Automatic monitoring and key process node distinguishing system and method for fine manipulation of oolong tea, withering of white tea, fermentation of black tea and scenting fragrance of scented tea
By combining differential channels with multi-stage adsorption tubes, automatic monitoring of aroma and identification of key process nodes during tea processing are achieved, solving the problem of relying on human experience in tea processing and improving the accuracy of identification and the reliability of the system.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- FUJIAN AGRI & FORESTRY UNIV
- Filing Date
- 2026-01-23
- Publication Date
- 2026-04-21
AI Technical Summary
In the current tea processing, the judgment of key processes for the formation of tea aroma relies on the experience of tea masters, which is highly subjective and difficult to quantify. Traditional methods are time-consuming and easily affected by environmental interference. Single sensor solutions lack selectivity, and adsorption tubes are prone to saturation, contamination, and blockage. There is also a lack of online monitoring and replacement prompts.
The differential fingerprinting technology using the CH0 control channel and the CH1-CH3 differential channels, combined with multi-stage adsorption tubes and photoionization detectors, enables automatic monitoring and identification of key process nodes in oolong tea processing, white tea withering, black tea fermentation, and scenting of sachets through mode selection and parameter set switching. The status of the adsorption tubes is monitored through a purification window and flow/differential pressure sensors.
It improves the robustness of tea processing identification, reduces engineering deployment costs, realizes automated identification and standardized control of key process nodes, and reduces misjudgment and maintenance difficulty.
Smart Images
Figure CN121899240A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of online detection and intelligent discrimination of volatile organic compounds in tea processing, specifically involving an automatic monitoring and key process node discrimination system and method for aroma of oolong tea during withering, white tea during withering, black tea during fermentation, and scented tea during scenting. Background Technology
[0002] Withering, fermentation, and scenting (for scented tea) are all crucial processes in tea processing, significantly influenced by raw material batches, temperature and humidity, ventilation, and equipment conditions. The aroma changes during withering of oolong tea and fermentation of black tea, and the scenting process of scented tea involve a dynamic balance between floral release and tea leaf adsorption, all of which rely heavily on the experience and judgment of tea masters. Traditionally, this relies on tea masters' comprehensive judgment through "smelling, observing, and touching the leaves," which is highly subjective and difficult to quantify and trace. While existing laboratory analytical methods can resolve components, they are time-consuming and difficult to apply online; single-TVOC or single-sensor solutions lack selectivity and are easily affected by background gases and humidity. Furthermore, the adsorption tube is a key component of differential detection, susceptible to saturation, contamination, and moisture blockage; the lack of online monitoring and replacement prompts can lead to long-term operational drift and misjudgments. Summary of the Invention
[0003] The purpose of this invention is to address the problems existing in the background technology by providing an automatic monitoring system and method for aroma detection and key process node identification in the processes of oolong tea processing, white tea withering, black tea fermentation, and scenting tea. The system employs a CH0 control channel and CH1-CH3 differential channels to form a differential fingerprint through "homogeneous comparison." Differential channels are configured with weak / medium / strong (or selective) adsorption tubes according to adsorption strength to achieve differential response to different VOC ranges. In the algorithm module, at least four sets of parameter sets Pz / Pw / Pf / Pj are preset. The device structure remains unchanged; the identification of stages and key process nodes / endpoints in oolong tea processing, white tea withering, black tea fermentation, and scenting tea is completed through "mode selection → parameter set switching." A purification window is set up: gas is supplied by a zero-gas module and switched via valve group to allow the system to perform a "recovery capacity test" under controllable conditions to confirm adsorption tube saturation / contamination; simultaneously, flow rate / pressure difference is used to identify blockage / wet blockage.
