Mist outlet control method of nano atomizer, nano atomizer and aromatherapy machine
By acquiring the parameters of the atomizer components, constructing a preset model, and combining PID and MPC control, the problems of large particle size and instability in traditional atomizers are solved, achieving nanoscale atomization and stable mist output.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-01-28
- Publication Date
- 2026-04-14
Smart Images
Figure CN121847385A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of intelligent control technology, specifically to a method for controlling the mist output of a nano-atomizer, a nano-atomizer, and an aromatherapy diffuser. Background Technology
[0002] Traditional atomizers, such as ultrasonic atomizers and piezoelectric atomizers, typically produce coarser particle sizes for various reasons. These include uneven distribution of vibration energy, such as the single-frequency vibration of piezoelectric ceramics forming a fixed standing wave within the resonant cavity, leading to concentrated liquid film rupture points and the generation of large particles; droplet merging effect, such as excessive surface tension of the liquid film or low flow velocity, causing droplets to collide and merge before detaching from the nozzle, resulting in increased particle size; and mismatch of structural parameters, such as a mismatch between the resonant cavity dimensions (length L, diameter D) and the vibration frequency f, which prevents the excitation of efficient liquid film breakup modes (such as the Rayleigh instability optimal mode).
[0003] In related technologies, the output particle size of atomizers is usually not fine due to various reasons. This application aims to solve how to achieve a finer output particle size of atomizers, such as nanometer scale, and how to ensure the stability of the output particle size. Summary of the Invention
[0004] The main objective of this invention is to provide a method for controlling the mist output of a nano-atomizer, a nano-atomizer, and an aromatherapy diffuser, in order to address the shortcomings of related technologies.
[0005] To achieve the above objectives, according to a first aspect of the present invention, a method for controlling the mist output of a nano-atomizer is provided, comprising: obtaining specified parameters from different components of the atomizer, including obtaining a driving frequency from orthogonally arranged dual-voltage ceramic plates; obtaining an amplitude from a driving module of the dual-voltage ceramic plates; obtaining a micro-pump flow rate from a piezoelectric ceramic micro-pump; determining a heating power based on a micro-heating plate module; determining a predicted atomization particle size based on the specified parameters as input to a preset model constructed based on specified physical properties of the liquid to be atomized; determining an optimal control quantity based on the future value and the predicted atomization particle size; determining a PID control quantity based on a given expected atomization particle size and the difference between the obtained adjustable parameters of the liquid to be atomized; and controlling the mist output based on the optimal control quantity and the PID control quantity.
[0006] Optionally, the preset model is: ,in, The surface tension of the atomized liquid is affected by temperature. , The surface tension of the liquid to be atomized at a specified temperature. For temperature coefficient, The density of the liquid to be atomized is affected by temperature. The flow rate of the liquid to be atomized is determined based on the micropump flow rate. The dynamic viscosity of the liquid to be atomized The length of the atomizer resonant cavity. For liquid film thickness, , Here is the conversion factor, and A is the amplitude. The coefficient is dimensionless. d is the driving frequency of the dual piezoelectric ceramics, and d is the predicted atomized particle size.
[0007] Optionally, determining the control quantity of PID control based on the given expected value of atomized particle size and the difference between the obtained adjustable parameters of the liquid to be atomized includes: determining the expected value of the corresponding adjustable parameter based on the given expected value of atomized particle size; and obtaining the adjustable parameters of the liquid to be atomized obtained by actual detection and the expected value of the adjustable parameters to determine the control quantity of PID control.
[0008] Optionally, determining the optimal control quantity based on future values and predicted atomized particle size includes: based on a pre-constructed discrete-time state-space model, minimizing the comprehensive error function with the predicted atomized particle size, flow rate, and temperature for the next N steps as the optimization objective, wherein preset physical constraints are satisfied during the optimization process, and the optimal control quantity is obtained by optimization solution.
[0009] Optionally, after adjustment by the PID controller, the state-space model is updated, and optimization is performed again based on the updated state-space model with the predicted values of atomized particle size, flow rate, and temperature for the next N steps as the optimization objective, minimizing the comprehensive error function.
[0010] Optionally, the pre-constructed discrete-time state-space model is as follows: ; in, , These are the current particle size, flow rate, temperature, liquid film thickness, dual frequency, and amplitude, respectively. for Adjustments to the dual frequencies, amplitude, and power respectively; For the output vector, For process noise, To observe the noise, For a matrix, These are time-varying parameters.
[0011] Optionally, the error function is: in, These are the weighting coefficients for particle size, flow rate, and temperature, respectively. Let λ be the change in the input, and λ be the smoothing weight of the control input. For predicting the time domain, d ref Qref T ref , respectively, are the predicted values of particle size, flow rate, and temperature, and d(k), Q(k), and T(k) are the current values of particle size, flow rate, and temperature, respectively.
