Multi-matrix intelligent fragrance diffusion system based on 4G-Biuoth networking and self-adaptive regulation and control method of multi-matrix intelligent fragrance diffusion system based on 4G-Biuoth networking

Abstract

The invention discloses a multi-matrix intelligent fragrance diffusion system based on 4G-Biuoth networking and a self-adaptive regulation and control method thereof, and relates to the technical field of intelligent environment control, and the multi-matrix intelligent fragrance diffusion system comprises a host unit, a regulation and control module and a plurality of slave units. According to the multi-matrix intelligent fragrance diffusion system of the 4G-Biuoth network and the self-adaptive regulation and control method of the multi-matrix intelligent fragrance diffusion system, wide-area coverage is realized through heterogeneous networking, and fragrance concentration and time domain distribution are dynamically regulated in combination with three-matrix driving, audio emotion mapping and DAI interaction technologies. According to the system, a multi-factor coupling model is adopted, people flow density, music characteristics and user input are fused, the intelligent level and the resource utilization rate of environment fragrance management are remarkably improved, and the system is suitable for refined fragrance control of large-scale commercial scenes.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent environmental control, and in particular to a 4G-Biuetooth networked multi-matrix intelligent fragrance diffusion system and an adaptive control method thereof. Background Art

[0002] Aroma diffuser refers to a device used to diffuse aroma or volatile substances. It usually diffuses fragrance, essential oils, disinfectants, etc. into the air through atomization, heating or natural volatilization to achieve the purpose of refreshing the air, regulating emotions, sterilization and disinfection.

[0003] However, traditional fragrance diffusion equipment has the following defects: (1) Defects of single fragrance control: Traditional aroma diffusers (such as CN202010123456.7) often use a single fragrance matrix, which cannot meet the fragrance switching needs of different time periods in commercial scenarios: morning, noon, and evening. When the fragrance needs to be changed, the fragrance canister must be manually replaced, which increases operation and maintenance costs and prevents dynamic adaptation.

[0004] (2) Network coverage and energy efficiency issues: Existing wireless aroma diffuser systems (such as US20210056789A1) mostly use Wi-Fi networking, which has the following problems: The slave devices are constantly online, resulting in excessive power consumption (measured >3W / node); network coverage is limited (a single AP supports a maximum of 32 nodes); the 4G module is only used for host communication, and a hierarchical network architecture is not established.

[0005] (3) Lack of cross-modal perception technology: Current music-responsive fragrance devices have technical limitations: Only a linear relationship between volume and fragrance concentration is established ; Unresolved audio spectrum features (such as low-frequency / mid-frequency / high-frequency energy distribution); Lack of dynamic mapping mechanism between music emotion model and fragrance parameters.

[0006] Insufficient user interaction capabilities: The existing APP control solution has interaction defects: it uses a fixed fragrance formula library and cannot support dynamic ratio adjustment; there is no priority management mechanism when the user's fragrance adjustment instructions conflict with the automatic control strategy; there is a lack of a three-dimensional visual fragrance adjustment interface, and the operation is not intuitive. Summary of the Invention

[0007] In response to the shortcomings of the existing technology, the present invention provides a multi-matrix intelligent fragrance diffusion system based on 4G-Biuetooth networking and its adaptive control method, which solves the problems of single fragrance control defects, network coverage and energy efficiency issues, lack of cross-modal perception technology and insufficient user interaction capabilities.

[0008] To achieve the above objectives, the present invention is implemented through the following technical solutions: a 4G-Biuetooth networked multi-matrix intelligent fragrance diffusion system, including a host unit, a control module and multiple slave units; The said comprises a 4G communication module, a central controller and an audio processing unit; The 4G communication module is used to receive the fragrance strategy instructions sent from the cloud; The central controller has a built-in dynamic control algorithm; The audio processing unit includes a high-fidelity microphone array and a digital signal processor for real-time acquisition and analysis of background music spectrum characteristics; The plurality of slave units are connected to the Bluetooth 5.2 Mesh network and include a three-matrix drive circuit and a multi-channel sensor group; The three-matrix drive circuit is used to independently control the piezoelectric atomizers and micro air pumps of the morning, midday, and evening fragrance jars; The multi-channel sensor group includes an infrared pyroelectric sensor and an odor sensor; The infrared pyroelectric sensor detects the flow of people per unit time. ; The MQ-3 series odor sensor detects the residual concentration of ambient fragrance ; The control module performs the following operations: Trigger spice matrix switching based on GPS time information; Calculate the final output power based on the multi-factor coupling model;

[0009] in, For the emotional factor of music, Adjust the fragrance weight for users, Dynamic optimization through LSTM neural network.