[0004] To achieve the above objectives, the technical solution of this invention is: an automatic monitoring and key process node discrimination system for the aroma of oolong tea processing, white tea withering, black tea fermentation, and scented tea, comprising a gas inlet, a filter unit, a gas pump, a multi-stage flow limiting and diversion module, and a gas path diversion manifold connected in sequence; the gas path diversion manifold connects four parallel detection channels CH0~CH3, where CH0 is the control channel and CH1~CH3 are differential channels; each detection channel includes a corresponding adsorption tube and a photoionization detector (PID) to obtain four-channel signals C0~C3; it also includes a multi-channel acquisition / signal aggregation module, a data processing and discrimination unit, a process mode selection / parameter set management module, a control interface, a zero-gas filtration / purified gas module, an electronically controlled valve group, a temperature and humidity sensor, a flow / differential pressure sensor, and a display / storage unit; the multi-channel acquisition / signal aggregation module synchronously acquires and aggregates the outputs of the four PIDs to form four-channel time-series data, which is then sent to the data processing and discrimination unit; the process mode selection / parameter set management module... The system includes a set of parameters for tea processing (Pz), withering (Pw), fermentation (Pf), and scenting (Pj), and provides corresponding preprocessing windows, feature weights, thresholds, and discrimination windows to the data processing and discrimination unit. The data processing and discrimination unit performs baseline correction, drift compensation, and temperature and humidity compensation on the four-channel signals C0~C3, calculates differential features, and, under constraints of Pz, Pw, Pf, or Pj, identifies the stages and key process nodes for oolong tea processing, white tea withering, black tea fermentation, and scenting of oolong tea. Endpoint determination; the zero-gas filtration / purification module is connected to the gas path manifold via an electronically controlled valve assembly to form a purification window for adsorption tube health confirmation and gas path self-test / recovery; the flow / differential pressure sensor is used to output the flow rate Qk or differential pressure ΔPk of each detection channel to distinguish between blockage / wet blockage and adsorption tube saturation contamination; the display / storage unit is used to display and store the four-channel curves, differential characteristics, determination results, and adsorption tube health status, and store the corresponding timestamp data; the control interface is used to output prompts or linkage control signals.
[0005] Furthermore, the adsorption tube (5-0) of the control channel CH0 is an empty tube or an inert equal flow resistance tube, used to provide an environmental baseline signal C0 under the condition of uniform flow resistance; the differential channels CH1~CH3 are respectively configured with weak adsorption tubes, medium adsorption tubes, and strong adsorption or selective adsorption tubes to form differential response signals C1~C3 for different volatile organic compounds; wherein, the adsorption material of the weak adsorption tube is Tenax TA or Tenax GR, the adsorption material of the medium adsorption tube is Carbograph material or a Tenax and Carbograph composite material, and the adsorption material of the strong adsorption or selective adsorption tube is Carboxen material or activated carbon material.
[0006] Furthermore, the multi-level flow limiting and diversion module includes main channel flow limiting and branch channel flow limiting, so that the four detection channels can work stably within the preset flow range, and the flow deviation of the four detection channels is not greater than the preset threshold.
[0007] Furthermore, the differential features include at least:
[0008] Difference of integral: ΔI_k=∫(Ck−C0)dt;
[0009] Difference slope: S_k = d(Ck−C0) / dt;
[0010] Delay difference: T_k=t_k(α)−t_0(α), where t_k(α) is the α-proportional threshold time when the channel signal Ck reaches its peak value, t_0(α) is the α-proportional threshold time when the channel signal C0 reaches its peak value, and k=1,2,3;
[0011] And the consistency between the channel ratio and the channel sorting.
[0012] Furthermore, the data processing and discrimination unit constructs the greening index Z, withering index W, fermentation index F, and scenting index J, respectively. Z, W, F, and J are weighted combinations of differential features and process parameters under the constraints of corresponding parameter sets Pz, Pw, Pf, and Pj, and are used to discriminate the greening endpoint, withering endpoint, fermentation endpoint, and key nodes / endpoints of scenting, respectively.
[0013] Furthermore, the control interface outputs "Continue shaking / Continue resting / Stop shaking" in the withering mode; "Continue withering / Adjust temperature and humidity or air volume / Turning or spreading for adjustment / End withering" in the withering mode; "Continue fermentation / Adjust temperature and humidity or ventilation / Turning or adjusting temperature and humidity or ventilation / End fermentation" in the fermentation mode; and "Continue scenting / Turning or adjusting temperature and humidity or ventilation / Initiating flowering / Replenishing flowering or re-scenting / End scenting" in the scenting mode.