[0012] According to a second aspect of the present invention, a nano-atomizer is provided, comprising: orthogonally arranged dual-voltage ceramic plates, a driving module for the dual-voltage ceramic plates, a piezoelectric ceramic micropump, a micro heating element module, and a control module; wherein the control module obtains specified parameters from different components of the atomizer, including obtaining a driving frequency from the orthogonally arranged dual-voltage ceramic plates; obtaining an amplitude from the driving module for the dual-voltage ceramic plates; obtaining a micropump flow rate from the piezoelectric ceramic micropump; determining a heating power based on the micro heating element module; determining a predicted atomization particle size based on the specified parameters as input to a preset model constructed based on specified physical properties of the liquid to be atomized; determining an optimal control quantity based on the future value and the predicted atomization particle size; determining a PID control quantity based on a given expected atomization particle size and the difference between the obtained adjustable parameters of the liquid to be atomized; and controlling the atomization based on the optimal control quantity and the PID control quantity.
[0013] According to a third aspect of the present invention, an aroma diffuser is provided, comprising the nano-atomizer described in the second aspect implementation.
[0014] According to a fourth aspect of the present invention, a control module is provided that acquires specified parameters from different components of an atomizer, including acquiring a driving frequency from orthogonally arranged dual-voltage ceramic plates; acquiring an amplitude from a driving module of the dual-voltage ceramic plates; acquiring a micropump flow rate from a piezoelectric ceramic micropump; determining a heating power based on a micro heating element module; determining a predicted atomization particle size based on the specified parameters as input to a preset model constructed based on specified physical properties of the liquid to be atomized; determining an optimal control quantity based on future values and the predicted atomization particle size; determining a PID control quantity based on a given expected atomization particle size value and the difference between the acquired adjustable parameters of the liquid to be atomized; and controlling the atomization output based on the optimal control quantity and the PID control quantity.
[0015] This embodiment describes a method for controlling the mist output of a nano-atomizer, a nano-atomizer, and an aromatherapy diffuser. The method includes obtaining specified parameters from different components of the atomizer, including obtaining the driving frequency from orthogonally arranged dual-voltage ceramic plates; obtaining the amplitude from the driving module of the dual-voltage ceramic plates; obtaining the micro-pump flow rate from the piezoelectric ceramic micro-pump; determining the heating power based on a micro-heating element module; determining the predicted atomization particle size based on the specified parameters as input to a preset model constructed based on specified physical properties of the liquid to be atomized; determining the optimal control quantity based on the future value and the predicted atomization particle size; determining the control quantity for PID control based on the difference between the given expected atomization particle size and the obtained adjustable parameters of the liquid to be atomized; and controlling the mist output based on the optimal control quantity and the control quantity of the PID control. This dual-control approach enables the nano-atomizer to output mist in a stable manner with extremely fine particle size. Attached Figure Description
[0016] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.
[0017] Figure 1 This is a flowchart of the mist control method for the nano-atomizer according to an embodiment of the present invention; Figure 2 This is a schematic diagram of the composition and structure of the nano-atomizer according to an embodiment of the present invention. Detailed Implementation
[0018] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.
[0019] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate for the embodiments of the invention described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.
[0020] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.
[0021] According to embodiments of the present invention, a method is provided, such as Figure 1 As shown, the steps include the following.
[0022] Step 101: Obtain specified parameters from different components of the atomizer, including obtaining the driving frequency from the orthogonally arranged dual-voltage ceramic plates; obtaining the amplitude from the driving module of the dual-voltage ceramic plates; obtaining the micro-pump flow rate from the piezoelectric ceramic micro-pump; and determining the heating power based on the micro heating element module. Through numerous experimental studies, it was found that the coarsening of particle size in traditional atomizers (such as ultrasonic atomization and piezoelectric atomization) is mainly due to the uneven distribution of vibration energy. The single-frequency vibration of the piezoelectric ceramic forms a fixed standing wave in the resonant cavity, resulting in concentrated liquid film rupture locations and the generation of large particles. The droplet merging effect occurs when the surface tension of the liquid film is too high or the flow rate is too low, causing droplets to collide and merge before leaving the nozzle, thus increasing the particle size. The mismatch of structural parameters, such as the mismatch between the resonant cavity size (length L, diameter D) and the vibration frequency f, makes it impossible to excite efficient liquid film breakup modes (such as the Rayleigh instability optimal mode).