[0010] Preferably, the three-matrix driving circuit adopts a circular layout, with the morning fragrance matrix located at the center of the circle, and the noon and evening fragrance matrices symmetrically distributed at 120°, and the spacing between each matrix is ​​≥15mm to prevent cross-contamination of fragrance.

[0011] Preferably, the audio processing unit performs the following operations: Perform 256-point FFT transformation on the music signal at 44.1kHz sampling rate; Calculate the energy weight coefficient of each frequency band; ; Output from the pre-trained music sentiment classification model .

[0012] Preferably, the user fragrance weight Input through the 3D polar coordinate interface of the mobile APP, in the interface; The radial axis represents the total fragrance concentration; The angle axis is divided into three 120° fan-shaped areas, corresponding to the proportions of morning, mid-day and evening fragrance respectively; Users can adjust the proportion of each fragrance in real time by dragging the slider, and the system will automatically normalize it.

[0013] Preferably, the LSTM neural network optimization process includes: Input layer: historical crowd flow data , fragrance residue , music factor ; Hidden layer: 3 layers of LSTM units; Output layer: dynamic coefficients , the objective function is to minimize the mean square error between the set value and the measured value of the fragrance concentration.

[0014] The present invention also discloses an adaptive control method for a multi-matrix intelligent fragrance diffusion system of a 4G-Biuetooth network, comprising the following steps: S1, network initialization phase; First, the host obtains the cloud configuration parameters through the communication module and builds the Bluetooth Mesh network topology; The slave unit nodes are then automatically assigned network addresses according to the preset geocoding; S2, multimodal data collection stage; First, use infrared pyroelectric sensors to count the density of people in the target area ; Then use the odor sensor to detect the residual concentration of the fragrance ; Finally, the background music signal is collected through the microphone array to extract the energy proportion of low frequency, medium frequency and high frequency; S3, dynamic regulation stage; When the low-frequency energy of the music is detected to account for more than 40%, the fragrance concentration enhancement mode is activated and the output power is increased to 150% of the baseline value; According to the fragrance ratio set by the user , dynamically distribute the driving power of each matrix according to the formula; ; The air pump speed and the vibration frequency of the atomizer are adjusted by the PID controller to make the actual output concentration error ≤5%.

[0015] The present invention provides a 4G-Biuetooth networked multi-matrix intelligent fragrance diffusion system and its adaptive control method. Compared with the existing technology, it has the following advantages: The 4G-Bluetooth networked multi-matrix intelligent fragrance diffusion system and its adaptive control method, with heterogeneous networking (4G+Bluetooth Mesh), three-matrix drive, and multi-modal algorithm as the core, realize the precise, intelligent, and low-power management of the fragrance system, which is suitable for refined environmental control in large commercial scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 Schematic diagram of the system architecture of the present invention; Figure 2 This is a schematic diagram of the host unit architecture of the present invention; Figure 3 This is a schematic diagram of the slave unit architecture of the present invention; Figure 4 This is a schematic diagram of the multi-channel sensor group architecture of the present invention; Figure 5 This is a cross-sectional view of the internal structure of the slave unit of the present invention; Figure 6 This is a schematic diagram of the infrared-fragrance linkage of the present invention; Figure 7 The DAI fragrance adjustment interface of the present invention; Figure 8 Schematic diagram of the multimodal control state machine of the present invention.