[0014] Furthermore, the data processing and discrimination unit performs adsorption tube health monitoring, including a two-level triggering mechanism: a first-level early warning calculates health indicators and triggers an early warning within a continuous sliding window; a second-level confirmation calculates the baseline Bk and recovery time Trec under clean window conditions; when Bk increases or Trec extends beyond a threshold, a replacement prompt is output; when Qk decreases or ΔPk increases beyond a threshold, a maintenance prompt is output.
[0015] This invention also provides a method for automatic monitoring and key process node identification of aroma in oolong tea processing, white tea withering, black tea fermentation, and scenting tea using any of the above-described systems, including:
[0016] S1. Process mode selection: Select the greening parameter set Pz, withering parameter set Pw, fermentation parameter set Pf, or scenting parameter set Pj;
[0017] S2. Sampling and differential detection: The processing environment gas enters the gas distribution manifold through the gas inlet, filter unit, gas pump and multi-stage flow limiting and distribution module and is distributed to CH0~CH3. C0 is obtained from the control channel and C1~C3 are obtained from the differential detection through the adsorption tube.
[0018] S3. Synchronous Acquisition and Preprocessing: Synchronously acquire C0~C3 and perform baseline correction, drift compensation and temperature and humidity compensation;
[0019] S4. Calculation of differential characteristics: Calculate the differential integral difference ΔI_k, differential slope S_k, delay difference T_k, channel ratio and sorting consistency;
[0020] S5. Key process node identification: Based on differential features, calculate the greening index Z, withering index W, fermentation index F or scenting index J under the constraints of Pz, Pw, Pf or Pj respectively, and output the stage identification and key node / end point identification results, while outputting prompts or linkage control signals.
[0021] S6. Adsorption tube health monitoring: Implements a first-level early warning and a second-level confirmation through the purification window, and outputs a replacement or maintenance prompt.
[0022] Furthermore, the greening index Z, withering index W, fermentation index F, and scenting index J satisfy the following:
[0023] Z=Σ i a_i·φ_i
[0024] W=Σ j b_j·ψ_j
[0025] F=Σ m c_m·η_m
[0026] J=Σ n d_n·ξ_n
[0027] Where i, j, m, and n represent the indices of the combined feature terms in the corresponding index, φ_i, ψ_j, η_m, and ξ_n are the combined features of the differential integral difference, differential slope, time delay difference, sorting consistency, and process parameters, respectively; a_i, b_j, c_m, and d_n are provided by the corresponding parameter sets Pz, Pw, Pf, and Pj.
[0028] Furthermore, the purification window is formed by switching the electric control valve group to the zero gas filtration / purified gas module (9) for gas supply, and the duration of the purification window is a preset duration; during the purification window period, the baseline Bk and recovery time Trec are calculated to eliminate the influence of environmental fluctuations.
[0029] Compared with the prior art, the present invention has the following beneficial effects:
[0030] (1) Control + multiple adsorption intensity differential channels improve the robustness of discrimination.
[0031] (2) The same device is suitable for multiple key processes such as withering, fermentation and scenting, and the engineering deployment cost is low.
[0032] (3) The purification window confirmation + flow / differential pressure assistance can reduce false alarms and facilitate maintenance.
[0033] (4) The results are traceable, which facilitates process standardization and linkage control. Attached Figure Description
[0034] Figure 1 This is a schematic diagram of the system structure of the present invention.
[0035] Figure 2 This is a flowchart of the method of the present invention.
[0036] Figure 3 This is a schematic diagram of the two-stage triggering logic for adsorption tube health monitoring.
[0037] In the diagram: 1-Gas inlet; 2-Filter unit; 3-Air pump; 4-Multi-stage flow limiting and diversion module; 4-1-Gas path diversion manifold; 5-0~5-3-Adsorption tube; 6-0~6-3-PID; 7-Data processing and discrimination unit; 7-1-Multi-channel acquisition / signal aggregation module; 8-Control interface; 9-Zero gas filtration / purified gas module; 10-Electrically controlled valve group; 11-Temperature and humidity sensor; 12-Flow / differential pressure sensor; 13-Display / storage module; 15-Process mode selection / parameter set management module. Detailed Implementation
[0038] The technical solution of the present invention will now be described in detail with reference to the accompanying drawings.