[0023] The stability of the mist output is affected by vibration frequency drift, liquid flow fluctuations, and temperature changes. Temperature changes (adjusted by a micro-heating element module) affect various parameters. This embodiment can be addressed through dual closed-loop feedback control, where the outer loop is a particle size closed loop, using a laser particle size analyzer to measure the particle size d in real time. meas The inner ring is a joint control of vibration parameters such as frequency f, amplitude A, and micropump flow rate Q; a piezoelectric ceramic micropump is used, and the control algorithm can compensate for changes in liquid viscosity and maintain a constant volumetric flow rate Q.
[0024] Two piezoelectric ceramic plates are orthogonally arranged on both sides of the resonant cavity. A sine wave / square wave signal (frequency f1, f2) is output by a high-frequency driving power supply. The amplitude of the driving voltage determines the deformation of the piezoelectric ceramic (i.e., the vibration amplitude A). The driving power supply has a built-in amplitude controller (such as a PWM-based voltage regulation module), which controls the amplitude A by changing the effective value (Vrms) of the output voltage.
[0025] Piezoelectric ceramic micropumps achieve precise flow rate regulation by periodically deforming a piezoelectric actuator to compress the liquid within the pump chamber. The flow rate Q is related to the actuator's vibration frequency and amplitude, as well as the pump chamber's geometric parameters (such as chamber volume and valve opening). The micropump is integrated into the liquid supply pipeline, with its inlet connected to a storage tank and its outlet connected to a resonant cavity. A flow servo controller uses a PID algorithm to adjust the piezoelectric actuator's drive signal frequency f_pump and voltage V_pump to compensate for changes in liquid viscosity (such as temperature fluctuations) or pipeline resistance, maintaining a constant volumetric flow rate Q.
[0026] The heating element heats the liquid through the Joule heating effect (P = I²R, where I is the current and R is the resistance), thereby reducing the viscosity (μ decreases as temperature increases), preventing droplet coalescence or decreased breakup efficiency due to excessive viscosity. The heating element is attached to the outer wall of the liquid storage tank or the inner wall of the resonant cavity (requiring an insulating heat insulation layer for protection), and the liquid temperature T is monitored in real time by a temperature sensor. The temperature PID controller adjusts the driving current I of the heating element based on the error between the set temperature T_ref and the measured T, thereby controlling the heating power P (P = V×I, where V is the supply voltage).
[0027] In this implementation, two piezoelectric ceramic plates (frequency f1, f2) are orthogonally arranged to excite the superposition of multiple standing waves within the resonant cavity, forming a uniform distribution of liquid film shear force. Multi-frequency vibration breaks the nodal limitations of a single standing wave, dispersing the liquid film rupture locations and reducing the generation of large particles. A tapered-expanding microchannel (contraction ratio ε = 3-5) is set at the nozzle tip, generating controllable turbulence through geometrical abrupt changes in the channel to promote initial droplet breakup; the expanding section reduces the flow velocity, minimizing secondary collisions and merging. The liquid film thickness h is adjusted by controlling the amplitude A of the piezoelectric ceramics, making h close to the Rayleigh instability optimal thickness to maximize droplet breakup efficiency. Step 102: Based on the specified parameters as input to the preset model constructed based on the specified physical properties of the liquid to be atomized, determine the predicted atomization particle size; and determine the optimal control quantity based on the future value and the predicted atomization particle size; and determine the control quantity of PID control based on the given expected atomization particle size and the difference between the obtained adjustable parameters of the liquid to be atomized.
[0028] Determining the optimal control quantity based on future values and predicted atomized particle size belongs to MPC control, i.e., model predictive control. Determining the control quantity of PID control based on the difference between the given expected atomized particle size and the adjustable parameters of the liquid to be atomized belongs to PID control.
[0029] As an optional implementation of this embodiment, the preset model is: ,in, The surface tension of the atomized liquid is affected by temperature. , The surface tension of the liquid to be atomized at a specified temperature. For temperature coefficient, The density of the liquid to be atomized is affected by temperature. The atomized liquid detachment velocity is determined based on the micro-pump flow rate. The dynamic viscosity of the liquid to be atomized The length of the atomizer resonant cavity. For liquid film thickness, , Here is the conversion factor, and A is the amplitude. The coefficient is dimensionless. d is the driving frequency of the dual piezoelectric ceramics, and d is the predicted atomized particle size.
