to remove high-frequency noise and obtain smooth fluorescence intensity information; performing differential calculation on the smoothed fluorescence intensity information to obtain the instantaneous change rate of enzyme activity; The instantaneous change rate of the enzyme activity is time-fitted using a nonlinear least squares method to obtain an enzyme activity decay curve showing the change of fluorescence intensity over time; Real-time monitoring of the concentration of each component in the volatile organic compound and calculation of the dynamic concentration ratio of the preset aroma substance; Based on the dynamic concentration ratio, a transformation relationship model between the components of volatile organic compounds is established using time series regression analysis; Extracting dynamic change characteristics of aroma substances according to the transformation relationship model and constructing aroma volatility characteristic spectrum; The temperature fluctuation characteristics, enzyme activity decay curve and aroma volatilization characteristic spectrum are normalized to generate the multidimensional characteristics.

3. The tea leaf fixing process adjustment method according to claim 1, wherein: Dynamically construct an LSTM prediction model based on the multidimensional features, and output a sequence of enzyme activity prediction values ​​for the prediction time segment, including: Performing time-series alignment processing on the temperature fluctuation characteristics, enzyme activity decay curve and aroma volatility characteristic spectrum to construct a multi-dimensional feature input sequence; Input the multi-dimensional feature data collected in real time into the trained LSTM network model; Outputting a sequence of original enzyme activity prediction values ​​for the prediction time segment through forward propagation calculation of the LSTM network; Smoothing the original enzyme activity prediction value sequence to obtain the enzyme activity prediction value sequence; The LSTM prediction model is constructed through the following steps: Initialize the basic structure of the LSTM prediction model and configure the number of neurons in the input layer, hidden layer, and output layer; The sliding time window method is used to divide the sample training features of the same tea category into training sample sets; The LSTM prediction model is iteratively trained using the training sample set to optimize network weight parameters until the LSTM prediction model training is completed.

4. The tea leaf fixing process adjustment method according to claim 1, wherein The enzyme activity prediction value sequence is input into the reinforcement learning optimized fuzzy PID controller to generate a temperature-speed adjustment instruction, including: Establishing a membership function of input variables and a membership function of output variables of a fuzzy PID controller, wherein the input variables include the deviation and the deviation change rate of the enzyme activity prediction value sequence, and the output variables include the temperature adjustment amount and the speed adjustment amount; Construct a reinforcement learning agent and configure its state space, action space, and reward function. The state space includes the current PID parameters and control effect evaluation indicators. The action space is the adjustment range of the PID parameters. The reward function is calculated based on the chlorophyll retention rate and energy consumption ratio. Dynamically optimize the parameter correction strategy of the fuzzy PID controller based on the interactive information generated by the interaction between the reinforcement learning agent and the killing device; The enzyme activity prediction value sequence is input into the optimized fuzzy PID controller to calculate the fuzzy control output information of the temperature adjustment amount and the speed adjustment amount; Defuzzifying the fuzzy control output information to generate an accurate initial temperature-speed adjustment instruction; Performing feasibility verification on the initial temperature-speed adjustment instruction according to the current working state of the fixing chamber; The initial temperature-speed adjustment instruction that passes the verification is recorded as the temperature-speed adjustment instruction and output.

5. The tea leaf fixing process adjustment method according to claim 4, wherein: Construct a reinforcement learning agent and configure its state space, action space, and reward function. The state space includes the current PID parameters and control effect evaluation indicators. The action space is the adjustment range of the PID parameters. The reward function is calculated based on the chlorophyll retention rate and energy consumption ratio, including: The state space is represented by formula (1), which is as follows: ； In formula (1), is the state space, is the parameter combination of the fuzzy PID controller, is the proportional coefficient of the fuzzy PID controller, is the integral coefficient of the fuzzy PID controller, is the differential coefficient of the fuzzy PID controller, To evaluate the control effect, is the chlorophyll retention rate deviation, is the energy consumption ratio deviation; The action space is expressed by formula (2), which is as follows: ； In formula (2), is the action space, is the adjustment amount of the proportional coefficient, , for Limit adjustment amplitude, is the adjustment amount of the integral coefficient, , for Limit adjustment amplitude, is the adjustment amount of the differential coefficient, , for Limit adjustment amplitude; The reward function is expressed by formula (3), which is as follows: ； In formula (3), is the reward function, is the weight coefficient of chlorophyll retention rate, is the weight coefficient of energy efficiency, is the absolute value deviation of chlorophyll retention rate, To take the upper limit cut-off energy consumption ratio deviation, ; Initialize the policy network and value network of the reinforcement learning agent and configure training hyperparameters, including the learning rate and discount factor; The interaction information generated by the interaction between the reinforcement learning agent and the killing device, and the parameter correction strategy of the fuzzy PID controller dynamically optimized according to the interaction information, include: Collecting interaction information generated by the real-time interaction between the reinforcement learning agent and the control device, including control state, executed actions, rewards obtained, and new control state; Building an experience sample library based on the interaction information, and extracting training data from the experience sample library by random sampling; Using the training data, the parameter weights of the strategy network and the value network are updated to generate a new parameter adjustment strategy; Applying the new parameter adjustment strategy to parameter update of the fuzzy PID controller, wherein the parameter update includes proportional coefficient update, integral coefficient update and differential coefficient update; Perform value evaluation on the updated fuzzy PID controller, and perform stability verification on the fuzzy PID controller that has passed the value evaluation to ensure that the control parameters are within a preset safety range; The new parameter adjustment strategy that has passed stability verification is output and executed.