[0039] like Figure 1As shown, this invention provides an automatic monitoring and key process node discrimination system for the aroma of oolong tea during withering, white tea during withering, black tea during fermentation, and scented tea during aroma processing. The system includes a gas inlet 1, a filter unit 2, an air pump 3, a multi-stage flow limiting and diversion module 4, and a gas path manifold 4-1 connected in sequence. The gas path manifold 4-1 connects four parallel detection channels CH0~CH3, where CH0 is the control channel and CH1~CH3 are differential channels. Each detection channel includes corresponding adsorption tubes 5-0~5-3 and photoionization detectors PID 6-0~6-3 to obtain four channel signals C0~C3. The system also includes a multi-channel acquisition / signal aggregation module 7-1, a data processing and discrimination unit 7, a process mode selection / parameter set management module 15, a control interface 8, a zero-gas filtration / purified gas module 9, an electronically controlled valve group 10, a temperature and humidity sensor 11, a flow / differential pressure sensor 12, and a display / storage unit 13. The multi-channel acquisition / signal aggregation module 7-1 monitors the four PID... The outputs from 6-0 to 6-3 are synchronously acquired and aggregated to form four-channel time-series data, which is then sent to the data processing and discrimination unit 7. The process mode selection / parameter set management module 15 includes at least the withering parameter set Pz, the fermentation parameter set Pw, the fermentation parameter set Pf, and the scenting parameter set Pj, and provides the data processing and discrimination unit 7 with corresponding preprocessing windows, feature weights, thresholds, and discrimination windows. The data processing and discrimination unit 7 performs baseline correction, drift compensation, and temperature and humidity compensation on the four-channel signals C0 to C3, calculates differential features, and completes the withering of oolong tea and white tea under the constraints of Pz, Pw, Pf, or Pj, respectively. The system identifies the stages of withering, black tea fermentation, and scenting of tea, and determines key process nodes / endpoints; the zero-gas filtration / purification module 9 is connected to the gas path manifold 4-1 via the electronically controlled valve group 10 to form a purification window for confirming the health of the adsorption tube and for gas path self-checking / recovery; the flow / differential pressure sensor 12 outputs the flow rate Qk or differential pressure ΔPk of each detection channel to distinguish between blockage / wet blockage and adsorption tube saturation contamination; the display / storage unit 13 displays and stores the four-channel curves, differential characteristics, discrimination results, and adsorption tube health status, and stores the corresponding timestamp data; the control interface 8 outputs prompts or linkage control signals.
[0040] like Figure 2 As shown, the present invention also provides a method for automatic monitoring and key process node identification of aroma in oolong tea processing, white tea withering, black tea fermentation, and scenting tea using any of the above-described systems, including:
[0041] S1. Process mode selection: Select the greening parameter set Pz, withering parameter set Pw, fermentation parameter set Pf, or scenting parameter set Pj;
[0042] S2. Sampling and differential detection: The processing environment gas enters the gas path manifold 4-1 through gas inlet 1, filter unit 2, air pump 3 and multi-stage flow limiting and diversion module 4 and is then distributed to CH0~CH3. C0 is obtained from the control channel and C1~C3 is obtained from the differential detection through the adsorption tube.
[0043] S3. Synchronous Acquisition and Preprocessing: Synchronously acquire C0~C3 and perform baseline correction, drift compensation and temperature and humidity compensation;
[0044] S4. Calculation of differential characteristics: Calculate the differential integral difference ΔI_k, differential slope S_k, delay difference T_k, channel ratio and sorting consistency;
[0045] S5. Key process node identification: Based on differential features, calculate the greening index Z, withering index W, fermentation index F or scenting index J under the constraints of Pz, Pw, Pf or Pj respectively, and output the stage identification and key process node / end point identification results, while outputting prompts or linkage control signals.
[0046] S6. Adsorption tube health monitoring: Implements a first-level early warning and a second-level confirmation through the purification window, and outputs a replacement or maintenance prompt.
[0047] The following are specific implementation examples of the present invention.