[0030] In this optional implementation, temperature T affects multiple physical quantities during atomization, including surface tension σ, dynamic viscosity μ, liquid density ρ, and even vibration frequency or amplitude. Therefore, T should be a key input parameter in the model, influencing these sub-items. f1 is the driving frequency of the main piezoelectric ceramic, which excites the fundamental mode standing wave of the resonant cavity; f2 is the driving frequency of the auxiliary piezoelectric ceramic, which orthogonally vibrates and excites higher-order standing waves; A is the vibration amplitude, which determines the initial disturbance intensity of the liquid film; the micropump supply flow rate is the in-situ pump flow rate, Q directly affects the flow velocity, v=Q / A. 流道 . The surface tension-dominant coefficient and the viscous resistance-dominant coefficient, respectively, depend on the nozzle structure, such as the nozzle diameter and the multi-frequency vibration coupling coefficient, and are determined by the phase difference between the two frequencies.
[0031] Surface tension σ(T) decreases with increasing temperature and can usually be fitted using linear or polynomial equations, such as... ,in It is the surface tension at a reference temperature T0 (e.g., 25℃), k TIt is the coefficient of surface tension as a function of temperature. Dynamic viscosity μ(T) decreases with increasing temperature, μ(T) = μ0·exp(E / (R(T - T_ref))), where E is the activation energy, R is the gas constant, and T_ref is the reference temperature. Density ρ(T) changes less with temperature, but for liquids it can be approximated as ρ(T) = ρ0[1 - β(T - T0)], where β is the volume expansion coefficient. v (droplet detachment velocity) is also temperature-dependent because temperature affects the viscosity and surface tension of the liquid, thus affecting its flow characteristics. Changes in liquid density ρ affect the effective liquid film height. σ(T), μ(T), and ρ(T) change with T, thus affecting the final particle size d.
[0032] The model transforms input parameters from the control system—amplitude A, micropump flow rate Q, heating power P, and dual frequencies f1 / f2—into key variables such as liquid film thickness h, liquid viscosity μ, surface tension σ, and vibration frequency. Ultimately, the output particle size d is determined by the mathematical expression of the particle size model. For example, decreasing σ (e.g., by heating) or increasing v (increasing flow rate Q) can decrease d; decreasing μ (by heating) or thinning h (by increasing amplitude A) can decrease d; increasing the nozzle length L prolongs the shear time, thus increasing d. The superposition of dual-frequency standing waves disrupts fixed nodes (preventing concentrated rupture of the liquid film at a fixed location), and multi-frequency vibration coupling terms... Making the droplet size distribution more uniform, for example, by increasing the frequency ratio f2 / f1 > 1.2, will enhance the randomness of the standing wave, leading to a reduction in droplet size.
[0033] In summary, the atomization particle size model parameters (σ, μ, ρ) all describe the continuous liquid phase state before atomization; the inputs of the control system (A, Q, P, fi) drive the liquid film instability and rupture process by changing these parameters (such as reducing σ and μ by increasing heating power P); the core optimization path is to reduce σ and μ (increase temperature), increase v (increase Q), thin h (increase A), and tune f2 / f1 (disperse rupture location), which can achieve a stable output with d≤100nm.
[0034] Step 103: Control the fog output based on the optimal control quantity and the control quantity of the PID control.
[0035] As an optional implementation method in this embodiment, after adjustment by the PID controller, the state space model is updated, and based on the updated state space model, the optimization is performed again with the predicted values of atomized particle size, flow rate, and temperature for the next N steps as the optimization target, in order to minimize the comprehensive error function.
[0036] MPC is model-based planning, and its planning quality highly depends on the accuracy of the mathematical model. However, there are always deviations between the real world and the mathematical model: the model cannot cover all physical details (such as the minute aging of piezoelectric ceramics, trace impurities in liquids); model parameters (such as liquid viscosity, thermal conductivity) change slowly with time, temperature, and composition; and to facilitate calculation, the model is simplified by linearization, sacrificing some nonlinear accuracy.
[0037] PID (Programming Error Correction) disregards the model and relies solely on actual sensor data. When MPC (Multi-Planning Calculation) makes a "perfect plan" based on a slightly biased model, the actual system may have deviated from the predicted trajectory. The role of PID is to use the measured error to compensate for this "planning error," ensuring the system ultimately reaches the goal, rather than the model's "theoretical optimum." This is analogous to MPC relying on an accurate map (model) for navigation, planning the shortest path. However, in reality, road construction (disturbance) or map errors (model mismatch) can occur. PID, on the other hand, acts like the driver's eyes, immediately swerving around obstacles instead of blindly following incorrect navigation into dead ends.
[0038] MPC and PID work at different paces, forming a perfect "fast and slow combination": the MPC layer (slow pace, such as once per second) performs complex optimization calculations, which takes a long time, so it runs at a lower frequency (such as once per second) to plan the control baseline for the system for a period of time in the future.
[0039] The PID layer (fast-paced, such as 10-100 times per second) is simple to calculate, responds extremely fast, runs at a very high frequency, and constantly monitors errors and makes millisecond-level fine adjustments.