6. The tea leaf fixing process adjustment method according to claim 4, wherein: The enzyme activity prediction value sequence is input into the optimized fuzzy PID controller to calculate the fuzzy control output information of the temperature adjustment amount and the speed adjustment amount, including: performing deviation calculation on the enzyme activity prediction value sequence to obtain the enzyme activity prediction value deviation and the deviation change rate; Input the enzyme activity prediction value deviation and the deviation change rate into the fuzzification interface and convert them into corresponding fuzzy language variables; Matching applicable temperature regulation fuzzy rules and speed regulation fuzzy rules through the fuzzy PID controller; Perform inference operations on the matched temperature regulation fuzzy rules and speed regulation fuzzy rules to generate fuzzy control output information containing temperature regulation amount and speed regulation amount; Defuzzification is performed on the fuzzy control output information to generate an accurate initial temperature-speed adjustment instruction, including: The temperature adjustment amount in the fuzzy control output information is calculated using the center of gravity method to obtain the initial temperature adjustment amount; The speed regulation value in the fuzzy control output information is calculated using the maximum membership method to obtain the initial speed regulation value; The initial temperature adjustment amount and the initial speed adjustment amount are combined in a time series to generate an initial temperature-speed adjustment instruction.

7. The tea leaf fixing process adjustment method according to claim 1, wherein: According to the temperature-speed adjustment instruction, the edge execution nodes of each temperature zone in the fixing chamber are controlled, and the edge execution nodes adjust the hot air temperature, the fixing drum speed and the axial segmented wind speed, including: Decomposing the temperature-speed adjustment instruction into independent control instructions for each temperature zone, wherein the independent control instructions for each temperature zone include a preheating zone temperature instruction, a main fixing zone temperature instruction, and a stabilization zone temperature instruction; Performing signal conversion on the temperature command of the preheating zone to generate a gas valve opening control signal to adjust the gas combustion intensity in the preheating zone; Performing power modulation conversion on the temperature instruction of the main fixing zone to obtain the output power of the heating element to adjust the hot air temperature of the main fixing zone; Performing PID calculation on the temperature command of the stable zone to obtain a fan speed adjustment value to adjust the temperature range of the stable zone; Furthermore, the temperature-speed adjustment instruction is input into the variable frequency drive to adjust the operating frequency of the driving unit of the fixing cylinder.

8. The tea leaf fixing process adjustment method according to claim 1, wherein: Collecting the fixation endpoint determination information, wherein the fixation endpoint determination information includes the chlorophyll fluorescence threshold, the volatile organic compound ratio, and the leaf surface temperature, and terminating the fixation when the fixation endpoint determination information meets the preset determination conditions, including: Performing real-time sampling on the chlorophyll fluorescence threshold to obtain a current fluorescence intensity value; Comparing the current fluorescence intensity value with a preset fluorescence threshold value to calculate the fluorescence intensity decay rate; When the fluorescence intensity decay rate reaches a preset enzyme inactivation threshold, a first termination signal is obtained; and, monitoring the concentration of each component in the volatile organic compound in real time, and calculating a volatile organic compound ratio, wherein the volatile organic compound ratio includes a real-time concentration ratio of linalool to linalool; When the volatile organic compound ratio reaches a preset aroma balance threshold, a second termination signal is obtained; and, collecting the blade surface temperature in real time, calculating the average value of the blade surface temperature, and obtaining the blade surface temperature average value; When the average surface temperature of the blade is lower than a preset surface temperature threshold, a third termination signal is obtained; The method further comprises: When the first termination signal and the second termination signal are received simultaneously, the completion termination instruction is immediately triggered; When only the first termination signal or the second termination signal is received, the delay determination program is started to detect the third termination signal within a preset time.

9. A tea withering process regulating system, characterized in that: The adjustment method according to any one of claims 1 to 8, wherein the adjustment system comprises: The fixing equipment comprises a fixing cylinder having a fixing chamber, wherein the fixing chamber contains tea leaves to be fixed; A driving unit, connected to the fixing cylinder in a transmission manner; The hot air component includes at least two fans, and the at least two fans are distributed in the withering cylinder according to a preset manner; A temperature sensor is arranged in the withering cylinder; An olfactory sensor is arranged in the withering cylinder; An image acquisition unit is arranged in the fixing cylinder; A control unit is electrically connected to the driving unit, the hot air component, the temperature sensor, the olfactory sensor and the image acquisition unit respectively, and the control unit is used to execute the method described in any one of claims 1 to 8.

10. A computer-readable storage medium having computer program instructions stored thereon, characterized in that: The computer program instructions implement the method according to any one of claims 1 to 8 when executed by a processor.

Images