[0048] Example 1: Four-channel differential structure
[0049] The gas enters the gas manifold after passing through the inlet, filter, gas pump, and flow-limiting diversion, and is then divided into four paths. CH0 is equipped with an empty tube / inert equal flow resistance tube as the baseline; CH1 to CH3 are equipped with weak / medium / strong (or selective) adsorption tubes in sequence, and output C0 to C3 respectively.
[0050] Example 2: Adsorption tube type and adaptation logic (example)
[0051] Weak adsorption tubes are suitable for low molecular weight / low boiling point VOCs that respond quickly and are easily desorbed; medium adsorption tubes are suitable for medium polar VOCs such as alcohols and esters; strong adsorption or selective adsorption tubes are suitable for aromatic or more polar VOCs.
[0052] By using a gradient configuration of "weak-medium-strong", comparable differential curves can be generated for the same batch of gas under different adsorption intensities.
[0053] Example 3: Calculation of Differential Features
[0054] Using the control channel as the baseline, calculate the differential signal Dk=Ck−C0.
[0055] The integral difference ΔI_k = ∫Dk dt, the differential slope S_k = dDk / dt, the time delay difference T_k = t_k(α) − t_0(α), and the channel sorting consistency are calculated. Among them, T_k is used to characterize the time difference of "adsorption-re-release", which is often used to distinguish the changes in the proportion of different volatile components.
[0056] Example 4: Four Process Differentiation Outputs (Same Framework, Different Parameter Sets)
[0057] Pz (Tooth Turning Mode): Outputs "Continue shaking / Continue resting / / Adjust temperature, humidity or airflow / Stop shaking".
[0058] Withering mode Pw: Outputs "Continue withering / Adjust temperature, humidity or air volume / Turn over or spread out to adjust / End withering".
[0059] Fermentation mode Pf: Outputs "Continue fermentation / Adjust temperature and humidity or ventilation / Turn over or adjust thickness / End fermentation".
[0060] Scenting mode Pj: Outputs "Continue scenting / Turn over or adjust temperature and humidity or ventilation / Remove flower residue (remove flower residue) / Replenish flowers or rescent / End scenting".
[0061] Example 5: Figure 3 As shown, the adsorption tube replacement cycle is related to the health monitoring algorithm (core).
[0062] Where Rk represents the gas response result or intermediate feature quantity (such as the differential sensor output, normalized response value, or feature after integration / slope processing) obtained by the k-th differential channel within a preset time window, and k is the differential channel number.
[0063] Level 1 Early Warning (Online): Calculate health indicators for each differential channel and monitor them continuously, for example:
[0064] (1) Recovery / stabilization index: baseline offset Bk before and after the clean window;
[0065] (2) Recovery time: The time required for the signal to fall back to the threshold after entering the clean window (Trec);
[0066] (3) Differential collapse: Whether the long-term difference between the differential channel and the control channel gradually disappears (indicating saturation);
[0067] (4) Sorting consistency anomaly: The relative relationship between weak / medium / strong channels is distorted over a long period of time (indicating contamination / abnormality).
[0068] An alert is triggered when an indicator exceeds the threshold for N consecutive sliding windows.
[0069] Secondary confirmation (clean air window): The valve group switches to zero gas to form a clean air window, and calculates Bk and Trec under controllable conditions. If Bk continues to rise or Trec continues to lengthen, it is determined that the adsorption tube is approaching saturation / contamination, and a replacement prompt is output.
[0070] Clog / wet blockage differentiation: Combine flow rate Qk or pressure difference ΔPk. If Qk decreases or ΔPk increases and persists, it is primarily identified as clog / wet blockage, and a maintenance prompt (cleaning, drying, or replacement) is output.
[0071] The above are preferred embodiments of the present invention. Any changes made to the technical solution of the present invention that do not exceed the scope of the technical solution of the present invention shall fall within the protection scope of the present invention.