[0040] The entire system forms a high-level closed loop: the actuator executes the fused instructions → the atomizer produces a new atomization effect → the sensor measures the new state → the state is fed back to the MPC for the next round of prediction and optimization, and also fed back to the PID for real-time error calculation → and so on.
[0041] Real-world example: Handling a sudden increase in liquid viscosity. Suppose the viscosity of the atomizing fluid suddenly increases due to a cooling environment; the system would respond as follows: 1. Disturbance occurs: Increased viscosity causes atomized particles to coarsen, and the sensor detects that the particle size increases from 100 nanometers to 115 nanometers.
[0042] 2. PID First Response: The PID controller immediately detects the huge error of +15 nanometers. In the next high-speed control cycle (e.g., 10 milliseconds later), it calculates that a significant increase in amplitude and flow rate is needed to compensate, and immediately outputs a large compensation command in an attempt to "brake".
[0043] 3. MPC Replanning: Almost simultaneously, MPC also experiences a state change in its next optimization cycle (e.g., after 1 second). Its model understands the underlying cause of "increased viscosity." Therefore, it not only plans to increase amplitude and flow rate to cope with the current coarse particles, but also proactively plans to increase heating power to fundamentally reduce liquid viscosity. Its new "first control variable" output is a comprehensive optimization scheme.
[0044] 4. Collaborative Recovery: In the initial stage of disturbance, the rapid compensation of PID control dominates, quickly curbing the particle size expansion trend. Subsequently, the optimization scheme of MPC (including temperature increase) begins to take effect, and the system smoothly returns to the target particle size under the fine adjustment of PID. MPC ensures that the entire process is completed within constraints such as safe temperature and maximum amplitude, while PID guarantees the speed and stability of recovery.
[0045] The core of MPC is model-based multi-objective, multi-constraint, long-term optimization. It solves a "sequence" problem, and the output is a sequence of control reference values, typically only the first one is executed. The core of PID is error-based single-point, fast, local feedback. It solves a "current" problem; the output is an incremental compensation value, which is an instantaneous correction calculated based on the currently measured error. In actual control, the outputs of MPC and PID are dynamically and weighted to generate the final command. For example, u_final = α × u_mpc + (1-α) × (u_mpc + β × Δu_pid), where u_final is the final control command sent to the actuator (such as a piezoelectric ceramic driver or micropump). u_mpc: the optimal control quantity calculated by MPC; Δu_pid is the compensation increment calculated by PID based on the real-time error. α and β are dynamically adjusted fusion weights and gains.
[0046] The system operates smoothly, and the MPC model predicts accurately. At this point, α is large (close to 1), and the final instruction u_final is almost equal to u_mpc given by the MPC. The compensation effect of PID is suppressed to a small extent, only playing a role in "steady-state error elimination." The system is mainly driven by the strategic planning of the MPC.
[0047] In the event of a sudden disturbance / model mismatch, when the sensor suddenly detects a significant deviation in particle size (e.g., due to a sudden change in viscosity), it indicates a large bias in the MPC prediction. At this point, the controller rapidly decreases α and increases β. The incremental compensation Δu_pid of the PID controller is significantly amplified, becoming the dominant force in control and quickly suppressing the deviation. The system is then driven by the rapid response mechanism of the PID controller.
[0048] Layered control uses MPC and PID to control different variables. For example, MPC leads global optimization, planning the optimal combination of amplitude A, flow rate Q, and power P to control particle size d and mist flow rate F. PID assistance: Based on the flow rate Q planned by MPC, a high-speed PID loop precisely follows this flow rate reference to compensate for fluctuations in the micropump itself and changes in pipe resistance. In this case, the PID setpoint Q_setpoint is derived from the MPC output Q_mpc, but the PID directly controls the micropump driver.
[0049] As an optional implementation of this embodiment, determining the control quantity of PID control based on the given expected value of atomized particle size and the difference between the obtained adjustable parameters of the liquid to be atomized includes: determining the expected value of the corresponding adjustable parameter based on the given expected value of atomized particle size; and obtaining the adjustable parameters of the liquid to be atomized obtained by actual detection and the expected value of the adjustable parameters to determine the control quantity of PID control.
[0050] Among the above-mentioned optional implementation methods, the composite control strategy adopts a three-layer architecture of bottom-level PID execution + model predictive control (MPC) optimization + adaptive parameter adjustment. The core logic is that the bottom-level adaptive PID is responsible for fast response of single variables (such as basic regulation of flow rate and temperature); the middle-level MPC is based on a multivariable state-space model to continuously optimize the long-term performance of multiple objectives (particle size, flow rate, temperature); and the top-level adaptive mechanism identifies system model parameters online, corrects the MPC prediction model and PID parameters, and deals with time-varying characteristics (such as changes in liquid viscosity).