[0017] In the figure: 1. Host unit; 11. 4G communication module; 12. Central controller; 13. Audio processing unit; 2. Slave unit; 21. Three-matrix drive circuit; 22. Multi-channel sensor group; 221. Infrared pyroelectric sensor; 222. Odor sensor; 3. Control module. DETAILED DESCRIPTION

[0018] The technical solutions in the embodiments of the present invention are clearly and completely described. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. All other embodiments derived by persons of ordinary skill in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0019] See also Figures 1-8 , an embodiment of the present invention provides a technical solution: a multi-matrix intelligent fragrance diffusion system of 4G-Biuetooth networking, comprising a host unit 1, a control module 3 and multiple slave units 2; The host unit 1 includes a 4G communication module 11, a central controller 12 and an audio processing unit 13; The 4G communication module 11 is used to receive fragrance strategy instructions issued by the cloud; The central controller 12 has a built-in dynamic control algorithm; The audio processing unit 13 includes a microphone array and a digital signal processor (DSP) for real-time acquisition and analysis of background music spectrum characteristics; Multiple slave units 2 are connected to the master unit 1 via a Bluetooth 5.2 Mesh network and include a three-matrix drive circuit 21 and a multi-channel sensor group 22; The three-matrix drive circuit 21 is used to independently control the piezoelectric atomizers and micro air pumps of the morning, noon, and evening fragrance jars; The multi-channel sensor group 22 includes an infrared pyroelectric sensor 221 and an odor sensor 222; Infrared pyroelectric sensor 221, detects the flow of people per unit time ; Odor sensor 222, detects the residual concentration of ambient fragrance .

[0020] Control module 3 performs the following operations: Trigger spice matrix switching based on GPS time information; Calculate the final output power based on the multi-factor coupling model;

[0021] in, For the emotional factor of music, Adjust the fragrance weight for users, Dynamic optimization through LSTM neural network.

[0022] Furthermore, the three-matrix driving circuit 21 adopts a circular layout, with the morning fragrance matrix located at the center of the circle, and the noon and evening fragrance matrices symmetrically distributed at 120°, and the spacing between each matrix is ​​≥15mm to prevent cross-contamination of fragrance.

[0023] Furthermore, the audio processing unit 13 performs the following operations: Perform 256-point FFT transformation on the music signal at 44.1kHz sampling rate; Calculate the energy weight coefficient of each frequency band; ; Output from the pre-trained music sentiment classification model (VGGish network) .

[0024] Furthermore, the user's fragrance weight Input through the 3D polar coordinate interface of the mobile APP, in the interface; The radial axis represents the total fragrance concentration; The angle axis is divided into three 120° fan-shaped areas, corresponding to the proportions of morning, mid-day and evening fragrances; Users can adjust the proportion of each fragrance in real time by dragging the slider, and the system will automatically normalize it.

[0025] Furthermore, the LSTM neural network optimization process includes: Input layer: historical crowd flow data , fragrance residue , music factor ; Hidden layer: 3 layers of LSTM units (128 neurons per layer); Output layer; dynamic coefficient , the objective function is to minimize the mean square error between the set value and the measured value of the fragrance concentration.

[0026] The present invention also discloses an adaptive control method for a multi-matrix intelligent fragrance diffusion system of a 4G-Biuetooth network, comprising the following steps: S1, network initialization phase; First, the host unit 1 obtains cloud configuration parameters through the 4G communication module 11 and builds a Bluetooth Mesh network topology; Then the slave unit 2 node automatically assigns a network address according to the preset geocoding; S2, multimodal data collection stage; First, the infrared pyroelectric sensor 221 is used to count the density of people in the target area. ; Then the scent sensor 222 is used to detect the residual concentration of the fragrance. ; Finally, the background music signal is collected through the microphone array to extract the energy proportion of low frequency, medium frequency and high frequency; S3, dynamic regulation stage; When the low-frequency energy of the music is detected to be greater than 40%, the digital signal processor (DSP) outputs =0.85, the central controller 12 triggers the fragrance enhancement mode, increases the output power to 150% of the baseline value and prioritizes the activation of the evening fragrance matrix; According to the fragrance ratio set by the user , dynamically distribute the driving power of each matrix according to the formula; ; The air pump speed and the vibration frequency of the atomizer are adjusted by the PID controller to make the actual output concentration error ≤5%.

[0027] 1. System Architecture Implementation The host unit 1 is composed of a 4G communication module 11, a central controller 12 and an audio processing unit 13; 4G communication module 11: uses Quectel EC25 module, supports LTE Cat4, and has a built-in SIM card slot (model: Nano-SIM).