Claims
1. A system for automatically monitoring and identifying key process nodes of aroma during the withering of oolong tea, the withering of white tea, the fermentation of black tea, and the scenting of flower tea, characterized in that, The system includes a gas inlet (1), a filter unit (2), a gas pump (3), a multi-stage flow limiting and diversion module (4), and a gas path manifold (4-1) connected in sequence. The gas path manifold (4-1) connects four parallel detection channels CH0~CH3, where CH0 is the control channel and CH1~CH3 are the differential channels. Each detection channel includes a corresponding adsorption tube (5-0~5-3) and a photoionization detector PID (6-0~6-3) to obtain four-channel signals C0~C3. The system also includes a multi-channel acquisition / signal aggregation module (7-1), a data processing and discrimination unit (7), a process mode selection / parameter set management module (15), a control interface (8), a zero gas filtration / purified gas module (9), an electronically controlled valve group (10), a temperature and humidity sensor (11), a flow / differential pressure sensor (12), and a display. / Storage unit (13); Multi-channel acquisition / signal aggregation module (7-1) synchronously acquires and aggregates the outputs of four PIDs (6-0~6-3) to form four-channel time-series data and sends it to the data processing and discrimination unit (7); Process mode selection / parameter set management module (15) includes at least the greening parameter set Pz, withering parameter set Pw, fermentation parameter set Pf and scenting parameter set Pj, and provides the data processing and discrimination unit (7) with corresponding preprocessing windows, feature weights, thresholds and discrimination windows; The data processing and discrimination unit (7) performs baseline correction, drift compensation and temperature and humidity compensation on the four-channel signals C0~C3, calculates differential features and completes the stage identification and key process node / end point discrimination of oolong tea greening, white tea withering, black tea fermentation and flower tea scenting under the constraints of Pz, Pw, Pf or Pj respectively; The zero-gas filtration / purification module (9) is connected to the gas path manifold (4-1) via the electronically controlled valve group (10) to form a purification window for confirming the health of the adsorption tube and for gas path self-testing / recovery; the flow / differential pressure sensor (12) is used to output the flow rate Qk or differential pressure ΔPk of each detection channel to distinguish between blockage / wet blockage and adsorption tube saturation pollution; the display / storage unit (13) is used to display and store the four-channel curves, differential characteristics, discrimination results and adsorption tube health status, and store the corresponding timestamp data; the control interface (8) is used to output prompts or linkage control signals.
2. The automatic monitoring and key process node discrimination system for oolong tea processing, white tea withering, black tea fermentation, and scented tea aroma processing as described in claim 1, is characterized in that, The adsorption tube (5-0) of the control channel CH0 is an empty tube or an inert equal flow resistance tube, used to provide an environmental baseline signal C0 under the condition of uniform flow resistance; the differential channels CH1~CH3 are respectively configured with weak adsorption tubes (5-1), medium adsorption tubes (5-2), and strong adsorption or selective adsorption tubes (5-3) to form differential response signals C1~C3 for different volatile organic compounds; wherein, the adsorption material of the weak adsorption tube (5-1) is Tenax TA or Tenax GR, the adsorption material of the medium adsorption tube (5-2) is Carbograph material or a Tenax and Carbograph composite material, and the adsorption material of the strong adsorption or selective adsorption tube (5-3) is Carboxen material or activated carbon material.
3. The automatic monitoring and key process node discrimination system for oolong tea processing, white tea withering, black tea fermentation, and scented tea aroma processing as described in claim 1, is characterized in that... The multi-level flow limiting and diversion module (4) includes main channel flow limiting and branch channel flow limiting, so that the four detection channels can work stably within the preset flow range and the flow deviation of the four detection channels is not greater than the preset threshold.
4. The automatic monitoring and key process node discrimination system for oolong tea processing, white tea withering, black tea fermentation, and scented tea aroma processing according to claim 1, characterized in that, The difference features include at least: Difference of integral: ΔI_k=∫(Ck−C0)dt; Difference slope: S_k = d(Ck−C0) / dt; Delay difference: T_k=t_k(α)−t_0(α), where t_k(α) is the α-proportional threshold time when the channel signal Ck reaches its peak value, t_0(α) is the α-proportional threshold time when the channel signal C0 reaches its peak value, and k=1,2,3; And the consistency between the channel ratio and the channel sorting.