[0051] As an optional implementation of this embodiment, determining the control quantity of PID control based on the given expected value of atomized particle size and the difference between the obtained adjustable parameters of the liquid to be atomized includes: determining the expected value of the corresponding adjustable parameter based on the given expected value of atomized particle size; and obtaining the adjustable parameters of the liquid to be atomized obtained by actual detection and the expected value of the adjustable parameters to determine the control quantity of PID control.
[0052] The goal of adaptive PID is to quickly coarsely adjust single variables such as flow rate and temperature to compensate for the response delay of MPC. Its parameters (proportional coefficient Kp, integral coefficient Ki, and derivative coefficient Kd) are adjusted online according to system characteristics. Specifically, this involves determining the desired flow rate and temperature based on the desired particle size, acquiring real-time flow rate / temperature data, and determining the error based on these two parameters. and its rate of change; calculate the PID parameter adjustment amount using a fuzzy inference engine. Update PID controller parameters: , Similarly; output control quantity .
[0053] Taking flow control as an example, the input variable is defined as "flow error". and "error change rate" The output is "PID parameter adjustment amount". . big Positive, decrease To avoid overshoot, increase Acceleration Points; middle Negative, increase Improve response speed, increase Suppress oscillations; Small just, Damped high-frequency oscillation.
[0054] As an optional implementation of this embodiment, determining the optimal control quantity based on future values and predicted atomized particle size includes: taking the predicted atomized particle size, flow rate, and temperature for the next N steps as optimization targets, minimizing the comprehensive error function based on a pre-constructed discrete-time state-space model, wherein preset physical constraints are satisfied during the optimization process, and the optimal control quantity is obtained by optimization solution.
[0055] As an optional implementation of this embodiment, the comprehensive error function is: in, These are the weighting coefficients for particle size, flow rate, and temperature, respectively. λ represents the change in the input, and λ is the smoothing weight of the control input (to avoid frequent jumps). For predicting the time domain, d ref Q ref T ref , respectively, are the predicted values of particle size, flow rate, and temperature, and d(k), Q(k), and T(k) are the current values of particle size, flow rate, and temperature, respectively.
[0056] In this optional implementation, MPC is based on a state-space model and aims to minimize the comprehensive error function J by optimizing the predicted values of particle size, flow rate, and temperature for the next N steps (prediction time domain) while satisfying physical constraints (such as upper limit of amplitude and lower limit of heating power).
[0057] The physical constraints of the atomizer need to be transformed into mathematical inequalities and embedded in the optimization problem: Amplitude constraint; ; Frequency constraints: ; Heating power constraints: ; Flow constraints: .
[0058] MPC employs a predict-optimize-execute rolling model, where prediction utilizes the current state x(k) and... Predict the state in the next N steps During optimization, the constrained quadratic programming problem (QP) is solved to obtain the optimal control sequence. During execution, only the optimal control value at the current moment is implemented. And collect the state at the next moment. To update the model.
[0059] Furthermore, the method also includes an adaptive mechanism, including updating the state-space model: First, an extended Kalman filter (EKF) is used to estimate unmeasurable states (such as liquid film thickness h(k)), and the model matrices A(k) and B(k) are updated using recursive least squares (RLS). Specifically, the residuals are defined... ; Calculate the covariance matrix P(k) and the gain matrix Update model parameters Similarly, update .
[0060] The adaptive mechanism also includes PID parameter self-tuning (low-level parameter optimization). When MPC detects that the low-level PID control error continues to exceed the threshold, it triggers PID parameter self-tuning: the current operating condition is determined by clustering algorithm, such as "high viscosity liquid" or "low temperature environment"; the PID parameters are adjusted to the optimal value based on the historical operating condition-parameter mapping table (offline calibration) or online gradient descent method; after implementing the new parameters, the error is monitored to see if it converges. If it does not converge, it is re-optimized.
[0061] For example, assuming the liquid viscosity suddenly increases due to a decrease in temperature during atomization, the execution flow of the composite control strategy is as follows: Data acquisition and state estimation: The sensor detects a decrease in flow rate (due to increased flow resistance caused by increased viscosity) and the actual detected particle size is coarser (increased droplet merging); the current liquid film thickness h(k) is estimated to increase, and the state vector x(k) is updated; Adaptive PID fast response: When the flow error > 0 and the error change rate < 0 (error continues to expand), fuzzy rules trigger the "medium error + negative change rate" strategy, increasing Kp (improving flow regulation sensitivity) and decreasing Kd (suppressing high-frequency noise). The PID output uPID increases the micro-pump drive voltage, raising the flow rate Q(k) to near the target value. MPC multi-objective optimization is performed based on the updated state-space model. The particle size, flow rate, and temperature for the next 5 steps are predicted. The objective function J is optimized, and it is found that the amplitude A (to reduce the liquid film thickness h and promote breakage) and the heating power P (to reduce the viscosity μ) need to be increased simultaneously. Considering the constraints (Amax=10μm, Pmax=5W), the optimal control sequence is generated: ΔA(k)=+2μm, ΔP(k)=+1W.