[0028] Operating frequency bands: B1 / B3 / B5 / B7 / B8 / B20.

[0029] Data transmission rate: 150Mbps downlink, 50Mbps uplink.

[0030] Central controller 12: Chip model: STM32H743VIT6 (ARM Cortex-M7 core, main frequency 480MHz).

[0031] Memory configuration: 1MB Flash + 512KB SRAM.

[0032] Communication interface: connect to 4G module via SPI bus, and connect to Bluetooth host via UART interface.

[0033] Audio processing unit 13: Microphone array: 4 MP34DT05TR-M digital microphones (SNR = 94dB).

[0034] DSP chip: TI TMS320C5517 (processes FFT analysis of audio signals).

[0035] Sampling parameters: 44.1kHz sampling rate, 256-point Hanning window, frequency resolution 172Hz.

[0036] The slave unit 2 includes a three-matrix drive circuit 21, a multi-channel sensor group 22 and a Bluetooth Mesh networking module; Bluetooth Mesh Networking: Chip model: Nordic nRF52840 (supports Bluetooth 5.2 protocol).

[0037] Network topology: Friend node mode is used, and the maximum number of relay hops is 3.

[0038] Power consumption control: sleep current ≤ 1.5μA, working current ≤ 8mA (@0dBm transmit power).

[0039] Three matrix driving circuit 21: Spice jar capacity: Morning (50mL) / Noon (80mL) / Evening (50mL).

[0040] Atomization component: piezoelectric atomizer (resonant frequency 1.7MHz) + micro air pump (flow rate 0.8L / min).

[0041] Annular layout parameters: center spacing ≥15mm (measured cross-contamination rate <3%).

[0042] Multi-channel sensor group 22: Infrared pyroelectric sensor 221: Panasonic AMN34111 (detection angle 110°, response time 0.3s).

[0043] Odor sensor 222: Model MQ-3 (detection range 10-1000ppm, warm-up time ≥24h).

[0044] 2. Implementation of Adaptive Control Methods Step 1: Network initialization 1. The host unit 1 obtains the network configuration parameters (JSON format) sent from the cloud through the 4G communication module 11.

[0045] 2. Parse geocoding rules (for example: slave address = region code (2 bytes) + device ID (1 byte)).

[0046] 3. Establish a Bluetooth Mesh network, set the host unit 1 as the Provisioner node and the slave unit 2 as the LowPower Node.

[0047] Step 2: Multimodal data acquisition Traffic flow detection: Count the number of times the infrared pyroelectric sensor 221 is triggered every 5 seconds , calculate the density; ; Audio feature extraction: The audio processing unit 13 performs pre-emphasis processing on the signal (coefficient α=0.97).

[0048] (2) Calculate the 13-dimensional Mel-frequency cepstral coefficients (MFCC, taking the first 13 dimensions) of the audio processing unit.

[0049] Output sentiment classification results (classification labels: soothing / active / neutral) through the pre-trained VGGish network.

[0050] Step 3: Dynamically control execution The control module 3 calculates the output power according to the following formula: Basic power calculation: ; (k=0.6, calibrated through experiments).

[0051] Multi-factor fusion: When the music emotion is "active", the enhancement mode is activated: ; ( The value ranges from 0 to 1 and is calculated by weighting the spectrum energy output by the audio processing unit 13).

[0052] 3. Implementation of key modules 1. Three-matrix drive circuit 21 drive control Ring layout anti-pollution design: Calculation of distance between adjacent spice jars; ; Drive timing control: adopt time-sharing multiplexing mechanism (morning matrix duty cycle duty cycle ≤ 40%).

[0053] 2. Audio-Fragrance Coupling Implementation Spectral energy weight training: (1) Dataset: GTZAN music dataset (10 genres, 1000 samples) (2) The training process code is as follows: model=Sequential() model.add(LSTM(128,input_shape=(30,13)))#Input 30 frames of MFCC features model.add(Dense(3,activation='softmax'))#output three categories model.compile(optimizer='adam',loss='categorical_crossentropy').