5. The automatic monitoring and key process node discrimination system for oolong tea processing, white tea withering, black tea fermentation, and scented tea aroma processing according to claim 1, characterized in that, The data processing and discrimination unit (7) constructs the greening index Z, withering index W, fermentation index F and scenting index J respectively. Z, W, F and J are weighted combinations of differential features and process parameters under the constraints of corresponding parameter sets Pz, Pw, Pf and Pj, and are used to discriminate the greening endpoint, withering endpoint, fermentation endpoint and key process nodes / endpoints of scenting respectively.
6. The automatic monitoring and key process node discrimination system for oolong tea processing, white tea withering, black tea fermentation, and scented tea aroma processing according to claim 1, characterized in that, The control interface (8) outputs "Continue shaking / Continue to stand / Adjust temperature and humidity or ventilation / Stop shaking" in the withering mode; "Continue withering / Adjust temperature and humidity or air volume / Turning or spreading for adjustment / End withering" in the withering mode; "Continue fermentation / Adjust temperature and humidity or ventilation / Turning for adjustment / End fermentation" in the fermentation mode; and "Continue scenting / Turning or adjusting temperature and humidity or ventilation / Start flowering / Replenish flowering or rescenting / End scenting" in the scenting mode.
7. The automatic monitoring and key process node discrimination system for oolong tea processing, white tea withering, black tea fermentation, and scented tea aroma processing according to claim 1, characterized in that, The data processing and discrimination unit (7) performs adsorption tube health monitoring, including two-level triggering: "early warning-confirmation"; the first-level early warning calculates health indicators and triggers an early warning within a continuous sliding window. The secondary verification calculates the baseline Bk and recovery time Trec under cleanroom conditions; when Bk increases or Trec extends beyond the threshold, a replacement prompt is output. A maintenance prompt is output when Qk decreases or ΔPk increases beyond the threshold.
8. A method for automatic monitoring and key process node identification of aroma in oolong tea processing, white tea withering, black tea fermentation, and scenting tea using the system described in any one of claims 1-7, characterized in that, include: S1. Process mode selection: Select the greening parameter set Pz, withering parameter set Pw, fermentation parameter set Pf, or scenting parameter set Pj; S2. Sampling and differential detection: The processing environment gas enters the gas path manifold (4-1) through the gas inlet (1), filter unit (2), air pump (3) and multi-stage flow limiting and diversion module (4) and is then distributed to CH0~CH3. C0 is obtained from the control channel, and C1~C3 are obtained from the differential detection through the adsorption tube. S3. Synchronous Acquisition and Preprocessing: Synchronously acquire C0~C3 and perform baseline correction, drift compensation and temperature and humidity compensation; S4. Calculation of differential characteristics: Calculate the differential integral difference ΔI_k, differential slope S_k, time delay difference T_k, channel ratio and sorting consistency; S5. Key process node identification: Based on differential features, calculate the greening index Z, withering index W, fermentation index F or scenting index J under the constraints of Pz, Pw, Pf or Pj respectively, and output the stage identification and key process node / end point identification results, while outputting prompts or linkage control signals. S6. Adsorption tube health monitoring: Implements a first-level early warning and a second-level confirmation through the purification window, and outputs a replacement or maintenance prompt.
9. The method for automatic monitoring and key process node identification of aroma in oolong tea processing, white tea withering, black tea fermentation, and scenting tea according to claim 8, characterized in that, The greening index Z, withering index W, fermentation index F, and scenting index J satisfy the following: Z=Σ i a_i·φ_i W=Σ j b_j·ψ_j F=Σ m c_m·h_m J=S n d_n·ξ_n Where i, j, m, and n represent the indices of the combined feature terms in the corresponding index, φ_i, ψ_j, η_m, and ξ_n are the combined features of the differential integral difference, differential slope, time delay difference, sorting consistency, and process parameters, respectively; a_i, b_j, c_m, and d_n are provided by the corresponding parameter sets Pz, Pw, Pf, and Pj.
10. The method for automatic monitoring and key process node identification of oolong tea processing, white tea withering, black tea fermentation, and scented tea aroma processing according to claim 8, characterized in that, The purification window is formed by switching the electric control valve group (10) to the zero gas filtration / purified gas module (9) for gas supply. The duration of the purification window is a preset duration. During the purification window, the baseline Bk and recovery time Trec are calculated to eliminate the influence of environmental fluctuations.