[0062] The model is adaptively updated. After control is implemented, the state x(k+1) at the next time step is collected, and the residual e(k+1) is calculated. The RLS updates the model matrices A(k+1) and B(k+1) to correct the influence of viscosity changes on the model. If the error does not converge completely, the adaptive PID fast response and MPC multi-objective optimization are repeated until stability is achieved.
[0063] By employing a hierarchical collaborative approach combining MPC multi-objective optimization, adaptive PID fast response, and online adaptive mechanism, the composite control strategy effectively addresses the issues of particle coarsening and unstable mist output caused by structural limitations and time-varying liquid characteristics in nano-atomizers, achieving high-precision and high-stability atomization control.
[0064] As an optional implementation method in this embodiment, the pre-constructed discrete-time state-space model is as follows: ;in, These are the current particle size, flow rate, temperature, liquid film thickness, dual frequency, and amplitude, respectively. for Adjustments to the dual frequencies, amplitude, and power respectively; For the output vector, For process energy, To observe the noise, For a matrix, These are time-varying parameters.
[0065] In the aforementioned optional implementation methods, the dynamic characteristics of the atomization system are determined by the coupling of multiple physical fields, including liquid flow, vibration energy transfer, and heat conduction. Based on first-principles calculations and experimental data, a discrete-time state-space model is constructed.
[0066] Time-varying matrix , The key coupling item The effect of temperature on particle size; The influence of liquid film thickness Amplitude conversion factor.
[0067] , .
[0068] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.
[0069] According to an embodiment of the present invention, a nano-atomizer is also provided, comprising orthogonally arranged dual-voltage ceramic plates, a driving module for the dual-voltage ceramic plates, a piezoelectric ceramic micropump, a micro heating element module, and a control module; wherein, the control module obtains specified parameters from different components of the atomizer, including obtaining the driving frequency from the orthogonally arranged dual-voltage ceramic plates; obtaining the amplitude from the driving module for the dual-voltage ceramic plates; obtaining the micropump flow rate from the piezoelectric ceramic micropump; determining the heating power based on the micro heating element module; determining the predicted atomization particle size based on the specified parameters as input to a preset model constructed based on specified physical properties of the liquid to be atomized; determining the optimal control quantity based on the future value and the predicted atomization particle size; determining the control quantity for PID control based on the given expected atomization particle size and the difference between the obtained adjustable parameters of the liquid to be atomized; and controlling the mist output based on the optimal control quantity and the control quantity for PID control.
[0070] refer to Figure 2 A schematic diagram illustrating the composition of a nano-atomizer.
[0071] According to an embodiment of the present invention, an aroma diffuser is also provided, including a nano-atomizer.
[0072] According to an embodiment of the present invention, a control module is also provided, which obtains specified parameters from different components of the atomizer, including obtaining the driving frequency from orthogonally arranged dual-voltage ceramic plates; obtaining the amplitude from the driving module of the dual-voltage ceramic plates; obtaining the micro-pump flow rate from the piezoelectric ceramic micro-pump; determining the heating power based on the micro heating plate module; determining the predicted atomization particle size based on the specified parameters as input to a preset model constructed based on specified physical properties of the liquid to be atomized; determining the optimal control quantity based on the future value and the predicted atomization particle size; determining the control quantity of PID control based on the given expected atomization particle size value and the difference between the obtained adjustable parameters of the liquid to be atomized; and controlling the mist output based on the optimal control quantity and the control quantity of PID control.
Claims
1. A method for controlling the mist output of a nano-atomizer, characterized in that, Specific parameters are obtained from different components of the atomizer, including the driving frequency from the orthogonally arranged dual-voltage ceramic plates and the amplitude from the driving module of the dual-voltage ceramic plates; Obtain the flow rate of a micropump from a piezoelectric ceramic micropump; Heating power is determined based on the micro heating element module; The predicted atomization particle size is determined by using the specified parameters as input to a preset model constructed based on the specified physical properties of the liquid to be atomized. The optimal control quantity is determined based on the future value and the predicted atomized particle size; the control quantity of PID control is determined based on the given expected value of atomized particle size and the difference between the obtained adjustable parameters of the liquid to be atomized. The fog output is controlled based on the optimal control quantity and the control quantity of the PID control.