[0054] (3) Weight distribution: α=0.6, β=0.3, γ=0.1 (test set accuracy 82.3%).

[0055] DAI fragrance interaction implementation Control module 3 provides a three-dimensional slider control: Polar coordinate system parameters: Radial range: 0%~100% (step accuracy 1%); Angle division: morning (0°-120°), noon (120°-240°), evening (240°-360°); Hybrid algorithm tolerance processing: , Δ is the PID adjustment amount.

[0056] Example 1: Shopping mall promotion scenario Environmental parameters: Crowd density ; Background music: low-frequency energy accounts for 45% (“active” emotion label detected).

[0057] System Response: 1. Control module 3 activates the fragrance enhancement mode, and the output power is increased to ; 2. The three-matrix driving circuit 21 adjusts the evening fragrance ratio to 70% (dynamically adjusts μ to 0.7 through LSTM); 3. Increase the air pump speed to 8500rpm and adjust the atomizer frequency to 1.9MHz.

[0058] Example 2: User-defined fragrance Operation process: 1. The user sets the morning:noon:evening time in the APP to 0.2:0.5:0.3; 2. Control module 3 automatically shuts down the morning matrix and distributes power to the noon matrix; ; 3. The ambient concentration is monitored in real time by the odor sensor 222, and PID adjustment is triggered (proportional coefficient Kp=0.4) when the detection deviation is greater than 5%.

[0059] The experimental data are shown in the following table; 1. Concentration control accuracy comparison

[0060] 2. Power consumption test results

[0061] The present invention has the following effects through the above embodiments: 1. Precision Control Performance Improvement Through multi-sensor fusion (infrared + odor sensor) and closed-loop PID algorithm, the fragrance concentration error is reduced by 62% compared with traditional systems (static scene error ≤5.2%, dynamic peak scene ≤12.7%).

[0062] The multi-factor coupling model (people flow / music / interaction factors) achieves dynamic compensation, with the response deviation in music mutation scenarios less than 15.3% (compared to 41% in traditional systems).

[0063] Timing adaptability The three-matrix ring layout supports millisecond-level switching of morning / midnight / evening fragrances (switching time <50ms), meeting the multi-time requirements of commercial scenarios. The GPS time synchronization error is ≤1 second, ensuring consistency of regional fragrance strategies.

[0064] 2. Energy Efficiency and Resource Optimization Energy-saving features Bluetooth Mesh networking reduces slave standby power consumption to as low as 0.48W (traditional Wi-Fi solutions ≥3W), and the 4G host's on-demand wake-up mechanism reduces communication energy consumption (average daily power consumption is reduced by 68%).

[0065] Spice utilization The dynamic adjustment algorithm saves more than 30% of fragrance consumption (measured data: under the same traffic flow, the traditional system consumes 150mL / day, while the system of the present invention consumes ≤105mL / day). The resonant frequency of the atomizer is adaptively adjusted (1.5-2.1MHz), and the atomization efficiency is improved by 40%.

[0066] 3. Cross-modal collaborative innovation Music-Fragrance Emotional Mapping Based on MFCC+VGGish audio feature analysis, the emotion matching accuracy reaches 82% (test data set: GTZAN). When the low-frequency energy accounts for >40%, the promotion mode is automatically triggered (the concentration is increased to 150%, and the evening fragrance accounted for +30%).

[0067] Multimodal priority management Supports seamless switching between automatic / music / manual modes (state machine response time <200ms), and the conflict resolution mechanism ensures that the deviation of the user-set weight Wuser is ≤5%.

[0068] 4. Commercial Application Value Scalability A single host can manage 256 slaves (the upper limit of the Bluetooth 5.2 protocol), covering over 10,000 square meters of commercial space. The spice jars are modularly designed, and the replacement time is less than 30 seconds (traditional equipment ≥5 minutes).

[0069] Enhanced user experience The DAI perfume app provides a three-dimensional visual interface (polar coordinate slider), with an operation response time of <300ms, and supports users to create exclusive fragrance brand logos (can store ≥100 sets of personalized formulas).

[0070] In summary, this invention takes heterogeneous networking (4G+Bluetooth Mesh), three-matrix drive, and multimodal algorithm as its core to achieve precise, intelligent, and low-power management of the fragrance system, which is suitable for refined environmental control in large-scale commercial scenarios.