2. The method for controlling the mist output of the nano-atomizer according to claim 1, characterized in that, The preset model is: ,in, The surface tension of the atomized liquid is affected by temperature. , The surface tension of the liquid to be atomized at a specified temperature. For temperature coefficient, The density of the liquid to be atomized is affected by temperature. The flow rate of the liquid to be atomized is determined based on the micropump flow rate. The dynamic viscosity of the liquid to be atomized The length of the atomizer resonant cavity. For liquid film thickness, , Here is the conversion factor, and A is the amplitude. The coefficient is dimensionless. d is the driving frequency of the dual piezoelectric ceramics, and d is the predicted atomized particle size.
3. The method for controlling the mist output of the nano-atomizer according to claim 2, characterized in that, The control quantity for PID control is determined based on the difference between the given desired atomized particle size and the adjustable parameters of the liquid to be atomized. Determine the expected value of the corresponding parameter to be adjusted based on the given expected value of the atomized particle size; The control quantity for PID control is determined by obtaining the actual measured parameters of the liquid to be atomized and the expected values of the parameters.
4. The method for controlling the mist output of the nano-atomizer according to claim 3, characterized in that, Determining the optimal control quantity based on future values and predicted atomized particle size includes: The optimization objective is to optimize the atomized particle size, flow rate, and temperature predictions for the next N steps. Based on a pre-constructed discrete-time state-space model, the comprehensive error function is minimized, while satisfying preset physical constraints during the optimization process. The optimal control quantity is obtained by solving the optimization problem.
5. The method for controlling the mist output of the nano-atomizer according to claim 4, characterized in that, After adjustment by the PID controller, the state-space model is updated, and based on the updated state-space model, the optimization is performed again with the predicted values of atomized particle size, flow rate, and temperature for the next N steps as the optimization objective, minimizing the comprehensive error function.
6. The method for controlling the mist output of the nano-atomizer according to claim 4, characterized in that, The pre-constructed discrete-time state-space model is as follows: ； in, , These are the current particle size, flow rate, temperature, liquid film thickness, dual frequency, and amplitude, respectively. , These are the adjustment amounts for dual frequency, amplitude, and power, respectively. For the output vector, For process noise, To observe the noise, For a matrix, These are time-varying parameters.
7. The method for controlling the mist output of the nano-atomizer according to claim 5, characterized in that, The comprehensive error function is: in, These are the weighting coefficients for particle size, flow rate, and temperature, respectively. Let λ be the change in the input, and λ be the smoothing weight of the control input. For predicting the time domain, d ref Q ref T ref , respectively, are the predicted values of particle size, flow rate, and temperature, and d(k), Q(k), and T(k) are the current values of particle size, flow rate, and temperature, respectively.
8. A nano-atomizer, characterized in that, include: The orthogonally arranged dual-voltage ceramic plates, the dual-voltage ceramic plate drive module, the piezoelectric ceramic micropump, the micro heating plate module, and the control module; The control module obtains specified parameters from different components of the atomizer, including the driving frequency from the orthogonally arranged dual-voltage ceramic plates; the amplitude from the driving module of the dual-voltage ceramic plates; the micro-pump flow rate from the piezoelectric ceramic micro-pump; and the heating power based on the micro heating element module. The predicted atomization particle size is determined based on the specified parameters as input to a preset model constructed based on the specified physical properties of the liquid to be atomized; the optimal control quantity is determined based on the future value and the predicted atomization particle size; and the control quantity of PID control is determined based on the given expected value of atomization particle size and the difference between the obtained adjustable parameters of the liquid to be atomized. The fog output is controlled based on the optimal control quantity and the control quantity of the PID control.
9. An aroma diffuser, characterized in that, Including the nano-atomizer as described in claim 8.
10. A control module, characterized in that, Specific parameters are obtained from different components of the atomizer, including the driving frequency from the orthogonally arranged dual-voltage ceramic plates and the amplitude from the driving module of the dual-voltage ceramic plates; Obtain the flow rate of a micropump from a piezoelectric ceramic micropump; Heating power is determined based on the micro heating element module; The predicted atomization particle size is determined by using the specified parameters as input to a preset model constructed based on the specified physical properties of the liquid to be atomized. The optimal control quantity is determined based on the future value and the predicted atomized particle size; the control quantity of PID control is determined based on the given expected value of atomized particle size and the difference between the obtained adjustable parameters of the liquid to be atomized. The fog output is controlled based on the optimal control quantity and the control quantity of the PID control.