[0071] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0072] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A 4G-Biuetooth networked multi-matrix intelligent fragrance diffusion system, characterized by: It comprises a host unit (1), a control module (3) and a plurality of slave units (2); The host unit (1) includes a 4G communication module (11), a central controller (12) and an audio processing unit (13); The 4G communication module (11) is used to receive the fragrance strategy instructions sent from the cloud; The central controller (12) has a built-in dynamic control algorithm; The audio processing unit (13) includes a microphone array and a digital signal processor (DSP) for real-time acquisition and analysis of background music spectrum characteristics; The plurality of slave units (2) are connected to the master unit (1) via a Bluetooth 5.2 Mesh network and include a three-matrix drive circuit (21) and a multi-channel sensor group (22); The three-matrix driving circuit (21) is used to independently control the piezoelectric atomizers and micro air pumps of the morning, midday and evening fragrance jars; The multi-channel sensor group (22) includes an infrared pyroelectric sensor (221) and an odor sensor (222); The infrared pyroelectric sensor (221) detects the flow of people per unit time. ; The odor sensor (222) detects the residual concentration of the ambient fragrance ; The control module (3) performs the following operations: Trigger spice matrix switching based on GPS time information; Calculate the final output power based on the multi-factor coupling model; ； in, For the emotional factor of music, Adjust the fragrance weight for users, Dynamic optimization through LSTM neural network.

2. The 4G-Biuetooth networked multi-matrix intelligent fragrance diffusion system according to claim 1, characterized in that: The three-matrix driving circuit (21) adopts a circular layout, with the morning fragrance matrix located at the center of the circle, and the noon and evening fragrance matrices symmetrically distributed at 120 degrees, and the spacing between each matrix is ​​≥15mm to prevent cross contamination of fragrance.

3. The 4G-Biuetooth networked multi-matrix intelligent fragrance diffusion system according to claim 1, characterized in that: The audio processing unit (13) performs the following operations: Perform 256-point FFT transformation on the music signal at 44.1kHz sampling rate; Calculate the energy weight coefficient of each frequency band; ; Output by pre-trained music emotion classification model (VGGish network) .

4. The 4G-Biuetooth networked multi-matrix intelligent fragrance diffusion system according to claim 1, characterized in that: The user's fragrance adjustment weight Input through the 3D polar coordinate interface of the mobile APP, in the interface; The radial axis represents the total fragrance concentration; The angle axis is divided into three 120° fan-shaped areas, corresponding to the proportions of morning, mid-day and evening fragrances; Users can adjust the proportion of each fragrance in real time by dragging the slider, and the system will automatically normalize it.

5. The multi-matrix intelligent fragrance diffusion system of 4G-Biuetooth networking according to claim 1, characterized in that: The LSTM neural network optimization process includes: Input layer: historical crowd flow data , fragrance residue , music factor ; Hidden layer: 3 layers of LSTM units (128 neurons per layer); Output layer; dynamic coefficient , the objective function is to minimize the mean square error between the set value and the measured value of the fragrance concentration.

6. The adaptive control method of a 4G-Biuetooth networked multi-matrix intelligent fragrance diffusion system according to any one of claims 1 to 5, characterized in that: The following steps are involved: S1, network initialization phase; First, the host obtains the cloud configuration parameters through the communication module and builds the Bluetooth Mesh network topology; The slave nodes are then automatically assigned network addresses according to the preset geocoding; S2, multimodal data collection stage; First, use infrared sensors to count the density of people in the target area ; Then use the odor sensor to detect the residual concentration of the fragrance ; Finally, the background music signal is collected through the microphone array to extract the energy proportion of low frequency, medium frequency and high frequency; S3, dynamic regulation stage; When the low-frequency energy of the music is detected to account for more than 40%, the fragrance concentration enhancement mode is activated and the output power is increased to 150% of the baseline value; According to the fragrance ratio set by the user , dynamically distribute the driving power of each matrix according to the formula; The air pump speed and the vibration frequency of the atomizer are adjusted by the PID controller to make the actual output concentration error ≤5%.

Images